AIDS: The Second Decade (1990)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

AIDS and Adolescents The committee finds no credible evidence that the threat of HIV infection or AIDS will cease in the near future in the United States, as noted in Chapter 1. Therefore, the committee believes it important to sustain effective HIV prevention programs for young people at or before the age at which they begin practicing behaviors that risk transmission of this deadly virus. Serological studies of HIV infection, surveys of sexual and drug use behaviors, and reports from clinics for the treatment of Hug use and sexually transmitted diseases (STDs) all indicate that some young people begin practicing behaviors that risk HIV transmission during and in some cases before their early teens. By the end of the teenage years, the majority of young persons in America report having begun sexual intercourse, and one-half report some experience with illicit drugs.) Evidence from HIV seroprevalence studies conducted among patients admitted to 37 metropolitan hospitals during 1988-1989 suggests that the HIV prevalence rate is vanishingly small among 11-year-olds but begins rising at age 12 and continues to rise throughout the teenage years.2 These behavioral and epidemiological facts suggest that HIV pre- vention efforts should begin at least by early adolescence3 and that they l See, for example, Tables 3-7 and 3-11. 2 Dr. Michael E. St. Louis, HIV Seroepidemiology Branch, Center for Infectious Diseases, CDC, per- sonal communication, April 4, 1990. 3Adolescence, as heated in the psychological literature' is not synonymous with the teenage years. It is generally said to begin between 10 and 13 years of age and to end between 18 and 21 years (Santrock, 1981). This stage of life is characterized by significant physical, psychological, and social 147

148 ~ AIDS: THE SECOND DECADE should continue throughout this period. Federal funding of such pro- grams has ample precedent in that this country has supported programs to prevent other health problems of adolescents (e.g., unintended teenage pregnancy, drug use). The committee believes that the nation must recog- nize the importance of and unique opportunities for preventing the spread of HIV in the teenage population and that federal agencies must inter- vene accordingly. Adolescents deserve special attention because patterns of health behavior and risk taking are often established during the teenage years. By targeting prevention programs to adolescents, the United States may not only be protecting its youth but also preventing future problems in the adult population. Despite the obvious benefits associated with reaching adolescents, less is known about effective interventions for this age group than for adults. Much of the accumulated knowledge about AIDS prevention has been gained from adult programs, primarily programs for adult gay men. Adolescent health behavior is likely to be different from that of adults, however, and programs designed for adults require modification to accommodate the behavioral, social, and developmental diversity found in the adolescent population. In this chapter, the committee descnbes the scope of the AIDS prob- lem among adolescents, insofar as data on AIDS cases and HIV infection are available for this population. Subsequently, it reviews what is known about the distribution of nsk-associated behaviors in the adolescent pop- ulation, as well as the prevalence of sexual intercourse, condom use, and drug use and the confluence of these high-nsk behaviors. Finally, the committee considers what should be done to prevent further spread of HIV infection in this population. THE EPIDEMIOLOGY OF AIDS AND HIV AMONG ADOLESCENTS Before reviewing available data on the scope of the HIV/AIDS problem among adolescents, the committee notes the inadequacies of those data, a deficiency leading the committee to the conclusion that the precise degree of infiltration of HIV infection into the adolescent population is presently unknown. The relatively few cases reported to date among 13- to 19-year olds (see Table 3-1) do not accurately reflect the scope of the problem, nor should these data be taken as grounds for complacency. changes. Definitions of the exact time of entry into and exit from adolescence vary from study to study, depending on such factors as the theoretical view that has been adopted, the cultural context of the adolescent, and biological and social development factors, as well as the issue or problem of interest. (See Gold and Petronio [1980] for further elaboration of varying definitions of adolescence.)

AIDS AND ADOLESCENTS I 149 Rather, the pattern of the epidemic in the adolescent population described below should be viewed as an opportunity for primary prevention that should not be overlooked. AIDS case statistics are probably the most reliable epidemiological data base currently available. Yet the counts of current AIDS cases repre- sent HIV infections that were acquired several years before the diagnosis of AIDS was made. The current best estimates of the mean incubation period of the disease (i.e., the mean time between HIV infection and the onset of clinically diagnosable AIDS) are eight to ten years, but as data spanning more time become available and as more effective prophylac- tic treatments become available, it seems likely that this estimate will increased Furthermore, it is possible that the median incubation period for teenagers may be longer than eight years. Natural history studies of hemophiliacs infected with HTV suggest that the incubation period may be longer for chidden (not newborns) and adolescents than it is for adults (Goedert et al., 19891.s Yet even assuming that the incubation periods for adolescents and adults are equivalent, it is likely that few persons that were infected during their teenage years would also be diagnosed as AIDS cases during their teens. Even with the assumption of a median incubation period of eight years, fewer than one-half of persons infected with HIV at age 13 would be expected to develop AIDS dunng their teenage years,6 and even fewer of those infected in the late teens would develop AIDS before age 4The current estimate is that the majority of HIV-seropositive individuals will go on to develop AIDS, and it is not impossible that 100 percent of seropositive individuals may eventually develop full-blown disease (IOM/NAS, 1988:35-36); see also the projections of Lui, Darrow, and Rutherford (1988) and Longini et al. (1989). However, new evidence suggests that the incubation distribution of the disease may be quite different for children, adolescents, and adults, as discussed in footnote 5. SGoedert and colleagues (1989:1144, Table 3) report estimated annual incidence rates for AIDS after HIV infection as 0.83 per hundred for 1- to 11-year-old children; 1.49 per hundred for 12- to 17-year- old adolescents; 2.39 per hundred for 18- to 25-year-old adults; 3.40 per hundred for 26- to 34-year-old adults; and 5.66 per hundred for 35- to 70-year-old adults. The ages cited are the age of the person at the time of HIV infection. It should be noted that annual AIDS incidence rates are not uniformly distributed over time following infection. Thus, for example, rates are close to zero during the first two years following infection, and they rise during the next four to six years. It should also be noted that incubation periods may vary across transmission categories for adolescents. Indeed, some teens may progress from infection to disease more quickly than others. According to case reports, teens who acquired HIV infection through drug use or sexual behavior have progressed more rapidly than adults from infection to AIDS (K. Hein, Adolescent AIDS Program, Montefiore Medical Center, Bronx, N.Y., personal communication, 1989). Other data indicate that the median survival period is shorter for patients less than 20 years of age (9.0 months) than for patients between the ages of 20 and 29 years (13.0 months) or 30 to 39 years (13.2 months) (Lemp et al., 1990). 6It should be realized, of course, that bode the cumulative risk of infection and the rate of risky behav- iors can be expected to increase with age during adolescence (see the evidence presented below).

150 ~ AIDS: THE SECOND DECADE 20. Those persons who are diagnosed with AIDS during their teens will be drawn mainly from the group of persons whose incubation periods were markedly shorter than the median and who were infected during their early teens. Seroprevalence data for probability samples of individuals drawn from well-defined populations of epidemiological interest would provide a more reliable basis for inferring the prevalence of infection among teenagers. Unfortunately, with few exceptions, seroprevalence surveys conducted in this country have relied on samples of convenience, and most have not included teenagers. The largest samples that provide information on adolescents are derived from the routine HIV screening of applicants for military service and the Job Corps. These data cannot be generalized with knowable margins of error to other populations, but they can provide some insight into segments of the teenage population in which the infection may be established. HIV prevalence estimates derived from blinded testing of newborns for HIV antibody (i.e., CDC's neonatal surveillance activity, which is described in Chapter 1) provide a reliable indicator of the prevalence of infection among women delivering children. (Infants, whether infected or not, carry the maternal antibodies to HIV at birth if the mother is infected with the virus.) Analysis of these data by the age of the mother could provide important information about HIV infection among teenage women who bear children. Unfortunately, tabulations of HIV seroprevalence by mother's age are not presently available for most states. hndeed, the committee notes that, in some states, data on the age of the mother are not being collected. To provide better information about HIV infection and AIDS among adolescents, the committee recommends that the Centers for Disease Control make available to the research community AIDS-related data that permit separate consideration of teenagers and other age groups. Specifically, the committee recommends that: · data on AIDS cases be made available in a form that permits tabulation by specific ages or by narrow age groups (these data should be as complete as possible without threatening inadvertent disclosure of the iden- tity of any individual case);7 every state that participates in the neonatal surveillance . 7 The Committee on National Statistics at the National Research Council and the Social Science Re- search Council have jointly convened a panel to study the broad issues of confidentiality and data access in research. Their report will be available in approximately two years.

AI:DS AND ADOLESCENTS ~ 151 activity include the age of the mother coded in years or by narrow age group; and · CDC provide data from its family of surveys by specific ages or in narrow age groups, as well as by race, gender, and ethnicity. The Scope of the Problem AIDS Cases. As noted above, the small percentage of AIDS cases diagnosed in the adolescent population does not imply that AIDS and HTV are not a problem for teenagers. Indeed, as Verrnund and colleagues (1989) argue, a substantial fraction of the AIDS cases diagnosed among persons in their twenties reflect infections contracted during the teenage years. As of December 31, 1989, approximately 24,000 cases of AIDS had been reported among teenagers and young adults (ages 13-29~. Table 3-1 shows the distribution of reported AIDS cases by age at diagnosis, using the broad age categories into which CDC has coded the data released to the public. It can be seen that roughly one case in five is diagnosed among persons under the age of 30. The proportion of cases actually diagnosed among teenagers, however, is small. Figure 3-1 displays the case counts by age at diagnosis for persons diagnosed with AIDS between the ages of 13 and 29. Allowing, as noted earlier, for an incubation period that is rarely less than two years and a mean incubation period that may be eight years or longer, one would expect that nearly all of the AIDS cases diagnosed among persons in their very early twenties would reflect HIV infections contracted during adolescence.9 The overall impact of AIDS on teenagers and young adults is reflected in the fact that AIDS was the tenth leading cause of death among 15- to 24-year-olds as early as 1984; it had risen to the seventh leading cause of death for this age group in 1986 and the sixth in 1987 (Kilboume, Buehler, and Rogers, 1990~. The CDC's family of surveys collects data on HIV infection from several subpopulations, including clients attending drug treatment, STD, tuberculosis, and prenatal clinics, patients at general hospitals, and newborn infants. With the exception of the survey of infants, all of the other surveys rely on samples of convenience. Data collected through this program are intended to provide information on the prevalence and incidence of infection in selected populations, to provide early warning of the emergence of infection in new populations, and to target intervention programs and other resources. 9 It may seem intuitively appealing to argue that at least one-half of the 4,268 cases diagnosed at age 28 reflect infections contracted during the teens. This argument is not, however, logically required. Because there is a non-zero probability of an AIDS diagnosis being recorded in each year from roughly two years after HIV infection, it would be theoretically possible (with a suitably large number of HIV infections among persons in their early twenties) to observe 4,000 cases of AIDS among 28-year-olds, none of which had been contracted by teenagers.

152 ~ AIDS: THE SECOND DECADE TABLE 3-1 Distribution of AIDS Cases Reported Through December 31, 1989, by Age at Diagnosis Diagnosis No. of (years) Cases Percentage < 1 785 0.7 1-12 1,210 1.0 13-19 461 0.4 2~24 5,090 4.3 25-29 18,966 16.1 3~39 54,334 46.1 40~9 24,951 21.2 50t 11,984 10.2 Total 117,781 1OO.Oa Percentages may not sum to 100.0 because of rounding. SOURCE: Special tabulation provided by the Statistics and Data Processing Branch of the AIDS Program, Centers for Disease Control. 6,000 5,000 4,000 oh 111 an C ~ 3.000 Oh 2,000 1 ,000 4,996 2,714 1 ,979 34 38 67 90 188 ~1 863 ~ 289 3,194 3,794 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 AGE AT DIAGNOSIS FIGURE 3-1 Number of AIDS cases reported among 15- to 30-year-olds through December 31, 1989, by single years of age at diagnosis. SOURCE: Special tabulation provided by the Statistics and Data Processing Branch of the AIDS Program, Centers for Disease Control.

AIDS AND ADOLESCENTS | 153 HIV Seroprevalence. Since October 1985, the Department of De- fense has tested applicantsl° for military service for evidence of HIV in- fection; in January 1986 the armed forces began screening all active-duty personnel. The crude prevalence rate for 17- to 19-year-old applicants screened between October 15, 1985 and March 31, 1989 was 0.34 per 1,000 (Burke et al., 1990~. HIV prevalence rates among military appli- cantsli from October 1985 to March 1986 increased directly and linearly with age from 0.25 per thousand among 18-year-olds to 4.9 per thou- sand among 27-year-olds (Burke et al., 1987:132~.12 Among active-duty personnel during the period October 1985 to July 1989, prevalence rates were found to have a similar distribution by age. The observed rates were lowest among soldiers less than 20 years old (0.5 per thousand), peaked at 3.4 per thousand among 30- to 34-year-olds, and gradually declined among older military personnel (Kelley et al., 1990~. Figures 3-2a and 3-2b plot these age-specific prevalence rates for military applicants and active-duty personnel. The plots demonstrate that roughly parallel age trends are found for the two groups. Data tabulated separately for applicants from the New York-New Jersey metropolitan area suggest a similar trend. The prevalence of HIV among applicants from the New York-New Jersey metropolitan area is several times higher than the rate among applicants from the rest of the nation (see Figure 3-2c). In this case, prevalence rates among military applicants 30 years of age and older are somewhat higher than those of applicants aged 26-30. The other large group of young people who are routinely screened for HIV are persons applying to participate in the Job Corps. The Job Corps is a federal program that provides training and employment for so- cioeconomically and educationally disadvantaged youths. Of the 69,233 applicants between the ages of 16 and 21 years who were screened be- tween October 1987 and November 198S, 3.9 per thousand were infected (St. Louis et al., 1989~. This rate is much higher than the prevalence 10Among military applicants who had been screened between October 1985 and March 1986, 86 per- cent were male and 76 percent were white; 46 percent were less than 20 years old, and only 5 percent reported education beyond a high school diploma (Burke et al., 1987) 11The observed age-specific prevalence estimates largely reflect the rates among older males (434 seropositive among a total of 263,572 men). Overall, HIV prevalence was lower among women, and fewer were tested (26 seropositive among 42,489 women). However, among 17- and 18-year-olds, the prevalence rates for males and females were approximately equal (Burke et al., 1987). Other analyses of military data find prevalence rates among teenage females to be comparable to those among teenage males (Horton, Alexander, and Brundage, 1989; Kelley et al., 1990). 12The authors concluded that HIV prevalence increased linearly with age between 18 and 27 years (Burke et al., 1987:132). Estimated prevalence among males over the age of 27 was lower, however, than the prevalence rates observed among male applicants between 25 and 27 years old.

154 ~ AIDS: THE SECOND DECADE 6 r a' 4 Q - LLJ He 3 LL > 2 UJ cr ILL > I 1 o ~ Males - /\ \/ _,~ Females / `1/ , , , 1 1 7 1 8 1 9 20 21 22 23 24 AGE 25 26 27 28 29 > 30 FIGURE 3-2A Prevalence of HIV (rate per 1,000 persons), by age, among male and female applicants for military service [N = 306,061] (October 1985 through March 1986). SOURCE: Burke et al., 1987. 3 Q - LLJ Cal A LL G 1 > I o 17-19 20-24 25-29 4 30-34 35-39 40 ~ AGE FIGURE 3-2B Prevalence of HIV (rate per 1,000 persons), by age, among active duty military personnel [N = 1,752,191] (January 1987 through April 1988). SOURCE: Peterson et al., 1988.

AIDS AND ADOLESCENTS ~ 155 Males - a' 2.0 a) Q LLJ A LL ~ 1.0 0,0 Females 1 1 < 20 21-25 AGE GROUP 26-30 > 30 FIGURE 3-2C Prevalence of HIV (percent), by age, among male and female applicants for military service from New York-New Jersey metropolitan area [N = 44,139] (October 1985 through June 1987). SOURCE: Brundage et al., 1988. rates found among the youngest groups of applicants for military ser- vice, which may reflect differences in the populations represented in applications to these two organizations. HIV Seroprevalence in Childbearing Women. Anonymous anti- body testing of newborn infants provides information on the prevalence of HIV infection among childbearing women (because infants circulate maternal antibody during the first months of life). As noted earlier, data are not available for every state, and some states do not provide infor- mation on the age of the mother. Data have been published for New York City and the rest of New York State (Novick et al., 1989a). Among babies born in New York State between November 1987 and November 198S, the seroprevalence rate was 1.6 per thousand outside New York City and 12.5 per thousand for births in New York City. Figure 3-3 plots the age-specific rates of infection found in New York City. Although these data show an age trend similar to that found in other studies, the rates of infection even among teenage mothers are substantial.~3 A1- most 10 per thousand or ~ percent of black teenagers who delivered children in New York City during this period were infected with HIV. Reanalyses of ZIP Code-specific areas with high rates of drug use (determined by comparing rates of drug-related hospital discharges) found rates as high as 40 per thousand or 4 percent in some ZIP Codes areas of New York City (Novick et al., 1989a).

156 ~ AIDS: THE SECOND DECADE 35 o o o ~ 30 a' Q AL 25 a: by ILL o CC lo Oh 20 15 10 o _ , _ White Black Hispanic 32.2 _ 1~ N~f 20.3 17 14.7 < 20 20-29 30-39 40 + AGE OF MOTHER FIGURE 3-3 HIV seroprevalence rates (per 1,000 persons) among New York City women giving birth between November SO, 1987, and November 30, 1988. Of 12S,120 New York City women tested, race/ethnicity was unknown for 2,461, and age was unknown for 2,334. SOURCE: Novick et al., 1989a:Table 4. The prevalence of HIV infection among Hispanic teenage mothers was almost as high. In considering Figure 3-3, the committee would note that it is only the contrast with rates of almost 30 per thousand found among 30- to 39-year-old black mothers that make the observed teenage prevalence rates appear relatively low. HIV Seroprevalence in Other Studies. The seroprevaTence rates provided above on women bearing children in New York City are but- tressed by emerging data from serosurveys of nonprobability samples of hospital patients. Ernst and colleagues (1989), for example, recently reported that, among patients aged 15 to 24 at the Bronx Lebanon Hos- pital, 36 per thousand or 3.6 percent of males and 25 per thousand or 2.5 percent of females were infected with HIV. Table 3-2 summarizes se- lected results of this and other small-scale seroprevalence studies among teenagers and young adults. Although the observed prevalence of HIV infection varies from population to population and from study to study, one finding is clear: the AIDS virus is substantially seeded in some segments of the adolescent population. Moreover, although the variation in estimates argues for more comprehensive and standardized monitoring of the spread of infection in this population, the lesson for prevention is nonetheless apparent. HIV infection is already spreading in the teenage

AIDS AND ADOLESCENTS ~ 157 population, and in some locales and population subgroups, the spread has been substantial. Variation of AIDS and HIV Prevalence by Gender. Just as more AIDS cases are seen among adult males than among adult females, so too are there more cases of AIDS among teenage boys than among teenage girIs.~4 Of all AIDS cases reported as of December 31, 1989, males outnumbered females by a 9:1 ratio.is As the first row of Table 3-3 shows, however, the ratio of male to female cases is much lower among teenagers than among adults in the United States. Thus, it can be seen that the male-to-female ratio, which is roughly 1:1 among cases diagnosed in infants (younger than 1 year), increases to 4:1 among teenagers, to 6:1 among 20- to 24-year-olds, and ultimately to 16:1 for cases diagnosed among persons in their forties. In addition to variations by age, the ratio of male-to-female cases of AIDS and HIV infection varies substantially among populations. Eleven percent of AIDS cases among teenagers have been diagnosed in the New York metropolitan areas and the male-to-female ratio for those cases is approximately 2:1 (Table 3-31. The size of this ratio indicates that girls in the New York area are supporting a greater burden of disease than girls nationally.'7 HIV infection rates calculated for military applicants As of July 1, 1986, there were more than 35 million individuals in this country between the ages of 10 and 19 years, constituting approximately 14.6 percent of the total population. The ratio of males lo females is essentially 1:1 (1.044:1) (U.S. Bureau of the Census, 1987:17). 15As of December 31, 1989, females of all ages accounted for 11,524 cases of AIDS out of a total of 117,781 reported cases. 16As of December 31, 1989, a total of 461 AIDS cases had been reported among 13- to 19-year- olds; 50 of these cases were from the New York Primary Metropolitan Statistical Area. In 1983, the Office of Management and Budget changed the classification system used to define metropolitan areas in federal statistical reports. The basic tern' for such areas, formerly known as Standard Metropolitan Statistical Areas or SMSAs, was changed to Metropolitan Statistical Areas or MSAs. The basic concept of a metropolitan area, however, remained that of an area with a large population nucleus and the adjacent communities, e.g., suburbs, that have a high degree of economic and social integration with that nucleus. Within metropolitan complexes of one million or more population, separate Primary Statistical Areas (PMSA) may be designated under the new classification system. Any metropolitan area containing one or more PMSAs was, in turn, designated a Consolidated Metropolitan Statistical Area (CMSA). The New York PMSA, for example, includes the five counties that comprise the City of New York, plus the suburban counties of Westchester, Rockland, and Putnam. Adjacent areas include other PMSAs, e.g., the Nassau-Suffolk (Long Island) PMSA. The New York PMSA and the Nassau- Suffolk PMSA are 2 of the 12 PMSAs that constitute the Consolidated Metropolitan Statistical Area that bears the awkward name: "New York-Northern New Jersey-Long Island, NY-NJ-CT CMSA." (U.S. Bureau of the Census, 1985, Appendix 2). 17The higher proportion of cases among females of all ages in New York reflects the greater relative proportion of cases attributable to IV drug use and heterosexual transmission in the Northeast than other areas. (See for example, Figure 1-1 in Chapter 1.)

158 Cal - . ~ c5 V) - ._ no Pa o Ad o Cal Cal m Cal .O - V: C) C) Cal C) a ._ cq - Cal cat - Cal Cal ~0 - o To 50 C) & o ._ C) ~ ~ C:: Cat U) CO) ~0 .O ~ C) Cal Cat C) 1 m - PA := _` 00 00 ON - _ at 00 - en a_ Cat 00 _ ~ ~ — 00 C) _ C,> Ct ~ ~ O ~ _ _ _ _ ~ 0 ~ ~ ~ 00 0 00 _ ~ C) _ ~ c: ~ _ Cal Cal A:; ~ O O .= ~ ~ US ~ ~) ;— Ch · _ 00 ~ O Em ~ U. _ C) C) c~ O ._ ~ _ C5 C) C~ o ~ . O _ ~ V, ~ ~ ~ ~ ~ ~ oo ~ ~ =: _ _ _ Ct C~ C) o C) a o C~ C) ~ C) — C~ 00 ~ . Z - ~ _ 00 _ ~ O ~ ~ C ~ ~ ~ ~0 ;~ ._ ~ 0C - ~ 1 _ — _ ~ — C~ 00 o C5 ~ ~ Ce O O - C~ U: O ~ C) C) _ c: C) _ C~ ~ _ C,) C: 0 0 5 C~ ~ ~ - - ~4 S: ._ C~ l l _ _ 5 ~ ~ C) o4 0 a: et C~ - - C~ C) U) — — £ O_ _ ._ ._ ~ ~: ~ O O 3 ~ o Z Z a: - - o £ ~o C) o C) Ct Ct - £ - o C) ~o oo oo ~ C~ °° oo oo s~ oo ¢N ~ oo r_ C~~ ~ ~ (~ 0 ~ ~ ~ Z Z° ~ C 1 1 0 ~ — 3 1 t_ t_ 1 c- oo ·= C~ oo =\ ¢\ oo _ _ ~ .< CS~ _ _ ~ O _ [L, [L O V) ~ ~ C)

AIDS AND ADOLESCENTS ~ 159 TABLE 3-3 Percentage of AIDS Cases Reported Through December 31' 1989, Diag- nosed Among Males and Male-to-Female Case Ratios, by Age at Diagnosis, for All Cases Diagnosed in the United States and in the New York Metropolitan Statistical Area (MSA) Age at Diagnosis (years) Population, Statistic < 1 1-12 13-19 2~24 25-29 3~39 40~9 50+ United States % Male 49 57 80 M:F Ratio 1.0:1 1.3:1 4.0:1 5.7:1 (Base N): (785) (1,210) (461) (5,090) New York MSA % Male 52 M:F Ratio 1.1: 1 (Base N) (229) 54 1.2:1 (311) 85 70 76 2.3:1 3.2:1 (716) (50) 88 91 7.3:1 10.1:1 (18,966) (54,334) 79 85 92 3.8:1 5.7:1 11.5:1 94 91 15.7:1 10.1:1 (24,951) (11,984) 92 11.5:1 (3,077) (10,799) (5,209) (2,274) SOURCE: Special tabulation provided by the Statistics and Data Processing Branch of the AIDS Program, Centers for Disease Control. from the New York City area yield male-to-female ratios of roughly 1.4:1 (Brundage et al., 1988; Horton, Alexander, and Brundage, 19891. Among Job Corps applicants, the male-to-female ratio nationally is approximately 1:1 among 16- to 18-year-olds, but for applicants between the ages of 19 and 21, the ratio climbs to approximately 4:1 (St. Louis et al., 19891. Overall, the number of infected men is approximately equal to the number of infected women among 17- to 19-year-old military applicants (~.09:~), but among 17- and 18-year-old applicants, fewer men than women were found to be seropositive (0.9:1) (Burke et al., 19901. Surveys of clinic populations have found roughly equivalent rates of infection among teenage males and females. T. C. Quinn and cowork- ers (1988), for example, screened anonymous blood samples from 943 consecutive patients at STD clinics in Baltimore; among 15- to 19-year- old patients, 2.5 percent of 434 female patients tested positive for HIV, compared with 2 percent of 509 males in this age group. Among young adults (aged 20 to 24) attending the same clinics, the prevalence of HIV infection was higher but remained roughly equal for males and females (3.4 percent of 385 females versus 3.8 percent of 840 males). Among older patients, males were two to three times more likely than females to be infected. Among college students in a blinded serosurvey of a Among 25- to 29-year-olds, Quinn and coworkers (1988) found that 2.9 percent of 239 females and 6.9 percent of 598 males were infected with HIV. Among patients 30 years of age and older, the Quinn team found that 4.3 percent of 185 females and 11.4 percent of 636 males were infected with HIV.

160 ~ AIDS: THE SECOND DECADE TABLE 3-4 Distnbution of AIDS Cases (percentage) Reported Through December 31, 1989, by Race/Ethnicity Race/Ethnicity Age Group (years) White Black Hispanic Othera Total N ~12 21.9% 52.6% 24.6% 0.9 100% 1,995 13-19 43.4 35.8 18.4 2.4 100 461 20-24 48.9 31.6 18.5 1.0 100 5,090 25-29 53.4 28.6 17.1 0.9 100 18,966 30 + 57.9 26.3 14.8 1.0 100 91,269 All ages 56.1 27.4 15.5 1.0 100 117,781 aThis category includes persons of unknown ethnicity (IV = 226), Asians/Pacific Islanders (N = 596), and American Indians/Alaskan nanves (N = 121). SOURCE: Special tabulation provided by the Statistics and Data Processing Branch of the AIDS Program, Centers for Disease Control. nonprobability sample of persons visiting clinics at 17 colleges, the vast majority of HIV infections (87 percent) were found among males (Gayle et al., 1989~.~9 Variation by Race/Ethnicity in AIDS anti HIV Prevalence. A disproportionate share of the burden of adolescent AIDS cases is borne by minority youth. Table 3-4 displays the racial composition of AIDS cases for venous age groups. As shown, 36 percent of teenage cases have been diagnosed among black teenagers, who make up only 15.3 percent of the population; 18 percent of cases are found among Hispanics, who constitute less than 10 percent of the national teenage population.20 The table also shows that, compared with adult cases, a larger proportion of adolescent cases have been diagnosed among blacks and Hispanics (54 percent of teen cases versus 41 percent of cases among persons 30 years of age and older). Data on HIV infection also indicate a greater prevalence of HIV in- fection among blacks and Hispanics than among whites. Among teenage military applicants, seroprevalence rates are highest among blacks; the prevalence rate for black females (0.77 per 1,000) was four times greater 190f 13,810 specimens tested, 23 were positive for HIV, for a prevalence rate of 0.17 percent or 1.7 per thousand. Twenty of these 23 specimens were from males. 20The racial distribution of the teenage population is similar to that found among adults: 81 percent of 10- to 19-year-olds in the United States are white, 15.3 percent are black, and slightly less than 10 percent (9.5 percent) are Hispanic (U.S. Bureau of the Census, 1987:16-17). Because Hispanic persons may be of any race, the percentages noted here reflect overlapping groups (and thus in total exceed 100 percent). The Hispanic category includes anyone self-identified as Mexican, Puerto Rican, Cuban, Central or South American, or of other Spanish/Hispanic ongin.

AIDS AND ADOLESCENTS ~ 161 than that for white males (0.18 per 1,000) (Burke et al., 19901.2i Sero- prevalence surveys of teenage mothers of newborns in New York City found that 0.77 percent of Hispanics, 0.81 percent of blacks, and 0.17 percent of whites were infected (Novick et al., 1989a) (see Figure 3- 31. Among female military recruits, seroprevalence rates were highest among black, non-Hispanic women (Horton, Alexander, and Brundage, 1989; Burke et al., 19901. Among male applicants, the relative excess of cases among blacks is greatest for those men from the Northeast and north central regions of the country (Sharp et al., 1989~. Among Job Corps applicants, the rate of infection among blacks was approximately five times higher than that for whites (0.7 versus 0.14 percent), and the ratio of male-to-female cases was much closer to unity for blacks (1.2:1) than for whites (7.~:1) (St. Louis et al., 1989~. Mode of Transmission for AIDS Cases. Table 3-5 shows the distribution of all AIDS cases by gender and risk category. These data suggest that a substantial proportion of cases among adolescent boys and young men are related to homosexual contact: 37 percent of teen cases and 68 percent of cases among 20- to 24-year-olds. Yet the largest risk category for teenage boys is exposure to contaminated blood and blood products, which accounts for 44 percent of diagnosed cases among 13- to 19-year-olds. The relatively high rate of cases among teenage boys resulting from transmission through blood products is explained in part by hemophilia, a sex-linked genetic disorder that is almost exclusively a problem of males. The distribution across risk categories of AIDS cases diagnosed in teenage girls is different from that seen among teenage boys. Hetero- sexual contact accounts for 37 percent of cases among teenage females and 41 percent of cases among women 20 to 24 years of age; 28 and 40 percent of cases, respectively, are ascribed to IV drug use. Among women who have been diagnosed with AIDS, females between the ages of 13 and 24 years are more likely than older women to report exposure through heterosexual contact. Geographic Variation in AIDS and HIV Prevalence. The geo- graphic distribution of AIDS cases among teenagers and young adults is shown in Table 3-6. Teenage and young adult AIDS cases are found throughout the country, and the overall distribution of young adult AIDS cases does not show marked deviations from the distribution of adult 2~0f the 1,141,164 military applicants aged 17 to 19 screened for HIV infection between October 15, 1985, and March 31, 1989, 393 were found to be infected, giving an overall prevalence rate of 0.34 per 1,000 (Burke et al., 1990). Crude seroprevalence rates (per 1,000 applicants) were 1.00 for blacks, 0.29 for Hispanics, and 0.17 for whites.

162 ~ - C~ C) - o Cat on 6 Cat o Cat en o Cat so o 6 oo G) C) ~: oo o E~ . C) _ c) 6 oo e~ s: ~ ~ V C~ o V: P_ — ~ V ~o o~ ._ o U' _ ~ ~D t I '~ ~ ~ C'~ ._ _Q o m ~ .= E~ "o V ~ 00 ~ ~ ~ ~ _ _ ON ~t C~ ~ _ 2 ~ ~ ~ ~ _ ~ ~ ~ ~ U~ ~ ~ o _ oo o _ o 8 8 o o 8 8 8 8 8 8 ~ ^ - Ct ~o C) _ ~ CD _ - o~ o ._ ~ 3 CO _ ~ o ~ o 5 O _- O ~ _ _ C~] ;> - X C) C~ O ~ o :~: c~ . - o C4 . - c~ ~> ~ ~ ~ - o ~ . . ~ . - o o - ~ ~ o ~o ~ O oo ~ ~D - ~ ~o ~ c~ oo o - ~ ~ oo oo - ~ . . . . . ~ ~ ~ ~ '] ~t ~ ~ oo - c~ ~ ~ ~ . ~ ~ o ~ ~ ~ ~ o ~ ~ ~ - ~ - ~ - d c~ d — ~ ~ oo v~ ~ C~ O ~ - - - ~ ~ ~ o ~ ~ . ~ ~ . ~ - o ~ ~ 'e 'e 'c ~ 'e — — ~ c~ — C~ d' ~ °O ~ . ~ ~ ~ ~ - ~ ~ ~ - - - - ~ ~ ~ o ~ ~i ~ ~ ~ v~ ~ ~ ~ ~ ~o 00 00 ~ ~ C5 CC Ct CO Ct C: ~ - ) ~ C5 c~ v, ~ v' ~ ~ ~ + ~ ~ l ~ l ~ - ~ ~ u~ o — <5 ~ v~ o — 6 ~ ~ ~ <: ~ _ ~ ~ ~ ~ C) :{s ·= oc ~ Ct .' ~ r: fi s c, = ~ ~ c) ~ c~ o o ~ ~ - - x ~ . ·= c~ c,, u, ~ ec -~c x ~ ~ c) o c - ~ o co x ~ o c) -- ~ ~ x ~ _ ~ S C .`a ~ _ 5 ~ ~ ._ ~ ~ C _ c D 0= ~ S C D u, e ~D <,, O ·— _ C~ [_ _ C) ~ _ — 'S ~ . _ O as ~ O ._ — CC _ C~ O X '`— ~ — o ';? O_ ~ O ._ ~ ', ~ ,= X o 3 s ~ D ,= O — ~o U, C) C ~ O ~ c5 0 C ~ C _2 ~ o.' _ = e E -0 - - = ~A '= c, - v°, e; ~ ~o7 7 o ~ =.,_ CC- ~ .= C) =, E ~ ~ o co C.) — E ~ ° ~ c Z ~ ~ ~ >. O - o C) oc Ct o U, o ~ _ _ o ~: o ~o ._ U: U: C) o CC - - Ct C~ - .= - U, C) - D ~: ._ o o - C~ - D - Ct ._ U,

AIDS AND ADOLESCENTS ~ 163 AIDS cases. There is more geographic variation in the distribution of teenage cases, but the total number of cases diagnosed in adolescents (461) is still quite small. The largest observed deviation between the adult and teenage distributions is for cases outside metropolitan areas (43 percent of teenage cases versus 27 percent of cases among adults aged 30 and older.22 Behind the statistics presented in Table 3-6 are localized areas with very high prevalence rates of AIDS. Gayle, Manoff, and Rogers (1989) report, for example, that more than one-half of the AIDS cases among 13- to 19-year-olds came from five states and Puerto Rico.23 Similarly, as noted earlier, only 3.6 percent24 of the U.S. population lives in the New York metropolitan area, but 11 percent of teenage AIDS cases were from New York. The AIDS case data indicating a concentration of adolescent AIDS cases in urban areas, especially New York City, are paralleled by screen- ing data on infection rates among applicants for military service.25 Figure 3-4 shows the distribution of the prevalence of infection among individu- als who applied to enter the military between October 1985 and September 1989. The elevated peaks indicate pockets of infection, largely around the coastal areas of the country. However, cases of HIV infection have been identified among 17- to 19-year-old military applicants from 200 counties in 41 states and the Distnct of Columbia (Burke et al., 1990~. HIV prevalence among Job Corps applicants also shows considerable geographic vanation, although the variation appears to be race specific (St. Louis et al., 1989~. Seroprevalence rates among black and Hispanic lob Corps applicants were highest in the Northeast (10.2 and 7.7 per 22Hemophilia- and transfusion-associated AIDS cases comprise a larger proportion of the cases re- ported among teens than among older persons Hemophilia- and transfusion-associated AIDS cases are also more geographically dispersed than those associated with sexual and IV drug-using behav- iors. To examine whether these factors might account for the smaller proportion of AIDS cases re- ported in nonrnetropolitan areas by adults (versus teens), Table 3-6 was recomputed excluding these cases. The result is only slightly attenuated: 37 percent of teen AIDS cases (excluding hemophilia- and transfusion-associated cases) were reported in nonmetropolitan areas versus 27 percent of cases reported among persons aged 30 and older. 23This includes both the metropolitan and nonmetropolitan areas of these states. 24In the 1980 Census of the Population, slightly more than 8 million people were counted in the New York, New York, primary metropolitan statistical area (PMSA) out of a total U.S. population Of 226 million. (The New York PMSA includes the city of New York plus the counties of Westchester, Rockland, and Putnam). 25 Because these rates include applicants of all ages, caution must be exercised in generalizing from them. Although it iS known, for example, that teenagers and young adults constitute the vast majority of military applicants, HIV prevalence rates in this population are highest among persons in their late twenties.

164 - o ~ cs~ ~ M ;\ ~ oo - ~ {~N - c~ . - c~ o oC c~ a - Ct z C5 r: o ._ 0 00 G~ - S~ C) C~ C~ S 0D - S C) ~: - C~ C~ - s U. C~ CC a o L~ a: e~ o :e I _ U. O Ct z ° ~ - C) g 8 g g g o ~ _ ~ oo C~ ~ ~ U~ . . . . . ~ ~ ~ C~ ~ _ _ _ _ _ - o U) ~ ~o _ ~ r~ V~ _ C~ ~ C~ ~ ~ o ~ 3 _ _ c~ _ c: _ C~ Ct Ct - - o o - C) "o .O — . ~ - o z Cq . C': _ o Ct ~ C, 04 o~ .= ~o o ~ ~ _ . . . . . 00 ~ ~D ~o ~o - o o U~ o ~ ~ oo . . . . . o~ _ _ _ - U: C, o~ _ ~ ~ -?- Ce 1 1 1 _ ~ o U~ o — _ ~ ~ C~ ~ C) C) - ~ D ~ CC ~ o _ C) _ — S C) O — ~ . _ C _ C~ C) Le S o ~ 5 O ~ 3 O'- O _ ._ E- _ ._ O o o~ C~ _ O - 06- & C~ Ct: ~C _ _ _ _ o ~ O C) _ _ ~ S V' ~ ~ - ~— C; O .= & C~ O ~ 3 ._ Cq - 2 oc ~ — q) _ ~ _ C~ o ·— C~ £o >-o 0 z & ~- c.) 0 c ~ — _ 5 C) _ = ~ = m ~ .Q ~ 3 ~ - ~o o — o C.> . c ~ 3 ~oL>c Co o~ t~ o0~ O Z C~ C~ - . _ o C) . o V C) C) o o ~, ~ D c5 c C) z o V o ao- ._ · - S >: O Z _ =: _ O ~ C) 3 ~ ~ . Z s o - C~ ~ t— _ ._ o ~ ~ .' co V ~ - ~ = V' _ ;n ~ - 7 0 C4 ._ ~ _ ._ — S O ~ V .. _ ~— V . _ . C~ ~ U, ~ - ~ O - c U, - 5 . . O c¢ oo — S ~ C ~ ._ z 5 _ ~ Z '- et .O V, .= ~ S CL & 4.) 00 . _ m .. c~J U) ~ C _ =0 ~ Z O ~ Z ~0 s ': X X S ~— U~ C ~ 0 5 C)- C . _ _ Ct _ C) C V C~ _ O ~— mt ~ .Y C _ O ~ .0 ·- C) O . _ o ~ V ~ C - V, ~— _ . _ ~ Ce ~ 6 V - — ) ~ Co ~ C) cc ~ 13 ~ Z .= a=. a ._ C~ ~ _ ~ ~ . _ v _ _c' 3 .~ o ' . ~ s ~ _ _ _ o ~ — C5 c ~ _ , . _ V CL ~ o o _ C~ C _ CC . ~ U, — — t15 c, o ,,4 . _ = ~ C~ 3 E m~ · ,_ 0 ~ ~ ~ . 4) ; ~ c, _~ ._ _ _ c,~ _ t5 ~ ~ c . c~ 00 z- o c~ _ ·_ ·: c~ c . .= ~ — <: . C2: - c~ 0 .: c~ c~ _ ._ c ~ v . . x _ 00 ~ 0 0 ~ c~ c~ 6 . X ,~ s ~ _ s o o s~ C~ _ Lc - CO o os c ~: C) - Ct ~: C) C5 C) ~1 ~ Ct - ." - C~ - >- D : ~ C) Z ~ - o - Ct - .D - - ._ · C) _ & ~;L V~ _ . . ~ O O ~

AIDS AND ADOLESCENTS ~ 165 ID ~ _ ~~,~~ A ~ — , ~ 1 ~ ~ _~ ~' `~C: ~ FIGURE 3-4 Map of number of HIV-infected applicants for military service ages 17 to 25 by county, October 1985 through September 1989. Counts are plotted by county, but they were not plotted for counties with 10 or fewer seropositive military applicants. SOURCE: John Brundage, Division of Preventive Medicine, Walter Reed Army Institute for Research. thousand, respectively), whereas rates for white applicants were highest in the West (1.9 per thousand).26 Pockets of High-Risk Youth. As implied by the data presented earlier, not all teenagers are equally likely to come into contact with the AIDS virus. Urban-dwelling adolescents, particularly minority youths and disadvantaged teens, appear to be at increased risk for this health threat. The limited HIV seroprevalence data available suggest that run- away youth appear to be particularly at risk. A survey of more than 1,300 homeless youth seeking medical services in New York City, for example, found that 6.37 percent were infected with HIV (New York State Depart- ment of Health, 19891. Rates were higher among older youth and young adults; 8.83 percent of the 19- to 23-year-olds surveyed tested positive for HIV. The evidence regarding runaway and"problem" youths, however, is not unifonn. Thus, a seroprevalence study of 1,878 16- to 18-year-olds admitted to two Los Angeles County juvenile detention facilities between February and August 1987 found that the prevalence of HIV infection (0.16 percent) was much lower than that found among runaway youth 26Although rates for black and Hispanic Job Corps applicants varied geographically over a relatively wide range (4.9 to 10.2 and 0.5 to 7.7 per thousand, respectively), rates for whites evidenced less regional variability (range 1.2 to 1.9 per thousand).

166 ~ AIDS: THE SECOND DECADE in New York City despite high rates of reported {V drug use (10 per- cent), needIe-shanng among injectors (approximately 48 percent of mate injectors and 55 percent of female injectors), and intercourse with an IV drug user (16 percent of males and 10 percent of females reported this behavior) (Baker et al., 19891.27 The relatively low prevalence of HIV infection in this population may be related to the lower seroprevaTence rates in the L`os Angeles population in general (compared with New York). Nonetheless, these data indicate that, even among groups that might be expected to have high rates of infection (i.e., 16- to 18-year-old runaway and "problem" youths), there is ample reason for hoping that HIV infection can still be prevented before it makes more substantial inroads into these vulnerable populations. Conclusion The available data indicate that HTV infection has established itself in the teenage population, and some segments of this population report substantial infection rates. For example, the committee notes with great concern that almost 1 percent of black teenage mothers who delivered babies in New York City in 1988 were infected with HIV. Even though the available evidence is sufficient to conclude that HIV is present in the adolescent population, there are relatively few data available for accurately monitoring the spread of HIV in this group. One of the major difficulties in presenting a clear picture of the scope of the AIDS and HIV problem among adolescents is the paucity of available data for narrowly defined age groups. Often, researchers are left to patch together data from different sources that use inconsistent age groupings. This problem is exacerbated by the fact that the AIDS case data file available for public use defines two overly broad categories spanning adolescence (ages 13 to 19) and young adulthood (ages 20 to 29~.28 In trying to compose a picture of the epidemic among adolescents, the committee has found this grouping less than optimal. Similarly, sero- prevaTence studies that have included adolescents provide only enough information to determine that the virus is already seeded in this popula- tion and that it appears to affect minority youth from large urban areas 27Bakerand colleagues (1989) report that anonymous HIV antibody testing was done on 1,878 consec- utive admissions to the juvenile facility. Self-reported data on behaviors were obtained from a random subsample of 417 of these 1,878 juveniles: 11 individuals in this subsample refused to participate in the survey. 28Tabulations in this chapter requiring special breakdowns were provided by the CDC at the special request of the committee. The data file that is publicly available does not provide data by finer age categories.

AIDS AND ADOLESCENTS ~ 167 differentially. To understand the dimensions of AIDS and HIV infection among adolescents and to plan and target intervention programs, more detailed surveillance data are needed (see the previous recommendation). BEHAVIORS THAT PUT ADOLESCENTS AT RISK Not all teens are at risk for HIV infection. Some, by virtue of their low level of risky behavior or because of the absence of the virus among their potential partners, will remain uninfected. The vast majority of very young teenagers, as well as older adolescents who have not begun sexual intercourse and do not inject drugs, have little to worry about.29 But as these individuals get older, move to different geographic locations, or engage in new behaviors, their risk level will change. Sexually active teens and those who inject drugs are certainly more vulnerable to infection than adolescents who do not engage in these behaviors, but there may be considerable fluidity in adolescent risk taking. This variation in risk taking will affect who is at risk and how many are vulnerable at a . . . . particular point In time. The adolescent population contains pockets of teenagers whose be- havior puts them at relatively high risk of acquiring HIV infection. Sexually active teens and those who share injection equipment are espe- cially vulnerable if the virus is present within the population from which Heir sexual and drug use partners are selected. Some teens are already infected and thus may be capable of transmitting the AIDS virus to other adolescents. Furthermore, some teens have sex or share drug injection equipment with adults and thereby run an even greater risk of acquiring HIV infection. Even among the youngest of teens, a subset is engaging in un- protected intercourse and drug use. (Indeed, unprotected intercourse is more common among younger teens than among older ones.) The con- sequences of these behaviors appear to be more serious for very young teenagers than for older adolescents. As the data presented later in this section indicate, the earlier an individual initiates one type of risk behav- ior, the more likely it is that he or she will initiate others. And when intercourse begins at an early age, it is less likely to involve the use of contraceptives and more likely to result in sexually transmitted diseases 29A small number of teenagers who received blood transfusions or blood products prior to the im- plementation of mandatory screening in 1985 were infected by contaminated blood. Since the imple- mentation of screening programs, however, the number of persons (of all ages) who become infected through the contaminated blood transmission mode has declined dramatically (see Chapter 5).

168 ~ AIDS: THE SECOND DECADE than if it were begun later (Zelnik, Kantner, and Ford, 1981; Zeinik and Shah, 1983; Brooks-Gunn and Furstenberg, 1989; Hein, l989a).30 Describing the distribution of risk-associated behaviors among ado- lescents is not a simple task. No single statistic captures the complex dimensions of risk, and there is considerable variation in the prevalence of sexual and drug use behaviors across ages, genders, and racial sub- groups. Furthermore, the difficulty of the task is increased by the uneven quality of the available data. In the following sections the committee reviews the evidence available on sexual and drug use behaviors among adolescents. Although the review permits some relatively firm conclu- sions, it also highlights the need for more and better data to understand both the behaviors themselves and the individual and social factors that motivate and shape those behaviors. Sexual Behavior Data Sources Previous research on unintended adolescent pregnancy has provided valu- able data on adolescent sexual behavior. Three national surveys using probability samples of adolescents offer estimates of several important aspects of adolescent sexual behavior: (1) the National Longitudinal Sur- vey of the Labor Market Experience of Youth,3~ (2) the National Survey of Young Men and Women,32 and (3) the National Survey of Family Growth (NSFG).33 In addition to these efforts, the 1988 National Sur- vey of Adolescent Males (NSAM) provides information on the sexual behavior of teenage boys.34 30Leaving aside developmental immaturity, which affects planning and decision making, among young teens, some researchers suggest that adolescent girls may be more susceptible to gonococcal and chlamydial infections because of the anatomy and physiology of the adolescent cervix (Bell and Hein, 1984; Duncan et al., 1990). 31 The National Longitudinal Survey of the Labor Market Experience of Youth is an ongoing national probability sample of some 12,000 young people who were between 14 and 21 years of age when first interviewed in 1979. 32 The National Surveys of Young Men and Women (frequently referred to as the Kantner and Zelnik surveys) were conducted in three waves in 1971, 1976, and 1979, using independent samples. The 1979 survey used a national probability sample of approximately 1,700 females between the ages of 15 and 19, and approximately 900 males between the ages of 17 and 21, all of whom resided in metropolitan areas. Only the 1979 survey sampled young men. 33The National Survey of Family Growth is a periodic survey of probability samples of women be- tween the ages of 15 and 44 years. The survey was first conducted in 1973 and later in 1976. In 1982, the sample of 7.969 women was designed to represent all women in this age group and not just those who had been mamed at least once or those with children, as had been the case in earlier cycles. 34The 1988 National Survey of Adolescent Males used a probability sample of 1,880 boys between the ages of 15 and 19 years from the noninstitutionalized, never-married U.S. male population.

AIDS AND ADOLESCENTS ~ 169 Differences in design, substantive focus, population that was sam- pled, and time at which data were collected preclude making precise comparisons of the estimates denved from each survey.35 Nevertheless, reports of sexual behaviors do not stand in isolation. As discussed in a later section, the rates for adolescent pregnancies and STDs confimn the general sense provided by the above surveys of behavior that a substantial 35The National Longitudinal Survey of Youth is a panel study with annual follow-up comprising a nationally representative probability sample of 5,700 young women and 5,700 young men who were between the ages of 14 and 21 as of January 1, 1979. Individuals were selected from stratified area probability samples of dwelling units. Blacks, Hispanics, and disadvantaged whites are oversarnpled. Sampling criteria excluded young people who did not live within the 50 states and those who were institutionalized. Earlier waves focused primarily on labor market experiences, but information per- taining to education, marriage and fertility events, income and assets, family background, attitudes, and aspirations was also collected. Later waves have also collected data on drug and alcohol use as well as on family planning, child care, and maternal care. This survey attempts to retain the same interviewers from wave to wave; 78 percent of the 368 individuals who administered interviews in the 1984 survey reported participating in earlier waves (Campbell, 1984). Although the selection bias that has resulted from differential sample attrition does not appear to account for the lower level of sexual activity reported in this survey versus the level reported in others, the use of the same interviewers from wave to wave may affect a respondent's likelihood of being candid about reporting sensitive behaviors (Kahn, Kalsbek, and Hofferth, 1988). Reporting of drug use behavior in the 1984 wave is lowest among those who had the same interviewer in all prior waves (Mensch and Kandel, 1988b). The design of the National Survey of Young Men and Women has varied over the different waves conducted to date. This survey includes information on contraceptive use, pregnancies, pregnancy in- tention, and sex education experiences. For example, the 1971 wave interviewed 15- tO 19-year-old women living in households in the continental United States; women who lived in college dormito- ries were sampled separately. In 1979, both young women and men living in households in standard metropolitan statistical areas in the continental United States were eligible, but female participants were between the ages of 15 and 19 and males were between the ages of 13 and 17 years. Each wave includes different respondents. Because a larger fraction of female respondents in the 1979 wave had not completed their teen years, there may have been underestimation of the proportion of sexually active girls at each age. The National Survey of Family Growth (NSFG) collects data on fertility patterns, infertility, reproductive health, contraception, fertility intentions, childbearing, adoption, adolescent pregnancy, and unwed motherhood, pre- and postnatal care, and infant health along with information on social, economic, and family charactenstics. Surveys of this type began in 1955, and data were collected in 1960, 1965, 1970, 1973, 197S, 1982, and 1988; the last four years are cycles of the NSFG. This survey has used nationally representative samples of women between the ages of 15 and 44 years, although separate questionnaires have been designed for women under 25 years of age. Cycle III, which began in 1982, was the first to include all women regardless of their marital status. The NSFG (in contrast to other surveys that have included missing data codes) imputes values on variables for which the respondent did not provide an answer typically using so-called hot-deck imputation. This method is a commonly used nonresponse imputation procedure in which a sample is divided into two categories: those who respond to an item and those who fail to respond. The assumption is made that individuals who respond to several items in the same fashion (e.g., age and education) would also have similar characteristics as an item to which some subjects failed tO respond (e.g., fertility intentions). Data files for respondents who provided information on each item of interest are aggregated (making a deck) and then randomized (shuffling the deck) and stacked. When the investigator encounters a data file with a missing item, he or she turns to the group or deck of complete data files and selects the next one that matches on a preselected set of items. The response for the missing item is then taken from the response provided on its matched counterpart. This process permits the calculation of both means and variances for both groups.

170 ~ AIDS: THE SECOND DECADE portion of the youth of this country are sexually active and that many of these youths do not engage in protective behaviors that prevent HIV transmission (as well as unwanted pregnancies or STDs).36 Vaginal Intercourse Tables 3-7a and 3-7b present findings from the two most recent nations surveys that provide estimates of the age of initiation of intercourse for teens. These tables indicate that males are more likely than females to report intercourse at any age;37 by age 19, however, the majority of teenagers whether black or white, mate or female—report intercourse. Among 19-year-old males, 96 percent of blacks, 85 percent of whites, and 82 percent of Hispanics report having engaged in sexual intercourse. The proportions of females reporting intercourse are lower but still sub- stantial. Among females (born between 1964 and 1972),38 81 percent of blacks and 62 percent of whites reported that they had engaged in sex- ual intercourse before their 19th birthday. In addition, the proportion of teenagers reporting intercourse at an early age is not small. Roughly one- third of 15-year-old boys report having engaged in sexual intercourse, and 21 percent of teenage girls report sexual intercourse by their 16th birthday. In its last report, the committee noted that the proportion of per- sons of all ages who reported engaging in premantal intercourse had increased steadily since the early 19OOs (Turner, Miller, Moses, 1989:91- 98~. Newly available data from the 1988 NSFG evidence a continuing trend over recent decades for more females to begin intercourse dur- ing their teens (see Figure 3-5~. Thus, although roughly 30 percent of women born dunng the period 1944-1949 reported intercourse before age IS, more than half of the women born between 1965 and 1971 reported beginning intercourse before this age. The racial differences found in the proportions of sexually active 36For an overview of the sexual behavior of adolescents and unintended pregnancy, see C. D. Hayes, ea., Risking the Future: Adolescent Sexuality, Pregnancy, and Childbearing. Washington, D.C.: Na- tional Academy Press, 1987. 37 There is some speculation that boys are more likely than girls to exaggerate their sexual activity. Although responses from the surveys appear to be internally consistent, it is not possible to determine their validity. That is, it is not possible to say whether the responses are consistently biased toward overreporting of sexual behavior among males (see Zelnik and Kantner, 1980; Sonenstein, Pleck, and Ku, 1989a,b). See Chapter 6 for a review of the validity and reliability of survey measurements of sexual behaviors. 38Age definitions used for males and females differ slightly; see notes lo Tables 3-7a and 3-7b.

AIDS AND ADOLESCENTS ~ 171 TABLE 3-7a Percentage of Teenage Males Who Report Engaging in Premarital Sexual Intercourse, by Age and by Race/Ethnicity Mates' bOm 1968-1973 All White B1aCk HiSPaniC IS 32.6% 25.6% 68.6% 32.8% 16 49.9 46.7 70.1 47.2 17 65.6 59.1 89.6 87.6 18 71.6 71.4 82.5 52.8 19 85.7 84.5 95.9 82.2 N 1,880 752 676 385 NOTE: The National Survey of Adolescent Males (NSAM) was conducted during 1988 using a cohort of males born 1968-1973 who were 15 to 19 years of age at the time of the interview. a estimates are the percentage of males at a given age (e.g., 16-year-olds) who have engaged in premarital intercourse. Thus, the estimates for 16-year-olds, for example, reflect the experiences of boys where ages ranged from 16.00 to 16.99. SOURCE: Sonenstein, Pleck, and Ku (1989b:Table 1). TABLE 3-7b Percentage of Never Married Teenage Females Who Report Engaging in Sexual Intercourse, by Age and by Race/Ethnicity Females, born 1964 1972 Age Alla White Black Hispanic Before 14 4.1% 2.8% 8.0% b Before 15 10.3 8.2 17.4 b Before 16 20.9 18.9 31.2 b Before 17 35.2 33.0 50.4 b Before 18 49.7 48.8 65.8 b Before 19 62.7 62.0 80.7 b Before 20 68.2 66.5 87.7 b NC 727 326 317 gab NOTE: Estimates are the percentage of females who reported first intercourse before a given age among all females born 1964~1972 who had reached that age at the time of the interview. Thus, in calculating the percent of women who had their first intercourse before age 20, women whose age at the time of interview was younger than 20 were excluded from the calculation. aIncludes persons whose race was listed as other. b percentages not shown for Hispanic females because of small sample size. CNs shown are minimum unweighted sample counts. Ns for calculation of estimates of percent having intercourse at early ages are somewhat higher. SOURCE: Tabulated from 1988 National Survey of Family Growth Public Use Data tape.

172 ~ AIDS: THE SECOND DECADE 60 40 cL Before Age ~ 30 20 10 - . `~ `~; `~ ~~ Aid' As' Before Age 12_ I 1 1 1 1 1 1 1 `~ `9 `9CO ~,~ , ,~oO BIRTH COHORT FIGURE 3-5 Percentage of women reporting first sexual intercourse before ages 16 and 18 by birds cohort. NOTE: Percentages reporting intercourse before age 18 in 1968-1971 birth cohort calculated using only women aged 18 and over at the time of Me interview (unweighted n = 587). SOURCE: Tabulated from 1988 NSFG Public Use Data Tape using weighted data. teens and age at first intercourse appear real but remain largely unex- plained.39 Multivariate analysis of the predicted probability of having premarital sexual intercourse at specific ages indicates that even control- l~ng for a number of possibly confounding factors (e.g., age at menarche, mother's education, religious affiliation, family stability during adoles- cence) does not eliminate racial differences (Hofferth, Kahn, and Bald- win, 19871. One study of approximately 1,400 students from four urban junior high schools in Florida found racial differences in the progression of intimate relationships that would be consistent with the racial differ- ences seen in the age at first coitus (Billy and Ubry, 1985; Smith and Udry, 19851. The sequencing of heterosexual behaviors was more grad- ual and more clearly defined for white teens than for blacks who more often proceeded to intercourse without first engaging in noncoital sexual activities.40 Although the results from this study carrot be generalized to the national population, they have important implications for AIDS 39There are a number of theories in the demographic literature that address racial differences in adoles- cent intercourse. Some explanations rely on biosocial factors (see, for example, Udry and Billy 1987 discussion of pubertal development). Others look to sociological, economic, and sociocultural factors (Furstenberg et al., 1987). Empirical analyses have not led to any consensus on the most appropriate theoretical model or the most significant factors with which to explain such differences. 40For example, genital fondling, oral-genital contact, and so forth.

AIDS AND ADOLESCENTS | 173 prevention strategies. The researchers suggest that those who counsel teens need "to be sensitive to cultural and ethnic differences in sexual patterns. The lack of black adolescent involvement in precoital petting behaviors places increased importance on reaching these teens very early in their heterosexual relationships" (Smith and Udry, 1985:12031. Abstinent Teenagers In reviewing the tables that summarize the numbers of teens who report engaging in sexual intercourse, it is important to remember that there are teens who are reporting no coitus. Issues that claim the attention of researchers tend to cluster around problems. Thus, psychologists and sociologists who have studied adolescent sexual development have gen- erally looked at the behaviors that have resulted in unintended pregnancy and STDs. This emphasis has resulted in a more extensive data base on teens who have reported intercourse than on those who have not. It is important to note, however, that roughly one-th~rd of females and one- fifth of males remain virgins as they enter their twenties. These teenagers are more likely than nonv~rgins to be white or Hispanics and to report higher levels of religiosity; they are also more likely to score higher on intelligence tests, report higher expectations for academic achievement, and live in an intact family (Hayes, 1987; Rosenbaum and Kandel, in press; see Table 3-~. Data from the two youngest birth cohorts of the National Longitu- dinal Survey of Youth show that, among males born between 1963 and 1964, white teenage boys are most likely and black teenage boys are least likely to report no coital experiences. Among females, Hispanic teenage girls are most likely and blacks least likely to report no coital experience before age 19. Parental education is positively correlated and delinquency in the teenagers themselves is negatively correlated with self-reported abstinence. Yet although the demographic characteristics of teens who report no intercourse can be described, much less can be said with any certainty about the factors that encourage abstinence. Programs that have attempted to delay the onset of intercourse among teens have in general not been evaluated. Patterns of Heterosexual Behavior The age of initiation of heterosexual intercourse among adolescents indi- cates the beginning of the period of risk for HIV infection. Information on patterns of sexual intercourse, including frequency, number of sexual 4lRacial differences in virginity persist even in analyses that control for the effect of socioeconomic factors.

174 ~ AIDS: THE SECOND DECADE TABLE 3-8 Estimated Percentage of Males and Females (born 1963-1964) Who Had Not Engaged in Sexual Intercourse Before Their 19th Birthday by Selected Social and Demographic Charactenstics ~ Males Females Characteristic Percentage N Percentage N Mother's education < High school diploma 16% 519 25% 531 High school diploma 20 638 35 589 Some college 27 296 47 259 Fathers education < High school diploma 15 604 26 557 High school diploma 19 468 31 462 Some college 29 101 47 361 Race/ethnicity White 23 1,145 34 1,082 Black 5 207 19 203 Hispanic 17 381 50 95 Household structure at age 14 Intact 23 1,024 39 967 Nonintact 12 428 20 412 1980 Delinquent acts None 40 231 45 484 1 or 2 24 428 29 531 3 or more 12 745 27 325 Total Sample 20 1,453 33 1,380 NOTE: Data are from the 1963 and 1964 birth cohorts of the National Longitudinal Survey of Youth. Virginity is measured by self^reports in the 1984 wave of the survey when the sample was aged 19 to 21. Delinquent acts were measured when the sample was 16 to 17 years old. SOURCE: Denved from Rosenbaum and Kandel (in press:Table 1). partners, and the likelihood of using condoms, provides an indication of the degree of risk. Available data on these variables are summa- rized below, as is the limited information collected to date on adolescent heterosexual anal intercourse. Frequency. Data from the 1982 National Survey of Family Growth indicate that a substantial subset of sexually active adolescents have intercourse frequently. Among 15- to 19-year-old unmarried sexually experienced females, 25 percent reported intercourse two to three times per month, 20.9 percent reported intercourse once a week, and 16.3

AIDS AND ADOLESCENTS ~ 175 percent reported it more than twice a week (see K.A. Moore et al., 1987:Table 1.7~. Moreover, once women initiate sexual intercourse, they usually continue sexual activity. In surveys of approximately 1,800 never- married women conducted in 1976, Zelnik and Kantner (1977) found that only 14.3 percent of sexually experienced white females and 12.7 percent of sexually experienced black females between the ages of 15 and 19 reported a single episode of intercourse. More than 10 percent (11.7 percent of 15- to 17-year-olds and 25 percent of 18- to 19-year-olds) reported engaging in intercourse at least twice a week (K. A. Moore et al., 1987~. In addition, in the 1982 National Survey of Family Growth, 40.5 percent of sexually experienced, single, 15- to 19-year-old females who had had intercourse during the three months prior to the interview indicated that they had done so more than four times a month (K. A. Moore et al., 1987~. Number of Partners. Although the data on numbers of sexual part- ners of teenagers are limited, they indicate that a substantial fraction of teens have multiple sexual partners. Figure 3-6a displays the distri- bution of the number of sexual partners reported by young women in the 1988 National Survey of Family Growth and by young men in the 1988 National Survey of Adolescent Males. These figures show that, although roughly one-quarter of the 18- to 19-year-olds were sexually in- experienced (i.e., they reported having no partners), a substantial fraction reported having several sexual partners. For example, among 18- and 19-year-old women, 11 percent reported 6 or more sexual partners, and 5 percent reported 10 or more partners. For young women in their early 20s, the reported numbers are considerably higher; 20 percent of 20- to 24-year olds report 6 or more partners and 10 percent report 10 or more partners. In its previous report (Turner, Miller, and Moses, 1989:98-100), the committee contrasted parallel findings for teenage women born during the 1960s with data collected in 1970 on the number of premarital partners reported by adult women born during earlier periods of this century. The increasing number of premarital sexual partners reported by women in more recent birth cohorts suggest a major shift in the social nones governing nonma~ital heterosexual sexual behavior among young women. Unmarried young men in these surveys declared many more sexual partners than unmanied young women of the same age. As Figure 3-6b shows, 26 percent of 18- and 19-year-olds reported 6 or more partners

176 ~ AIDS: THE SECOND DECADE A. Females in 1988 30 ,.... . : ::: . `~20 _ ::: Q ·:.: :' Q 10 _ ::..:. ,:::,, ::.... :2 : : O I::::: , 30 LLJ 20 6 of LL C: CC LU Q 10 o .::. '..: ,... · .... ::::::: ::~ '~.:s .:,::., ........ ,............. '.:.." i:, .~,... _ :,:::: , . . :::: :,: .. : :,: :. ·...: :...... : ::. , . . . ~ ::..e ,.,:: . ::,::.: :.. . .. :::::::. :::: . :::::. ,'.:::2 : :,: ~ ......... : : ..::. .... . . . . . :::: .... .::: ... ::. :,:..:.. A9eS 18-19 ~ Ages 21-24 0 1 2 B. Males in 1988 _ ~ ~ . NUMBER OF SEX PARTNERS :;.::;; 0 1 2 3-5 6-9 10+ ~ A9eS 18-19 ~3 ::::: ~ a · 3 ·:::: ~ :: . ..1 ·:.:.:. .. STASH 1 1 1 I::::::' 1 .:.:.:., , ·: : :: : ~ 1 : . .::: 1 1 ·: :::: ~ 1 :'..:::] 1 ·.::~::'] 1 ::-::"] 1 .:::~-::d 1 ::::::] 1 a::::. 1 ::::::] 1 I:...] 1 :::: :: ~ 1 .~..~..:.d 1 : .-. -:: ~ 1 ! . ~ I ,~ —_:, , C u NUMBER OF SEX PARTNERS I..:.:,:: 11 + FIGURE 3-6 Number of lifetime sexual partners reported by: (a) females aged 18-19 and 2~24 in 1988, and (b) males aged 18-19 in 1988. SOURCES: 1988 National Survey of Adolescent Males, unpublished tabulation provided by F. Sonenstein (Urban Institute), J. Fleck, and L. Ku; and 1988 National Survey of Family Growth, unpublished tabulation provided by A. Campbell (National Institute of Child Health and Human Development).

AIDS AND ADOLESCENTS ~ 177 and 7 percent reported 11 or more. Nonetheless, more than half of the males of this age reported two or fewer sexual partners.42 Contraception and Condom Use. Condoms have been shown to provide protection against AIDS and other sexually transmitted diseases (Wigersma and Oud, 1987; Feldblum and Fortney, 1988; Darrow, 1989; Periman et al., 19901.43 However, the fact that a significant number of adolescents report a history of STDs indicates that many sexually active teens are not using condoms or are not using them properly (Bell and Hein, 1984~. The prevalence of STDs among adolescents is especially worrisome given that rates of HIV transmission are thought to increase when genital lesions are present (Nahmias et al., 1989; Teizak et al., 1989; Zewdie et al., 19891. Although contraceptive use among teens has increased, at least through the 1970s (K. A. Moore et al., 1987:Table 2.1, 390), the frac- tion of U.S. teens who use birth control at first intercourse and regularly thereafter is still small in comparison with teens from other developed countries (Iones et al., 1985~.44 Among 15- to 24-year-old women who were interviewed in the 1982 National Survey of Family Growth and who reported premarital intercourse during their teens, approximately one-half (47 percent) of white females and two-thirds (66 percent) of black females did not use any contraception at first intercourse (Kahn, Rindfuss, and Guilkey, in press). Moreover, young teens are less likely to use contraception than older adolescents and more likely to delay the use of birth control until after intercourse has been initiated. This tendency has resulted in probabilities of unintended pregnancy within six months of onset of intercourse that are nearly twice as high for teenage girls who are younger than 16 years of age than for those who wait until age 18 or 19 to initiate intercourse (Zabin, Kantner, and Zelnick, 1979~. Among women reporting adolescent intercourse in the 1982 National Survey of Family Growth, less than one-third of those who reported intercourse 42It should be noted that different rates of virginity reported in Figure 3-6 and Table 3-8 reflect data that are derived from separate surveys of different age groups. 43 In vitro studies of condoms (i.e., conducted under laboratory conditions) show that HIV does not pass through latex condoms. In viva studies of the effectiveness of condoms (i.e., data collected from actual use) are confounded by problems associated with improper and inconsistent use (see Feldblum and Fortney, 1988). 44 For example, of the 15- tO 20-year-old females interviewed in the United States in 1979, 70 percent reported using contraception at their last coitus. More than 80 percent of 16- to 17-year-old Swedish girls surveyed in 1978 reported contraception at last coitus, as did 89 percent of 16- to 18-year-old females from the Netherlands interviewed in 1981 (Jones et al., 1985).

178 ~ AIDS: THE SECOND DECADE prior to the age of 15 used some method of birth control at first premar- ital intercourse compared with half of those who initiated intercourse at age 17 or 18 (Mosher and Bachrach, 19871.45 Condoms were the method of contraception at first intercourse re- ported most often by 15- to 24-year-old sexually active females inter- viewed in the 1982 National Survey of Family Growth (Kahn, Rindfuss and GuiLkey, in press). Yet despite the popularity of condoms, only a minority of respondents (21.7 percent of white teens and 13.7 percent of nonwhite teens) reported condom use at first intercourse. After first inter- course, when it appears that sex is more likely to be planned, teenagers are considerably less likely to use condoms and more likely to use birth control pilIs.46 Only 12.2 percent of white sexually active 15- to 19-year- old female adolescents reported current use of condoms during the 1982 wave of the National Survey of Family Growth (Pratt et al., 1984~. Changes in teenage contraceptive practices during the 1980s have been reported in the last year for teenage males, and preliminary analyses of the 1988 NSFG suggest that parallel changes have occurred among teenage females. Sonenstein and colleagues (1989b), for example, found higher rates of reported condom use in the 1988 NSAM than had ever been found before. More than half of 15- to 19-year-old males (58 percent) claimed that they had used a condom during their last episode of intercourse. This rate of reported condom use is more than double (58 versus 21 percent) the rate reported in the 1979 National Survey of Young Men and Women (Sonenstein, Pleck, and Ku, l989b).47 Furthermore, the proportion of teenage males reporting condom use in 1988 did not appear to vary greatly with age (59.5 percent for 15-year-olds versus 55.2 percent for 19-year-olds). Finally, contrary to previous patterns of reported condom use, blacks were more likely (65.5 percent) than whites (54.4 percent) to report condom use. It is not possible to be certain that boys are not exaggerating their use of condoms. Given the onset of the AIDS epidemic, the socially 45A similar pattern of findings is reported from the 1982 National Surrey of Family Growth (see K. A. Moore et al., 1987, Table 2.6:400). Slightly less than half (42 percent) of the teens who participated in this survey and who reported intercourse by age 15 waited more than 12 months after first coitus to use birth control. Only 15 percent of those who started intercourse by ages 18 to 19 waited that long to take preventive action. 46In the 1982 National Sun/ey of Family Growth, almost two-thirds (63.9 percent) of 15- to 19-year- old females reported the pill as their current method of birth control (Pratt et al., 1984). 47 Although there are no available data specifying the age of purchasers, condom sales reportedly increased by more than 60 percent between 1987 and 1989 (Consumer Reports, 1989). Between 1985 and 1988, retail sales of condoms increased from approximately $180 million to $360 million (R. Rothenberg, "Condom makers change approach," New York Times, August 8? 1988, D- 1).

AIDS AND ADOLESCENTS ~ 179 responsible answer to this survey question in 1988 is clear. (See Chapter 6 for further discussion of the validity of such survey measurements.) Observed declines over this same time period in the use of female methods (e.g., intrauterine devices or lUDs, the pill, diaphragms, foam, jelly, or suppositories)48 lend some, albeit weak, support to the inference that condom use may actually have increased substantially among sexually active adolescents. Preliminary analyses of the 1988 wave of the National Survey of Family Growth, however, indicate a parallel trend for a greater proportion of sexually active teenage girls to report using condoms. Among sexually experienced 15- to 19-year-old females, 19.7 percent reported currently using condoms, compared with ~ I.4 percent in 1982.49 Other analyses5° of the methods used at first and last intercourse evidence a similar rise in the rates of reported condom use. Of sexually experienced females aged 15 to 19 in 198S, 47 percent reported using condoms at their first intercourse and 29 percent reported using condoms at their last intercourse.si These recent indications of increased condom use raise hopeful prospects for protecting this nations youth from HIV infection and may be evidence of the effectiveness of programs to promote greater condom use among sexually active teens. Nevertheless, the need for caution and concern remains. Half of all teenage boys in the 1988 NSAM and one- third of teenage females in the 1988 NSFG did not use any contraception at first intercourse. Although the rates of condom use appear to have increased substantially during the 1980s, a substantial proportion of sex- ually active teenagers are not using condoms. Indeed, only half of the teenage males at highest risk for infections in the 1988 NSAM sample had used a condom at last intercourse (see Table 3-91. Heterosexual Anal Intercourse. Although anal intercourse appears to be more likely than vaginal intercourse to transmit HIV (IOM/NAS, 48 In 1979, 28.0 percent of 17- to 19-year-olds reported the use of a female method of contraception compared with 21 7 percent in 1988 (Sonenstein, Pleck, and Ku, 1989b). 49W.F. Pratt, National Center for Health Statistics, personal communication, March 7, 1990. OC. F. Turner, National Research Council, unpublished tabulations, May 2, 1990. 51 As in earlier studies, the pill was the most frequently reported contraceptive used at last intercourse; 38 percent of sexually experienced 15- to 19- year aids reported using the pill at last intercourse versus 8 percent at first intercourse. 52Those teens considered to have the greatest chance of becoming infected reported either same-gender sex, the use of IV drugs or a sexual partner who injected drugs, a history of other STDs' or sex with a prostitute. This high-risk group constituted 8.7 percent of the sample (Sonenstein, Pleck, and Ku, 1989a).

180 ~ AIDS: THE SECOND DECADE TABLE 3-9 Risk Behaviors and Condom Use at Last Intercourse Among Sexually Active, Never-Marned Males Aged 15 to 19 Years in the United States, 1988 Risk Behaviors Percent of Active Males in Categoryb Percent Used Condoms Last Time Group 1: Any of the following Ever had male-male sexC Ever used IV drugs or was sex parmer of an IVDU 1.8 Ever had a sexually transmitted diseased Ever had sex with a prostitute 8.7 3.0 Group 2: Any of the following unless in Group 1 Ever had sex wide a stranger Ever had sex with someone with many parmers Five or more sex partners in last year Group 3: None of the above behaviors 2.9 1.2 37.7 20.0 31.1 7.8 53.6 51.2 65.7 20.8 66.4 16.8 44.5 39.9 40.6 36.g 65.9 Pin the three composite AIDS risk categories, 5.7 percent of categorical data were missing. b among Rose with valid responses. C Male-male sexual activity includes mutual masturbation, or insertive or receptive urogenital or anogenital intercourse. Asexually transmitted diseases include gonorrhea, syphilis, herpes, or venereal warts. SOURCE: Sonenstein, Pleck, and Ku (1989a). 1988:45), little is known about the number of teenage girls who have experienced anal intercourse. Small surveys of clinic populations indicate that some girls are engaging in this risky behavior and that condoms are rarely used. Jaffe and colleagues (1988) found that anal intercourse was reported by approximately one-quarter (25.2 percent) of the teenage girls attending an adolescent outpatient clinic in New York City. In addition, Kegeles and coworkers (1989) found that, of 104 girls aged 15 to 21 interviewed in a family planning clinic, 12 percent reported that they had engaged in anal intercourse. Given the potential risk for HIV infection associated with this practice, the committee believes that surveys of adolescent sexual activity should gather data on this behavior. Same-Gender Sexual Behavior Among Adolescents For some young people, sexual experimentation will include same-gender sexual behavior. The precise number of adolescents who experiment with same-gender sex is not known with any certainty. Using data from a 1970 survey of sexual behavior in a national probability sample, Pay and colleagues (1989) concluded that a minimum of 20 percent of American men have had sexual contact to orgasm with another male at some point in their lives, and a minimum of 6 percent have had such contact at least once during adulthood. The overwhelming majority of

AIDS AND ADOLESCENTS ~ 181 men who reported such contacts reported that their first male-male contact to orgasm occurred during their teenage years or earlier.53 In the 1988 National Survey of Adolescent Males, Sonenstein, Pleck, and Ku (1 989a) note that 3 percent of respondents (aged 15 to 19) reported same-gender sexual activity (mutual masturbation, oral or anal sex). There have been other surveys of adolescent homosexual behavior, but these efforts have not relied on probability sampling techniques and thus provide estimates that cannot be generalized to the broader adolescent population. The consensus among such studies nonetheless is that adolescent homosexual activity is not rare. An anonymous, self- administered questionnaire given to 512 high school students in the Bronx, for example, found that 10 percent of female students and 9 percent of male students reported same-gender sex (Reuben et al., 1988~. Among 279 homosexual and bisexual men who were interviewed by CDC in 1981 and 1982 about their medical history and lifestyle, more than half reported the initiation of same-gender sex by age 16; 20 percent reported initiation of same-gender sex by the age of 10 (Haverkos, Bukoski, and Amsel, 19891. As noted by Pay and colleagues (1989), however, much of the same-gender sexual activity during adolescence does not appear to be the first stage in the development of an exclusively homosexual identity or lifestyle. Rather, the role of same-gender sexual contact in the sexual development of heterosexual or bisexual youth has yet to be defined. Sexually Transmitted Diseases Different state requirements for reporting STDs and differential reporting by public versus private health care delivery systems make it difficult to estimate national STD incidence and prevalence rates. Moreover, rates that present the number of cases per 100,000 10- to 19-year-olds will underestimate the size of the problem because not all teens are sexually active.54 There are data available, however, indicating that STDs are a public health problem for teenagers in this country. In 1987, there were 7,041 cases of gonorrhea reported to CDC among 10- to 14-year-olds and 18S,233 cases among 15- to 19-year-olds, giving unadjusted rates of 42.7 per 100,000 and 1,028.1 per 100,000 for the two respective age groups 53 Of those reporting age at first contact, approximately one-half reported male-male sexual contact to orgasm before age 15; approximately 90 percent reported that their first contact occurred before age 22 (C. F. Turner, National Research Council, personal communication based on unpublished tabulations from 1970 Kinsey/National Opinion Research Center [NORC3 survey, January 20, 1990). 54Age-specific rates of various diseases are calculated by dividing the number of cases within an age group by the number of susceptible individuals in the group. Adolescents who do not report intercourse are not susceptible to STDs; therefore, including all 10- to 19-year-olds in the denominator of such a calculation will underestimate the rate.

182 ~ AIDS: THE SECOND DECADE (CDC, 198Se). Compared with the gonorrhea caseload, there were fewer primary and secondary syphilis cases reported for 1987; 229 cases were identified among 10- to 14-year-olds (1.4 per 100,000~. Still the number of cases of syphilis increased and 4,331 cases were diagnosed among 15- to 19-year-olds, a dramatic increase that yielded a rate of 23.7 cases per 100,000. The incidence of both gonorrhea and syphilis was 1.4 to 3.l times higher among females in these age groups than among males. On addition, although STD rates for adolescent males are substantial, teenage boys do not account for as large a proportion of male gonorrhea cases as do teenage girls for cases among women. In 1979, 38 percent of gonorrhea cases reported by females were diagnosed in adolescent girls, but only 17 percent of cases among males were In teenage boys (Bell and Hein, 1984~. There is some evidence to suggest that STDs are a greater problem in younger populations than in older ones. When the denominator for national STD rates is corrected to include only the sexually active portion of the adolescent and young adult populations, 10- to 19-year-olds have the highest rates of gonorrhea and syphilis. In 1976 the rates for gonorrhea among sexually active females 10 to 19 years old were approximately 3,500 per 100,000 (Bell and Hein, 19841. Gonorrhea rates dropped precipitously as age increased; among 20- to 24-year-old women, the rate was approximately 1,800 per 100,000, and among 25- to 29-year-old women it was approximately 700 per 100,000. Rates that decline with age may be attributable in part to the dif- ferential use of health care facilities. Younger women may be more likely than older women to use public facilities (as opposed to private ones), and therefore may be more likely to have their case reported to the health department. Yet even within the population of clients who use public clinics, rates of STDs decrease as age increases. This trend is also reflected in lower incidence rates of hospitalized cases of pelvic inflammatory disease among older females (Bell and Hein, 19841. Conclusion The specter of AIDS has not stopped teens from engaging in unprotected sexual intercourse. Furthermore, continued risk taking does not appear to be related to a lack of awareness of the threat posed by unprotected sexual contacts.55 It is clear that as Tong as the sexual behaviors that transmit 55CDC-assisted surveys conducted in 1988 found that the majority of students between the ages of 13 and 18 were well aware that sexual intercourse is a risk factor for HIV transmission (CDC, 1988a). Between 88.3 and 98.1 percent of the sample, which included students from areas with the highest cumulative incidence of AIDS, correctly identified sexual intercourse as a risk factor, and between 83.8 and 98.4 percent knew that IV drug use could spread HIV infection.

AIDS AND ADOLESCENTS I 183 HIV infection are found in the adolescent population, there remains the possibility that teenagers can acquire or transmit the AIDS virus. It is therefore important to heed the empirical data, note the populations and geographic areas in which the risky behaviors tend to occur, and direct intervention resources toward those areas to prevent AIDS. Such actions will not only improve the chances for preventing the further spread of HIV infection but may also help to deal with other long-standing problems of the adolescent population, namely, STDs and unintended pregnancies. Drug Use As noted earlier, adolescence is a time of experimentation with a variety of behaviors that may include sex and the use of drugs. Much drug use, even though it is considered to be a significant health threat to the individual as well as to society at large, does not constitute a direct risk factor for AIDS. IV drug use,56 on the other hand, can play a direct role in HIV transmission if drug injection equipment is shared. Moreover, the use of drugs that are not injected may pose an indirect threat. Even alcohol, for example, can play an indirect role in transmission by Towenng inhibitions and perhaps clouding the judgment of those who are drinking, thereby facilitating unsafe sexual practices (Stall et al., 1986; Faltz and Madoover, 1987; Flavin and Frances, 19871. Another drug that has the potential to affect transmission of the virus is crack, a smokable and potent fob of cocaine that is believed to heighten sensations of sexual arousal among males. Ethnographic studies indicate that some women exchange sex for crack (or for cash to buy crack) (Fullilove et al., 1990; Turner, Miller, and Moses, l989:Chapter 31. In a study of 13- to 19-year- old black male and female crack users in the San Francisco area (N = 222), virtually all (96 percent) were found to be sexually active, slightly more than half (51 percent) reported that they had combined sex with drug use, and one-quarter of both boys (25 percent) and girls (24 percent) had exchanged sexual favors for drugs or money (Fullilove et al., 1989; 19901. The potential for unsafe sexual practices under such conditions is considerable; crack users are unlikely to know the serostatus, sexual history, or drug habits of their partners, and there is small likelihood that condoms will be used. Indeed, Fullilove aIld colleagues (1990) found that 41 percent of teenage crack users in their sample reported a history of an STD, and only 26 percent of boys and 18 percent of girls reported condom use dunng their last episode of sexual intercourse. Thus, the 56Illicit drugs that are administered through intravenous injection include heroin, cocaine, ampheta- mines, and so-called speedballs, a combination of heroin and cocaine.

184 ~ AIDS: THE SECOND DECADE combination of sex and drugs in the adolescent' population may be a potent mixture that bolsters risky behaviors and contributes to the spread of HIV. The next section examines the prevalence of drug use in the adolescent population. Data Sources and Limitations The preponderance of the data presented in this section comes from three national survey programs. Monitoring the Future is an annual survey of approximately 17,000 high school seniors. The National Household Survey on Drug Abuse collects data from approximately 8,000 individuals who are 12 years of age or older; slightly more than one-quarter (2,246 persons or 28 percent) of the sample are between the ages of 12 and 17. The 1987 National Adolescent Student Health Survey examined the health-related knowledge, attitudes, and behavior of 11,419 eighth and tenth grade students. National drug use surveys that include teenagers are likely to under- estimate the number of teens using drugs. School-based surveys, such as Monitoring the Future or the National Adolescent Student Health Sur- vey,58 do not capture adolescents who have dropped out of school or those who are chronic absentees. Thus, survey efforts may miss an important and sizable segment of the at-risk adolescent population. (Johnston and colleagues t1989] report that high school dropouts constituted 15 percent of the age group surveyed by the National Adolescent Student Health Survey, and Mensch and Handel ~l98Sa] found that approximately 22 percent of the 12,000 young people interviewed in the National Lon- gitudinal Survey of Youth had dropped out of school at least once.) In 1984, there were more than a half million 14- to 17-year-olds who had dropped out of high school prior to graduation (Hayes, 1987, Table 2.3:38~. Moreover, Hispanic adolescents were more likely than other youth to have dropped out of school by age 15, and 14.3 percent of Hispanic teens had dropped out of school by age 16 or 17. Household surveys, such as the National Household Survey on Drug Abuse,59 also have limitations in that they miss the transient, the homeless, and the incarcerated. At the time the 1980 census was taken, 0.8 percent 57 Monitoring the Future, an ongoing annual survey, began in 1975. Each year, 125 to 135 schools are selected to provide a representative cross-section of U.S. high school seniors. The survey also has a longitudinal component that follows approximately 2,400 students in each class. 58The 1987 National Adolescent Student Health Survey included students from randomly selected classrooms chosen from a national probability sample of 217 schools in 20 states. 59The National Household Survey on Drug Abuse is an ongoing survey of individuals aged 12 and older who live in households in the United States. Samples have ranged from approximately 3,000 in

AIDS AND ADOLESCENTS ~ IS5 of all 15- to 24-year-olds were confined to institutions. Males incarcerated in prisons, jails, or juvenile detention/corrections facilities constituted the majority (83 percent) of the 360,000 individuals in this group (Wetzel, 19871. All of the groups omitted from household or school surveys are known to have higher drug use rates than the general population (Shatter and Caton, 1984; Rothman and David, 1985; Mensch and Kandel, 1988a; Yates et al., 1988; Baker et al., 1989; Rolf et al., in press). Inevitably, questions also arise regarding the validity of self-reported data on illegal activities, including the use of illicit drugs, that are col- lected in surveys. Comparisons of drug-related data from the 1984 waves of the National Longitudinal Survey of Youth with data from Monitoring the Future and the National Household Survey on Drug Abuse suggest that there is underreporting of illicit drug use (other than marijuana) in the National Longitudinal Survey. Moreover, youth who report limited involvement with a particular drug are more likely to deny such use (Mensch and Kandel, l988b), a finding that is supported by another, more geographically circumscribed study (Single, Kandel, and Johnson, 1975~. Moreover, data from the Monitoring the Future survey showed proportionately more drug use reported in the month prior to the inter- view than in the preceding 12 months, reflecting, perhaps, a telescoping of recall (Bachman, Johnston, and O'Malley, 1981~. Given that there is some degree of underreporting of drug use and some ur~de~representa- tion of the heaviest users, these surveys may be providing lower-bound estimates of adolescent Mug use in the United States. Initiation of Drug Use Adolescent drug users do not suddenly issue forth as full-blown addicts. Rather, there is an empirically observed progression in stages of drug use, beginning with licit drugs, such as beer and wine; continuing with cigarettes or hard liquor, or both, with some teens going on to marijuana use; and fewer finishing the progression with other illicit drugs, such as stimulants, inhalants, hallucinogens, sedatives, and nonprescription tranquilizers. Cocaine use is frequently preceded by or coincides with the use of other illicit drugs (B. A. Hamburg, Kraemer, and Jahnke, 1975; Yamaguchi and Kandel, 1984a; O'Donnell, 1985; Newcomb and gentler, 1989; Kandel and Davies, in press). Indeed, data from the 1984 wave of the National Longitudinal Survey of the Labor Market Expenence of 1972 to more than 8,000 in 1985. Individuals are selected using a multistage area probability design (National Institute on Drug Abuse [NII)A3, 1988)

186 AIDS: THE SECOND DECADE Youth found that cocaine use is almost always preceded by marijuana use (Kandel and Davies, in press).60 Not all drug users participate in every step of the sequence outlined above, nor does the use of a drug at a lower level inevitably lead to a higher stage of drug use (Newcomb and BentIer, 19891. When progression to the next level of drug use does occur, however, the use of drugs from the previous stages generally continues; indeed, heavy or frequent use of a familiar drug is characteristic of the transition to the next step (Kandel, 1984; Rosenbaum and Kandel, in press). Injection Practices Among Adolescents The sharing of injection equipment to administer illicit drugs is one of the major modes of transmission of HIV in the United States and Europe. Approximately 30 percent of AIDS cases in the United States report the use of injected illicit drugs as a risk factor for infection (see Chapter 11. Most persons who inject illicit drugs, such as heroin, first do so either in late adolescence or in early adulthood (Gerstein, Judd, and Rovner, 1979; Newcomb and gentler, 19881. If a person has not initiated illicit drug injection by the age of 25, it is unlikely that he or she will do so. Consequently, any examination of the risk of acqumug HIV infection among adolescents and young adults should consider exposure to the virus through shared injection equipment. Progression through the various stages of drug use does not in- evitably lead to injection; as the drugs become increasingly "harder," fewer and fewer adolescents progress to the next stage of the drug use sequence, leaving only a small number who will move to the later stage of injection. This progression raises important issues concerning the economy of structuring intervention efforts that are meant to deal with the hazards associated with the end point of this sequence. There is no pharmacologic inevitability that leads adolescents to seek stronger or more powerful psychoactive drugs, although there appear to be psychoso- cial factors associated with progressive drug use that may be helpful in identifying individuals for early intervention. These factors include hav- ing peers who are heavily involved in drug use, beginning drug use at a relatively early age, and reporting a variety of social and developmental problems, including poor relationships with parents, peers, and authority systems (Kandell, Kessler, and Margulies, 19781. The precise number of adolescents and young adults who inject drugs 60Among individuals who participated in that survey and reported the use of alcohol or cigarettes, marijuana, or other illicit substances, less than 1 percent reported experimentation with cocaine before the initiation of marijuana use (Kandel and Davies, in press).

AIDS AND ADOLESCENTS ; 187 is not known. However, data from the Treatment Outcome Prospective Study (TOPS)61 have shown that, of individuals in treatment in this country, a not insignificant percentage are teenagers. Of the 8,795 TOPS participants who reported IV drug use in the year before entenog treat- ment, 0.2 percent or 17 individuals were less than 16 years old, I.6 percent or 140 were 16 to 18 years old, and 7.3 percent or 642 were 19 to 21 years old (Ginzburg, 19884. The impact of AIDS on the number of drug users who inject, the frequency of injection, and the use of shared injection equipment is also unknown. Of cause for concern, however, are reports such as the anecdotal information supplied by Mata and Jorquez (19881. These investigators reported that, where needle-shanog persists among Mexican Americans who inject drugs, young users are the most likely age group to participate in this hazardous practice. Prevalence of Drug Use The number of adolescents who use specific drugs, such as cocaine or heroin, is not known.62 Estimates of the percentage of adolescents who report drug use rely on surveys that may provide a sense of the lower bound on these numbers. As discussed earlier, the degree of underreporting and underrepresentation in these surveys is not known, making it difficult to interpret differences in the rates of drug use found across studies. What is clear is that prevalence rates are higher in school surveys than in household surveys (compare for example, Tables 3-10 and 3-111. It is thought that these differences in estimates may be attributable to the different methods used to collect the data (i.e., a face-to-face interview in the home versus a self-administered questionnaire).63 Although drug use among high school students has declined since 61 TOPS data were collected from three annual cohorts (1979, 1980, and 1981) of individuals admitted to more than 40 drug abuse treatment programs in this country. The survey seeks to understand the natural history of drug use before, during, and after treatment. More than 11,000 subjects who entered treatment between 1979 and 1981 were interviewed at the point of intake into treatment; a subset of 4,600 was followed after discharge from treatment through 1986. 62 See Turner, Miller, and Moses (1989:Chapter 3); and Spencer (1989) for a review of national esti- mates. 63 Differences in anonymity and privacy are likely to affect responses. In the National Household Survey of Drug Abuse noted above, drug questions were asked aloud, and responses were marked on the answer sheet by the interviewer. This procedure raises questions about protecting the privacy of the respondent and the validity of the response given the presence of parents or others in nearby proximity. In the school surveys, questionnaires were entirely self-administered. In addition, subjects' ages vary across studies. The National Adolescent Student Health Survey was conducted in the fall when eighth graders were, for the most part, 13 years old and tenth graders were 15 years old. The Monitoring the Future survey was conducted in the spring when seniors were 17 and 18. The National Household Survey of Drug Abuse covers the same aggregate population as the student surveys except for the inclusion of 12-year-olds and the exclusion of 18-year-olds.

188 ~ AIDS: THE SECOND DECADE its peak in 1979 (see Table 3-10), the proportion of adolescents who use drugs remains considerable (Johnston, Bachman, and O'Malley, 1989~. The proportions of teens who report the use of specific Mugs vary by drug type and route of administration. Considerably more adolescents report the use of licit than illicit substances, and drugs that are swallowed, smoked, or inhaled are more popular than drugs that are injected. Alcohol use, for example, is extremely common, with more than three-quarters of eighth graders (74 percent) having had at least one drink (see Table 3-11~. From a health perspective, however, quantity and frequency of consump- tion are more significant than whether an individual has ever used alcohol. The data indicate that there are significant numbers of young teenagers who report drinking levels that may pose potential Meats to their health. More than one-quarter (26 percent) of eighth graders and 38 percent of tenth graders reported having had five or more drinks on at least one occasion in the two weeks prior to the National Household Survey (NIDA, 19881. It is clear from these data that many adolescents not only Wink but do so frequently and excessively. Heavy drinking, moreover, is known to be combined with increased sexual activity and decreased use of contraception (Ensminger, 19871. AIDS-related studies of adults also indicate that alcohol use is associated with unsafe sexual behavior (Stall et al., 1986~; however, much less is known about adolescent risk taking under the influence of alcohol. According to adolescent surveys, marijuana is the most popular illicit drug for this population.64 In 1985, nearly one-quarter (23.6 percent) of 12- to 17-year-olds reported experimentation win marijuana, and the majority (19.7 percent) reported using the drug during the past 12 months (see Table 3-101. Among eighth graders surveyed in 1987, 14 percent had used marijuana; a third (34 percent) of tenth graders and half (50 percent) of a group of high school seniors had also smoked marijuana (see Table 3-111. The use of marijuana, hashish, and hallucinogens decreased between 1979 and 1985, but the use of cocaine has fluctuated. Rates now appear to be approximately stable (see Table 3-101. In 1985, 5 percent of 12- to 17-year-olds reported experimentation with cocaine. In 1987, even among eighth graders, rates for cocaine use were substantial: 5 percent reported that they had used cocaine at least once (see Table 3-11~. By the twelfth grade, the lifetime prevalence of cocaine use had increased threefold to 15 percent. 64Alcohol is not classified as an illicit substance here, despite the fact that it is illegal for adolescents under the age of 18 to purchase alcoholic beverages.

AIDS AND ADOLESCENTS I 189 TABLE 3-10 Trends Between 1979 and 1985 in Percentage of 12- to 1?-year-olds Reporting Use of Selected Drugs During Lifetime and Dunng Previous Year Reported Ever Using the Drug Reported Using the Drug in Last 12 Months Drug 1979 1982 1985 1979 1982 1985 Manjuana and hashish 30.9% 26.7~o 23.6% 24.1% 20.6% 19.7% Inhalants 9.8 NAa 9.2 4.6 NA S. 1 Hallucinogens 7.1 5.2 3.3 4.7 3.6 2.7 Cocaine 5.4 6.5 4.9 4.2 4.1 4.0 Heroin 0.5 b b b b b Nonmedical use of any psychotherapeutics 7.3 10.3 12.1 5.6 8.3 8.5 Cigarettes 54.1 49.5 45.2 13.3 24.8 25.8 Alcohol 70.3 65.2 55.5 53.6 52.4 51.7 N 2,165 1,S81 2,246 2,165 1,581 2,246 a Not available. b toss than 0.5 percent. CIncludes stimulants, sedatives, tranquilizers, and illicit use of analgesics or painkillers that are generally available only by prescnption. SOURCE: NIDA National Household Survey on Drug Abuse (1988:Tables 6 and 10). Heroin use is rare among the adolescents whose responses are cap- tured by surveys. The NIDA household surveys found that 0.5 percent of respondents indicated that they had used heroin at least once. Twelfth graders who participated in Monitoring the Future reported somewhat higher lifetime prevalence rates, with 1 percent of seniors indicating heroin use (see Table 3-111.65 Although national heroin use data indicate that drug injection is relatively rare among adolescents, local surveys find considerable van- ation in the proportion of respondents who report needle use. Surveys of students in Massachusetts and in the Bronx found that roughly 1 to 2 percent reported {V drug use (Strunin and Hingson, 1987; Reuben et al., 19881. Yet among 213 adolescent runaways between the ages of 11 and 16 who were interviewed in Texas, 7 percent had injected drugs (Hudson et al., 19891. Furthermore, the percentage of students in a CDC-assisted school survey who reported IV drug use (see Table 3-12) ranged from 2.S percent for Michigan to 6.3 percent for Washington, D.C. (CDC, 1988a). Among male students, however, almost 9 percent (i.e., 8.7 percent) of 65 Among high school seniors who participated in Monitoring the Future, 2 percent of males and 1 percent of females reported heroin use.

190 ~ AIDS: THE SECOND DECADE TABLE 3-11 Percentage of Schoolchildren Who Reported Ever Using Selected Drugs, by Grade in School Between 1985 and 1989 Those Who Reported Using the Drug at Least Once Drug 8th Graders 10th Graders 12th Graders Man3uana 14% 34% 50% Inhalants 21 21 l 9 Cocaine S 9 15 Crack 2 3 5 Heroin NAa NA 1 Other opiates NA NA 9 Stimulants NA NA 22 Sedatives NA NA 9 Tranquilizers NA NA 11 Cigarettes NA NA 67 Alcohol 74 85 92 NOTES: Data on 8th and 10th grade students are drawn from the National Adolescent Student Health Survey (N > 4,400 for all drugs). Data on 12th grade students are taken from the Monitonng the Future survey (N = approximately 16,800 for marijuana, cocaine, cigarettes, and alcohol; N is approximately 13,440 for inhalants and 6,720 for crack because only four fifths of students were asked about inhalant use and only two-fifths were asked about crack use). a Not available. SOURCES: Association for the Advancement of Health Education, Appendix B (Sampling and Weighting Procedures); University of Michigan, press release of February 9, 1990. the students from Washington, D.C., reported that they had injected illicit drugs. Differential Patterns of Drug Use Gender differences in drug use appear to be substance specific, and the gap between male and female use has declined over time (Kandel, 1980~. Drug use is generally higher among teenage boys; girls, however, are more likely to smoke cigarettes and to use stimulants. The magnitude and direction of racial and ethnic differences in drug use present a complex and sometimes contradictory picture.66 Unlike surveys of sexual behavior, almost all national drug use surveys report that illicit drug use is lower among black youths than among whites, with Hispanics falling in between (Bachman, Johnston, and O'Malley, 1981; NIDA, 1988; Handel and Davies, in press). Surveys that report high levels of drug use among blacks are generally restricted to poor 66Unfortunately, not all surveys collect race- or ethnicity-specific data. For example, the Monitoring the Future survey, although reporting on the prevalence of use by region, rural-urban location, college plans' and gender' does not provide specific rates for racial or ethnic groups.

AIDS AND ADOLESCENTS ~ 191 TABLE 3-12 Percentage of Secondary School Students Reponing IV Drug Use, by Sex and by Age for Selected Cities and States' 1988 Gender Age Site Total Femaie Male 13-14 15-16 17-18 CaiifOmiaO (N = 7,013; RR = 64%) Michigan (N=991;RR= 100%) WaShingtOn' D.C. 6.3 (N = 1,275; RR = 100~) San FranCiSCO (N = 802; RR = 88~o) 4.1% 2.6% 5.7% 2.8% 2.8 2.1 3.4 3.2 3.2 4.6 8.7 b 4.0 3.7 2.4 5.1 1.4 3.9 3.9% 4.3% 1.3 8.9 2.4 NOTES: The Ns indicate the size of the total sample, and they do not account for item nonresponse. RR refers to the school response rate, that is, the number of schools participating in the survey divided by the number of schools selected in the sample design. a The sample from California excludes students in San Francisco and Los Angeles. b Not shown because sample size was less than 40. SOURCE: CDC(1988a:Tables 1 and 3). black communities (Brunswick, Merzel, and Messeri, 1985~; yet even, for example, in a survey of inner-city Baltimore youth, blacks reported less drug involvement than whites (Zabin et al., 19861. The reasons why white teens report greater involvement with drugs than blacks are not clear. Kandel and Davies (in press) suggest several possible explanations, including differential underreporting, differential levels of drug involvement, and differential use of publicly funded treat- ment centers and emergency rooms, which are important sources of drug-related data. There are also some preliminary findings that indicate differential validity in reported data. An analysis of two waves of the National Longitudinal Survey of the Labor Market Expenence of Youth found that blacks were more likely than whites to provide internally discrepant responses to questions on illicit drug use. Among those who admitted using illicit drugs in the past year in a self-administered form in the 1980 survey, blacks were more likely than whites to deny in a subsequent interviewer-administered survey that they had ever used il- licit substances (Mensch and Kandel, 198Sb).67 That blacks may be more inclined to provide socially desirable responses is supported by findings 67Inconsistent responses were twice as common among blacks and Hispanics than among whites. Approximately 13 percent of black and Hispanic males and 22 percent of black and Hispanic females who reported marijuana use in 1980 claimed never to have used this drug upon reinterview in 1984. Seven percent of white males and 10 percent of white females provided similar inconsistent responses.

192 ~ AIDS: THE SECOND DECADE from Monitoring the Future. In the 1983 wave of that survey, non-drug- using black teens were twice as likely as white teens to indicate that they would not report drug use even if they had been involved in it (Johnston, Bachman, and O'Malley, 1984~. Age appears to play an important role in the initiation of use of most drugs.68 Surveys have shown that, for most substances, drug initiation rates peak in the late teens (Kandel, 19801. Analyses based on complete drug history data for the 1980 wave of a longitudinal study of a probability sample of tenth and eleventh grade New York State students in 1971 (a total of 1,325 students) indicated that the risk of initiating marijuana use begins at age 10 and reaches its height at age 18, declining substantially thereafter. Participants in this retrospective survey were between 24 and 25 years of age when the 1980 data were collected; of those who had used marijuana by the time they were interviewed, 90 percent had done so by age 20. Thus, if initiation did not begin by the end of the teenage years, it was unlikely to occur. A similar pattern was observed for the use of psychedelics, which peaks at age 21 (Kandel and Logan, 19841. However, the risk of initiating cocaine use does not terminate by age 20 but continues for a number of years thereafter (Kandel and Logan, 1984; Johnston, O'Malley, and Bachman, 19881.69 A similar pattern has been found in the Monitoring the Future sur- veys. Two-fifths of that sample (approximately 6,000 high school seniors) were asked to name the grade level at which they began using particular drugs.70 As Table 3-13 shows, 2.9 percent of teens began experimenting with marijuana in the sixth grade, and nearly 13 percents had initiated marijuana use by the end of the ninth grade. However, only 4 percent initiated marijuana use in the twelfth grade. Initiation of cocaine-use, in contrast, was more common in the later high school grades. Research indicates that early initiators of marijuana use are at much greater risk of becoming involved with other illicit Mugs (Yamaguchi and 68 For example, a recent survey of 3,454 secondary school students found age-related differences in the context of marijuana initiation (Bailey and Hubbard, 1990). The authors suggest that developmental factors may influence the context of initiation of this drug. 69The use of prescribed psychoactives also continues to rise beyond adolescence, particularly for fe- males (Kandel and Logan, 1984). 70The figures for the twelfth grade are somewhat misleading because the survey was conducted before the school year was finished. Rates of initiation are thus low, not only because the period of risk is truncated by the survey but also because recent initiators, who are undoubtedly light users, may be less likely than more practiced users lo admit involvement. 71 Table 3-13 shows percent initiating drug use in particular grades. For marijuana, 2.9 percent reported first use in 6th grade and a further 10 percent report first use during 7th-8th grades.

AIDS AND ADOLESCENTS ~ 193 TABLE 3-13 Percentage of Students in the High School Class of 1987 Reporting Initiation of Use of Selected Drugs, by Grade of First Use Grade Never Used ~g 6Ih 7th-8th 9th 10th 11th 12th the Drug ManjUana 2.9% 10.0% 12.3% 12.3% 8.2% 4.4% 49.8% Inhalants 2.5 3.3 3.6 2.7 3.4 1.4 83.0 Ha11UCinOgenS 0.3 0.9 1.9 2.5 3.3 1.5 89.7 Cocaine 0.2 0.6 2.2 3.7 5.4 3.0 84.8 Heroin 0.1 0.1 0.3 0.4 0.2 0.1 98.8 opiates Other than heroin 0.6 1.0 2.0 2.0 2.5 1.0 90.8 ShmU1antS 0.6 3.8 5.7 5.4 3.8 2.4 78.4 SedatiVeS 0.4 1.5 2.5 1.9 1.5 0.8 91.3 TranqUi1iZerS 0.4 1.6 2.6 2.6 2.4 1.4 89.1 A1COhO1 8.8 22.6 24.5 19.3 11.5 5.5 7.8 Getting drunk 3.3 13.8 20.3 17.8 11.9 5.7 27.1 Cigarettes 21.0 19.4 10.9 7.2 5.7 2.9 32.8 Cigarettes (daily Use) 1.6 5.2 5.3 4.4 3.3 1.6 78.7 NOTE: The enmes in this table show percent reporting mat their first use occurred In particular grades. The sample size for all percentages is approximately 6,000 persons. SOURCE: Johnston, O'Malley, and Bachman (1988:Table 15). Kandel, 1984b). Among white males in the 1984 National Longitudinal Survey of Youth who used marijuana, other illicit drugs, and cocaine (N = 926), the mean age of onset of marijuana use was 14.9 years compared with 16.9 years for those who only used marijuana (N = 1,6901. The mean age of onset of marijuana use for white females (N = 583) who reported that they used alcohol (or cigarettes), manjuaIla, and other illicit substances was 14.S years, compared with 17.2 years for those who resmcted their drug use to marijuana (N= 1,645) (Kande] and Davies, In press).72 72Respondents in the 1984 National Longitudinal Survey of the Labor Market Experience of Youth, who were between the ages of 19 and 27, were by venue of their age at very low risk of initiating marijuana use. The same is not true for cocaine. As a result, the differences reported here may be attenuated as the sample ages and some of the "marijuana only" group are transferred into the "multiple drug" group. However, data for an older sample (the 1980 New York State follow-up cohort) found the age differences in marijuana initiation for the two groups to be similar in magnitude to those seen in the national survey. Among males, the multiple drug users (N = 189) initiated marijuana use at 15.6 years of age compared with 18.3 years for the "marijuana only" users (N = 165). Among females We

194 ~ AIDS: THE SECOND DECADE Although the vast majority of teens who experiment with marijuana will not go on to inject drugs, early initiators of drug use are nevertheless at increased risk of turning to crack and IV drugs in later years, thus also increasing their risk for HIV infection. Research on adolescent drug use has provided a sense of the behavioral characteristics that distinguish adolescents who use drugs from their abstemious peers; it has also, with the help of longitudinal surveys, separated the antecedents from the consequences of use (Kandel, 1980~. Unfortunately, there has been little systematic research on the predictors of early drug initiation. Identifying and reaching early initiators of drug use will be important in preventing or remolding behaviors that confer risk for a variety of health problems, including AIDS. Clustering of Risk Behaviors Research on adolescents has provided compelling evidence of covar~a- tion among sexual activity, alcohol use, drug use, and delinquency (Miller and Simon, 1974; Hayes, 19871. The early onset of sexual intercourse does not appear to be an isolated behavior, nor does drug use. Rather, each is part of a complex pattern of interrelated activities. R. Jessor and coworkers (1980) surmise that the co-occurrence of problem behaviors (e.g., the use of alcohol or illicit drugs, criminal acts) may reflect for some adolescents a choice of lifestyle rather than a choice of particu- lar behaviors; the co-occurrence of these behaviors may thus reflect a single underlying tendency. This concept of a lifestyle or syndrome has found some support in the literature on adolescent problem behavior.73 The confluence of r~sk-associated behaviors argues for pursuing an un- derstanding of sexual and drug use behaviors within the larger social context of adolescence. Just as there are developmental stages of drug use, there may be a predictable sequence of problem behaviors in which one behavior, rather than being functionally equivalent to another, may actually constitute a risk factor for subsequent behaviors (Robins and Wish, 1977; Rosenbaum and Kandel, In press). A number of recent studies have examined the association between drug use and sexual activity. Elliott arid Morse (1989) reanalyzed data from the 197~1980 waves of the National Youth Survey, a household probability sample of 2,360 adolescents aged 11 to 17 at the time of the first interview. At the time of the initial survey in 1976, the proportion of mean age of marijuana initiation for multiple drug users (N = 132) was 15.9 years, compared with 18.9 years for the ``marijuana only" users (N = 224) (Kandel, Murphy, and Karus, 19851. 73For an overview of this literature and its problems, see Osgood and coworkers (1988).

AIDS AND ADOLESCENTS 195 boys reporting sexual activity in the previous year ranged from 10 percent for those who had not used drugs to 23 percent for users of alcohol only, 48.3 percent for combined alcohol and marijuana users, and 72 percent for those using multiple illicit drugs. In 1975-1976, Ensminger and Kane (1985) conducted a follow-up study of students from a poor, inner-city Chicago neighborhood who had been in the first grade in 196 ~1967 when they were first interviewed. Slightly more than half of the original subjects (705 of the original 1,242 children, or 56.7 percent) participated in the follow-up survey. The investigators found that those who drank were two to three times more likely to be sexually active than those who abstained from alcohol. Similarly, males who had used marijuana at least once were ten times more likely to be sexually active than those who had not used it, and females who had used marijuana were seven times more likely to report initiation of sexual intercourse than girls who were not involved with . . marlJuana. There are also data on the reverse of this phenomenon showing higher rates of drug use among sexually active teens. A 1981 study of 2,557 inner-city adolescents aged 12 to 18 in four Baltimore junior and senior high schools found that, regardless of race or gender, drug use scores (which reflected the type of substance and frequency of use) were higher for sexually active adolescents than for virgins (Zabin et al., 1986~. Elliott and Morse (1989), having established a relationship between sexual activity and drug use, attempted to determine the temporal order. They found that 2.25 times as many males and nearly 5 times as many females initiated drug use before sex rather than initiating sex prior to drug use. It should be noted, however, that their definition of drug use in this study was fairly comprehensive and included alcohol. Because the average age of initiation for alcohol use is very low, including alcohol in the definition of drug use tended to lower the are of initiation of drug use in the Elliott and Morse study. ~ cat Using data from the National Longitudinal Survey of the Labor Mar- ket Experience of Youth, Kandel and coworkers (Kandel and Davies, in press; Rosenbaum and Kandel, in press) investigated the association between prior drug used and sexual initiation by age 16. This study con- trolled for such covariates as race, religion, parental education, family structure, and personality (including delinquency and school characteris- tics). Retrospective data from the two youngest cohorts (i.e., the 2,711 individuals who were 19 to 20 years old in 1984) showed that, for both 74 Drug use was broken down into the following categories: (1) alcohol or cigarettes, or both, (2) marijuana, and (3) other illicit drugs.

196 ~ AIDS: THE SECOND DECADE males and females, the association of drug use and sexual behavior is considerable, even after the covar~ates above are entered into the predic- tion equation. Early sex is 1.4 times more frequent for boys who have used alcohol or cigarettes, or both, than for boys who did not report any prior drug use; it is 2.7 times more frequent for boys who have used marijuana, and 3.4 times more frequent for boys who have used other illicit Mugs. The association is even stronger for females. Early sex is 1.S, 3.5, and 4.9 times more frequent, respectively, for female users of the three categories of drugs than for nonusers. The association is also stronger for whites and Hispanics than for blacks, which may reflect the earlier age of initiation of sexual intercourse for blacks and the relatively limited time for drug initiation to occur prior to the initiation of sexual activity. Kandel and Davies (in press) also examined the degree to which early sexual activity is associated with an increase in subsequent drug use. Using the entire National Longitudinal Survey of the Labor Market Experience of Youth sample from 198475 and controlling for a number of potentially shared selection factors, Kande! and Davies found that early sexual activity was the most important predictor of cocaine involvement. In addition, among the 93 percent of males and 86 percent of females who were sexually experienced, the earlier sex was initiated, the greater the incidence of subsequent cocaine use. The predicted probability of using cocaine was 20 percent for males who waited until age 17 to initiate sexual intercourse and 12 percent for females who waited until age 18. Moreover, the frequency of intercourse in the 30 days prior to the survey was higher for cocaine users, especially males.76 The association of sexual activity and substance use was also shown in Mott and Haunn's (1988) analysis of retrospective data from the National Longitudinal Survey of Labor Market Experience of Your. Their work indicated that, at a given age, teenagers who use alcohol or marijuana at least once a month are more likely than nonusers to begin sexual intercourse in the following year. Similarly, those who are sexually active are more likely than virgins to report subsequent drug use. Yet although Mott and Haunn note that sexual activity and drug use are statistically related, they point out that the majority of young 75In 1984, the 12,069 respondents in the sample were between l9 and 27 years old. 76The frequency of intercourse applied to young adults between the ages of 19 and 27; there were no data to indicate the frequency among teens younger than 19 years of age. It is possible, however, that the same pattern of sexual behavior may also be seen among younger teens.

AIDS AND ADOLESCENTS ~ 197 adolescents (i.e., those under 16) were abstinent and had never used alcohol or manjuana.77 Thus, drug use and early sexual activity are, indeed, related. Prior use of drugs, both licit and illicit, significantly increases the risk of early sexual activity among adolescents. Furtherrr~ore, early sexual activity in- creases the likelihood of involvement with cocaine, a drug for which the age of initiation is typically later than that reported for initiation of sexual intercourse. The studies descnbed above clearly document that sexual activity and drug use are associated; however, research has not yet re- vealed whether the association is a simple function of shared antecedents. Are early sexual activity and drug use in adolescents determined by a common set of behavioral attributes? Does one behavior constitute a unique risk for the over? Additional research will be needed to answer these and other important questions. Therefore, the committee recom- mends that the Public Health Service give high priority to studies of the early initiation of risk-associated behaviors, including drug use and sexual behaviors, and of progression to the practice of multiple high-risk behaviors. Before proceeding, one further point should be emphasized. Al- though the committee recognizes the need for further research, it would reiterate that there is already sufficient information to conclude that early initiation of sexual behaviors or drug use carries the risk of HIV trans- mission for some fraction of American youth. This finding motivates the committee's call for AIDS interventions that target teenagers before they begin practicing risky sexual and drug-using behaviors. Subpopulations of Teens at Higher Risk As discussed earlier, national surveys are unlikely to reach youths who have dropped out of school or those who are not living in households. This latter group, which may be fairly sizable, Includes adolescents who live on the street or in residential institutions (e.g., youth detention centers) or who have joined the Job Corps. Reaching these teenagers may be particularly important in that some of these youths may currently be at higher risk of acqu~nng HIV infection than the general population of adolescents. There are several studies that support this perception of increased risk. A recent national survey of juveniles in custody, for example, found that 63 percent of these youths used drugs regularly (Bureau of 77That is, any use of marijuana, or monthly use of alcohol, or any history of sexual intercourse.

198 ~ AIDS: THE SECOND DECADE At six in the evening, four nights a week, a small white van loads up with passengers at a garage on Jerome Avenue in the Bronx. The van belongs to Project Streetbeat, a program of Planned Parenthood. The passengers are counselors. On this chilly Thursday night, as always, the van speeds downtown on the Bronx River Parkway, towards the heart of the South Bronx in search of teenage prostitutes. Loose packets of condoms, bleach kits and comic books about AIDS prevention, clutter the front seat . . . Social worker Larry Bilick is director of Project Streetbeat, and tonight he is driving the van. "This is an unreported life on the whole; and unreported conditions that these young people are living under. We don't hear almost anything about them. These young people are not runaways that come from other areas. They are mostly from the Bronx, they are mostly throwaways, it is not like they have some place else to go to, if only they wanted to go back home. Many of them have been victims of physical abuse andlor sexual abuse when growing up. Many of them have come out of the social service system of the city of New York; are casualties of that system. And, the circumstances of their lives have led them to these streets . . . " The outreach workers hand out condoms, bleach kits to sterilize needles, and to new clients, a comic book explaining safe sex in the language of the streets. Because most of the teenagers are homeless, living in abandoned buildings or in cars, they also give away donated clothing and dignity packs: plastic bags filled with basic necessities like a toothbrush, a comb and a bar of soap . . . The staff, most of whom are bilingual, don't preach or lecture, but they do gently urge these young people to come to Project Streetbeat's office where the staff can work to connect them with medical services, drug treatment programs, and shelter. Casanova Street is the busiest prostitution block In the neighborhood, mainly because so many of these young people live in the abandoned cars in the junk yard at the top of the road. Froggy is 20 years old and has been working Hunt's Point for the last five years. She gets her name from a voice that has been gnarled by years of crack smoking and began using Project Streetbeats's condoms a year ago. "You must see people coming round giving protections, clothes, that makes me feel good, cuz it makes me feel like they care. You know, like I'm somebody for them, you know?" In ache year and a half since Project Streetbeat began, the staff has had more than 3,000 different encounters with hundreds of prostitutes aged 13-21. One counselor has taken it upon himself to meticulously research and record the deaths of the different young people who he has worked with in the streets. Since the program began, he knows of 10 teenagers who have died; some from AIDS, some brutally murdered by pimps or johns. lithe deaths of these teenagers are not the stuff of newspaper obituaries, but for the counselors, it's the motivation that keeps the van driving through these streets, night after night. EXHIBIT 3-1 Edited News Report on Teenage Prostitutes in the Bronx by David Isay. SOURCE: Crossroads, National Public Radio Network, broadcast on WAMU, Washington, D.C., January 12, 1990.

AIDS AND ADOLESCENTS ~ 199 Justice Statistics, cited in the AMA Council of Scientific Affairs tAMA- CSA], 1990~. It is estimated that more than 500,000 juveniles are taken into custody through public detention or correctional facilities each year (AMA-GSA, 1990:9871. Smaller surveys of institutionalized adolescents provide insight into the problems of youths who reside in such settings, although the findings may not be generalizable. For example, interviews with 378 female adolescents aged 12 to 1778 residing in detention centers in the Bronx in the late 1970s found that the average age of initiation of sexual intercourse in this sample was 12 years. Virtually all (99 percent) of this sample were sexually active (Hein et al., 1978~. STDs have been reported as problems for both teenage boys and girls in detention situations.79 Almost one-fifth (17 percent) of 262 respondents to a survey of 16- to 17-year-olds incarcerated in the Los Angeles area reported a history of STDs (Morris et al., 1989), and almost half (47 percent) reported drinking alcohol in situations that led to intercourse. Among 184 juvenile offenders recruited through the San Francisco Youth Guidance Center, 35 percent reported exchanging sex for drugs or money, and airnost half (46 percent) agreed that "sex without condoms is worth the risk of getting AIDS" (Temoshok et al., 1989~. Comparing 802 high school students with 14- to 18-year-old juvenile offenders (N = 113) residing in a detention facility in the San Francisco area, DiClemente and DuNah (1989) found significant differences in the prevalence of risk-associated behaviors. Youths in detention were significantly more likely than students to report ever injecting drugs (12.9 percent versus 3.7 percent); they were also more likely to report more than three sexual partners (86.4 percent versus 15.1 percent) and an earlier age of initiation of sexual intercourse. Street youth are at greater risk than children who live in more fa- vorable conditions for virtually all medical disorders of childhood (IOM, 1988~. Consequently, it is likely that they would also be at higher risk for such health threats as drug dependency, HIV infection, and other STDs. A major contributor to higher risk for these adolescents may be prostitu- tion, which is one method teens employ to survive. For any youngster, 78The mean age was 14.8 years. 79Between July 1983 and June 1984, 285 teenage girls and 2,236 teenage boys between the ages of 9 and 18 residing in a New York City detention center participated in screening for STDs (Alexander- Rodnguez and Vermund, 1987). Fewer boys than girls (3 percent versus 18.3 percent) were found to have gonorrhea, and prevalence rates for syphilis were also lower for the boys (0.63 percent versus 2.5 percent). Most of those infected had no clinical signs. In a separate survey of 2,672 youths in temporary detention in New York City (Hein, Marks, and Cohen, 1977), 2,064 boys and 374 girls were asymptomatic for venereal disease. Following culture, however, 1.9 percent of asymptomatic males and 6.9 percent of asymptomatic females were found to have gonorrhea.

200 ~ AIDS: THE SECOND DECADE prostitution can be a particularly hazardous survival tool, given the lim- ited negotiation skills of young teens compared to those of their older patrons. It is important to note that an understanding of the prevalence of risk- associated behaviors among street youths is compromised by difficulties in reaching and engaging this population in surveys and research. Studies of this population have necessanly used small samples of convenience, relying on volunteers recruited from shelters and the streets. Although it is not possible to provide an accurate demographic and risk profile for street youths from these studies, the available research does offer a clear sense of the problems associated with living on the streets, and several findings are consistently reported across such studies. Interviews with street youths indicate that this group is, indeed, at risk for HIV infection, and the epidemiological evidence suggests that the prevalence of HIV infection is higher among street youths at least in New York City (see Table 3-2) - than among adolescents in general. Moreover, although the precise number of homeless youths who use drugs is not known, small surveys of this population have consistently found that the majority report drug use (Shaffer and Caton, 1984; Rothman and David, 1985; Yates et al., 1988~. Most street youths who have participated in studies also report that they are sexually active, but few report the regular use of condoms (Shaffer and Caton, 1984; Yates et al., 1988; Hudson et al., 1989; Rotheram-Borus et al., 1989~. Another subpopulation of teens who are at higher risk for HIV infection comprises adolescents who are raped or sexually abused (Gellert and Durfee, 19901. Data from the 1987 wave of the National Survey of Children8° found that 7 percent of 18- to 22-year-olds in this country report experiencing at least one episode of nonvoluntary intercourse; approximately half of the experiences reported by women occurred before the age of 14 (K. A. Moore, Nord, and Peterson, 19894. Conclusion In assessing adolescents' risk of acquiring HIV infection, it is clear that some teens are more likely than others to engage in the behaviors that are known to transmit HIV. En addition, the number of adolescents at risk because of a specific behavior vanes in part according to the behavior. Although many teens experiment with alcohol, relatively few report the use of crack or injection of illicit drugs. Nevertheless, a substantial gOThe 1987 wave of the National Survey of Children followed up on experiences reported by subjects during the first two waves. Telephone interviews resulted in an 82 percent response rate from persons interviewed in the second wave. Analyses reported above are based on 1,121 respondents aged 18-22.

AIDS AND ADOLESCENTS | 201 proportion of teens are sexually active, and although recent data indicate that more teens than ever are using condoms, there remains a sizable subpopulation who are still engaging in unprotected intercourse. As is discussed at the end of this chapter, the differential distribution of risk- associated behaviors has important implications for intervention efforts. It is clear, however, that early onset of risky behaviors can have grave consequences, and it is therefore necessary to reach these youth before such behaviors begin. For youth who are already engaging in risky behaviors, it will be necessary to intervene as quickly and as effectively as possible to facilitate protective change in these behaviors. INTERVENING TO PREVENT FURTHER SPREAD OF INFECTION The section that follows considers intervention strategies to retard the spread of HIV infection among teens. These interventions are motivated by the risky behaviors detailed in the preceding section. An examination of the contentious question of appropriate goals for AIDS prevention efforts aimed at teenagers is followed by a description of interventions to reach different segments of the teenage population. The section concludes with a discussion of the resources needed to serve the segments of the teenage population who are at relatively high risk for HIV infection. Of paramount importance, however, is the continuing need for cou- pling careful evaluation research to future intervention Programs. Role of Evaluation Research 1 ~ The committee strongly believes that progress in the development of successful intervention programs will depend on an iterative process in which programs are implemented and their effects assessed, following which new and better interventions are designed and tested. Toward this end, the committee recommended in its first report that "planned variations of key program elements be systematically and actively incorporated into the design of intervention programs at an early stage" (Turner, Miller, and Moses, 1989:307~. The committee reiterates this recommendation, noting that such designs can provide invaluable evidence of the effectiveness (or ineffectiveness) of different intervention strategies. It is the committee's fimn opinion that this strategy~oupling AIDS prevention programs with rigorous research efforts to determine the effects of those programs~ffers the best hope of producing effec- tive behavioral interventions for the diverse subgroups that constitute the adolescent population. Such efforts are the key to improving the inade- quate knowledge base that presently hobbles efforts to design effective

202 ~ AIDS: THE SECOND DECADE interventions. To further encourage the use of such designs, an evaluation panel of the committee has issued a review of research and intervention strategies that can be used to evaluate mass media campaigns, testing and counseling programs, and interventions designed by community-based organizations (Coyle, Boruch, and Turner, 19901. In the following section, the committee offers, to the extent possible, suggestions for AIDS prevention strategies that currently appear promis- ing for retarding the future spread of HIV among teenagers. Owing to the lack of adequate scientific evidence, readers will note that, although the committee was able to descnbe with relative confidence the venous patterns of adolescent behavior that place young people at risk of ac- quiring or transmitting HIV, there is much less certainty about the types of programs that can help adolescents modify those risky behaviors to avoid infection. Thus, all suggestions for intervention programs carry a concomitant recommendation that the interventions be earned out in a manner that permits the collection of evidence of their effectiveness. It is only through such efforts that effective interventions will be designed. Goals of Intervention Programs for Teens Within the adolescent population, there is tremendous variation with respect to risk taking and potential exposure to the AIDS virus. Similarly, the committee notes that considerable controversy surrounds some of the adolescent behaviors that transmit HIV infection. This section considers in turn, the controversy concerning the general goals of AIDS prevention programs for teenagers and specific goals for three segments of the teenage population. General Goals With regard to the use of drugs that are associated with HIV transmission, there is a consensus both within our committee and in the nation as a whole that such drug use is, in itself, physiologically destructive and psy- chologically debilitating. Thus, AIDS prevention programs for teenagers and adults properly discourage drug use among all persons. For persons who do use drugs, AIDS prevention programs have the goals of (1) discontinuance of drug use, if possible; and (2) if discontinuance is not possible, implementation of practices to reduce the risk of HIV transmis- sion (e.g., by using sterile needIes). In its previous report, the committee concluded that, regardless of the availability of treatment opportunities 8} As discussed by Tumer, Miller, and Moses ( 1989:Chapter 5) and Coyle, Boruch, and Turner ( 1990), an optimal research strategy is likely to involve rigorous evaluations of selected AIDS prevention programs rather than routine evaluation of all interventions.

AIDS AND ADOLESCENTS ~ 203 and programs intended to prevent drug use, a substantial number of peo- ple in the fruited States would continue to inject drugs, at least in the short run. Some of these persons Will be teenagers and young adults. Consequently, the committee believes that AIDS prevention programs should encourage these drug users to seek treatment; they should also ensure that these young people are made aware of all effective methods for reducing their risk of contracting or transmitting HIV. The commit- tee wishes to emphasize that the goals of prevention and risk reduction are not contradictory. None of the current studies on "safer injection" programs82 have shown an increase in TV drug use as a result of making sterile needles available or promoting injection equipment stenlization using bleach. Indeed, it appears that safer injection programs may indi- rectly encourage IV drug users to seek treatment. Furthermore, although results are still preliminary, many studies indicate that such programs do reduce the risk of HIV transmission among IV drug users (Buning, Coutinho, and van Brussel, 1986; Jackson and Neshin, 1986; Jackson and Rotkiewicz, 1987; Chiasson et al., 1987; Watters, 1987; Buning et al., 1988; Stimson, 1988, 1989; Hartgers et al., 1989; van den Hock, van Haastrecht, and Coutinho, 19891. With regard to premarital and same-gender sexual behaviors, there is no similar consensus in this country. There is instead ample evidence that these behaviors are extremely controversial subjects. It will be seen from Table 3-14, for example, that there is a substantial divergence in the opinion of American adults about sexual intercourse among persons who are not married. Although there was some increase in tolerance of sexual intercourse among unmarried persons over the last two decades, roughly one-third of Americans report that they believe premarital sexual intercourse is "always wrong," whereas another one-third report that they believe it is "not wrong at all."83 Given these divisions in public opinion, it is not surprising that fed- eral AIDS education efforts have stumbled for several years over this issue. Much of the controversy involves disputes about the need to of- fer "realistic" advice regarding the protective value of condoms versus counterclaims that the AIDS epidemic requires moral education to pro- mote abstinence from sexual activity prior to maJnage and fidelity within 82 See Chapter 2 of this report and Tumer, Miller, and Moses (1989:Chapter 3) for a detailed review of research in this area. 83 Estimates are derived from surveys of probability samples of the noninstitutional adult population of the continental United States conducted by the General Social Survey program of the National Opinion Research Center (University of Chicago; see Davis and Smith, 1989). The estimates have been weighted to reflect the varying probabilities of selection for persons in households with different numbers of eligible adults.

204 ~ AIDS: THE SECOND DECADE TABLE 3-14 Distnbutions of Responses (percentage) to a Question About Premantal Sexual Relations Tabulated from Surveys of Probability Samples of U.S. Adults, 1972-1989 Attitude Toward Premarital Sex Almost Always Always Sometimes Not At All Year Wrong Wrong Wrong Wrong N 1972 35.7% 11.4% 25.2% 27.7% 1,537 1974 33.4 13.0 23.9 29.7 1,492 1975 30.6 12.4 25.2 31.7 1,427 1977 30.7 9.9 23.1 36.3 1,481 1978 29.1 12.3 20.4 38.2 1,494 1982 28.4 9.1 21.7 40.7 1,794 1983 28.1 10.7 24.6 36.6 1,561 1985 28.1 9.0 20.0 43.0 1,482 1986 28.2 8.8 22.8 40.2 1,425 1988 26.2 10.2 22.2 41.4 955 1989 27.7 8.8 23.1 40.4 971 All YearS 30.0 10.6 23.0 36.5 15,556 NOTE: The question was as follows: If a man and a woman have sex relations before mamage, do you think it is always wrong, almost always wrong, wrong only sometimes, or not wrong at all? Tabulations were weighted by the number of adults in the household so correct for different probabilities of wi~in-household selection in households of different size. The Ns shown in the table are unweighted. SOURCE: Tabulated from the General Social Survey conducted by the National Opinion Research Center, University of Chicago (Davis and Smith, 1989). mamage.84 Not surprisingly, this controversy has been most intense with regard to AIDS education for teenagers and young adults. In 1987, for example, the U.S. Senate voted 94 to 2 in support of an amendment to the Department of Health and Human Services appro- pnations bill that requiem all AIDS educational matenals and activities for young adults and school-aged children to emphasize "abstinence from sexual activity outside of a monogamous ma~Tiage."85 The impact of such legislation can be seen in the constraints that have been incorporated in AIDS prevention programs mounted by the Public Heals Service. CDC's "Guidelines for Effective School Health Education to Prevent the Spread of AIDS," for example, propose that, for young people who have already begun to engage in sexual intercourse. "school oro~ams should enable r- -an- - 84See, for example, TuIner, Miller, and Moses (1989:Chapter 7) from which this discussion is adapted. 85 Congressional Record, OCtOber 14,1987, S-14217.

AIDS AND ADOLESCENTS ~ 205 and encourage them to stop engaging in sexual intercourse until they are ready to establish a mutually monogamous relationship within the context of manage" (CDC, 19886:41. The CDC guidelines do recog- nize, however, that some young people will not follow this advice, and they recommend that school systems, in consultation with parents and health officials, provide information on other strategies to prevent HIV transmission. Guidelines such as these for teenagers and young adults do not reflect the pluralistic nature of beliefs in this country but attempt to impose a particular set of values on all adolescents. They also fly in the face of the data presented earlier in this chapter, which indicate that most contemporary teenagers in the United States begin sexual relations during their teens, that a sizable fraction begin intercourse in the early teens, arid that a substantial portion of sexually active teens report having intercourse at least once a month. Indeed, more than 1 million teenagers become pregnant each year,86 and more than 400,000 of these pregnancies occurred in young women 15 to 17 years of age (Hayes, 1987:5~55 2611. In the context of a deadly, sexually transmitted epidemic, the com- mittee believes that AIDS prevention programs must heed the data on nsk-associated behaviors reported by the adolescents themselves and not be sidetracked by wishful thinking about patterns of behavior some might hope teenagers would follow. In 1986, Surgeon General C. Everett Koop set a high standard for rational discourse on AIDS prevention for ado- lescents. On release of his report on AIDS, which dealt frankly with the behaviors that transmit HIV, the surgeon general observed:87 Controversial and sensitive issues are inherent in the subject of AIDS, and these issues are addressed in my report. Value judgments are absent. This is an objective health and medical report, which I would like every adult and adolescent to read .... Many people especially our youth are not receiving, information that is vital to their future health and well-being because of our reticence in 86In 1984, it iS estimated Mat just over 400,000 teenage girls obtained abortions (Hayes, 1987:261). More recently, Trussell (1988:262) has estimated that `'One out of every 10 women aged 15-19 in the United States becomes pregnant each year, a proportion that has changed little during the past 15 years." Trussell's calculations yielded an estimate of roughly 860,000 pregnancies (837,000 pregnan- cies among 15- to 19-year-olds and 23,000 among girls aged 14 or younger); however, he notes that these "are underestimates of pregnancies because spontaneous abortions are ignored and because data are tabulated by age at the resolution of pregnancy instead of age at conception" (p. 262). (Hayes [1987:54-55] estimated that 134,000 teenage pregnancies ended in miscarriages in 1984.) 87C. Everett Koop, public statement made at press conference upon release of Surgeon General's Report on Acquired Irnrnune Deficiency Syndrome, Washington, D.C., October 22, 1986.

206 ~ AIDS: THE SECOND DECADE dealing with the subjects of sex, sexual practices, and homosexuality. This silence must end. We can no longer afford to sidestep frank, open discussions about sexual practices homosexual and heterosexual .... As parents, educators, and community leaders we must assume our responsibility to educate our young. The need is critical and the price of neglect is high. AIDS education must start at the lowest grade possible as part of any health and hygiene pro:,~am. There is now no doubt that we need sex education in school and that it should include information on sexual practices that may put our children at risk for AIDS. Teenagers often think themselves immortal, arid these young people may be putting themselves at great risk as they begin to explore their own sexuality and perhaps experiment with drugs. The threat of AIDS should be sufficient to permit a sex education cumculum with a heavy emphasis on prevention of AIDS and other sexually transmitted diseases. This committee concurs wholeheartedly with these objectives. Specific Program Goals As noted previously, there is tremendous variation in the behaviors of different segments of the adolescent population and in the risk of HIV infection faced by these subgroups. The committee believes that inter- vention programs must reflect this diversity of risk and thus proposes specific goals for three groups of teens. Program Goals for Adolescents Who Are Not Engaging in Risk- Associated Behaviors. Intervention programs should seek to provide information, motivation, skills, and practical assistance to help these young people avoid future risks and to involve them in current AIDS prevention activities. The ultimate goal of AIDS prevention is to block HIV transmission, and programs should accommodate the range of chal- lenges young people will face and the variety of choices they may make. Abstinence, delay of intercourse until marriage, and other traditional be- havioral patterns are effective ways of eliminating the risk of sexually transmitted HIV infection if in fact these patterns are enacted. Because some teens, however, will choose to begin sexual activity, all teenagers should be educated about the protective value of condoms and sperrni- cides. In addition, all teens should be educated about the dangers posed by the use of illicit drugs. Intervention programs must reach youth in their very early teen or preteen years. In this regard, it is worth remembering that 23 percent of young women in the United States report that they have engaged in sexual intercourse by their 16th birthday, and 4 percent report engaging in sexual intercourse by their 14th birthday. Given the serious consequences often associated with such early initiation of intercourse, the committee 88For the birth cohort (N = 897) born 1962-1964, the National Survey of Family Growth found that

AIDS AND ADOLESCENTS ~ 207 recommends that AIDS prevention programs make special efforts to reach very young teens and, in some subpopulations, to reach youth before they enter adolescence. The committee also believes that interventions should ensure that youth who engage in male-male sexual contacts have sufficient knowl- edge to protect themselves in such encounters. As noted earlier in this chapter, estimates denved from a probability sample of Amencan men in 1970 (Fay et al., 1989189 suggest that a minimum of 20 percent of Amencan males have male-male sexual contact to orgasm at some point in their lives and that most of these men have their first such experiences during adolescence.90 Program Goals for Adolescents Who Are Engaging in Sexual In- tercourse but Who Are Not Using Illicit Drugs. Intervention programs for these young people should educate them about the dangers of drug use and seek to facilitate protective changes in their sexual behaviors. Although education about abstinence may be valuable, clear advice con- cerning the protections offered by condoms and spermicides should also be offered. In this regard, the committee wishes to draw attention to the parallel approach recommended by another National Research Council committee that addressed a more common and less deadly consequence of adolescent sexual behaviors: unwanted pregnancy. In recommending policies for dealing with adolescent pregnancy, that committee struggled with many of the same controversies that surround AIDS-related discus- sions. After two years of deliberations they concluded the following: Sexually active teenagers, both boys and girls, need the ability to avoid preg- nancy arid the motivation to do so. Early, regular, and effective contraceptive use results in fewer unintended pregnancies. Delaying the initiation of sexual activity will also reduce the incidence of pregnancy, but we currently know very little about how to effectively discourage unmarried teenagers from initiating intercourse. Most young people do become sexually active dur- ing their teenage years. Therefore making contraceptive methods available and accessible to those who are sexually active and encouraging them to 23.1 percent of women reported that they had had intercourse by their 16th birthday, and 4 percent reported intercourse by their 14th birthday (Hofferth, Kahn, and Baldwin, 1987:Table 3). 89These estimates are based on a national sample of the adult male population of the United States in 1970. 90There are only a few research efforts now under way tO study gay youth. Several of the projects now being conducted (one in New York as part of the HIV Center for Clinical and Behavioral Studies funded by the National Institute of Mental Health; others directed toward street youth in Baltimore, Maryland, and Belo Horizonte, Brazil [Rolf et al., 1989]) are promising but are unlikely by themselves to be sufficient. Clearly, more research will be needed to understand the determinants of same-gender sex among adolescents and to identify the most effective mechanisms for reducing AIDS-associated risk in that age group.

208 ~ AIDS: THE SECOND DECADE diligently use these methods is the surest strategy for pregnancy prevention. (Hayes, 1987:262) Program Goals for the Small Groups of Adolescents Engaging in Multiple High-Risk Behaviors and Those Adolescents Who May Already Be Infected with HIV. Intervention programs for these teens should make every effort to assist them in altering the behaviors that place them at risk and should seek to alter any social or economic conditions that support their risk taking. Teens who are using illicit drugs, especially those who inject drugs, should be encouraged to seek treatment. Such a policy in turn requires effective referral networks to treatment programs that are tailored to adolescent needs. For example, residential drug treatment programs have been designed to alter teen lifestyles, recreational activities, patterns of association, and other factors relevant to their Hug use. It should be noted, however, that for the most part methadone treatment programs are not available to adolescents. Those adolescents who continue to use needles should be urged not to share them or other injection equipment. Adolescents who are living on the streets, engaging in prostitution, or exchanging sex for crack will require additional services, such as shelter, counseling, and medical care. These adolescents who are at highest risk of infection should also be made aware of the availability of HIV testing and counseling, how this service is delivered (confidential versus anonymous testing), and the significance of the information provided. Teenagers known to be infected with HIV will require information and counseling regarding the potential consequences of their infection, including its possible effects on future childbearing and on sexual and Hug use partners. They should also be given advice about securing the medical and social services they may need in the future. The difficulties manifested by this last group of teens present the greatest challenges to AIDS prevention and arise from some of the nation's most severe social problems. They also highlight the lack of success previous efforts have had in dealing with these problems prior to the onset of the AIDS epidemic. These problems remain equally difficult to solve today, but AIDS has increased the price that will be paid for failing to deal with them. WHAT DO TEENS KNOW ABOUT AIDS? National surveys of high school students sponsored by the CDC (1988a)9~ 91In 1987 CDC (1988a) began collecting data from a national sample of high school students in grades 9 through 12. These data are intended lo help local departments of education assess local knowledge,

AIDS AND ADOLESCENTS ~ 209 together with local surveys92 have found that virtually all students have heard of AIDS and the majority know about the routes of transmission. Fewer teens, however, are aware of actions that can be taken to prevent the acquisition of infection (DiClemente, Zorn, and Temoshok, 1986; Strunin and Hingson, 19871. (See Tables 3-15, 3-16, and 3-17 below.) Moreover, misconceptions remain among adolescents concerning the role of casual contact in the transmission of HIV (DiClemente, Zorn, and Temoshok, 1986; Strunin and Hingson, 1987; CDC, 1988a, DiClemente, Boyer, and Morales, 1988; DiClemente, 1989; Siege] et al., 19891. Findings from a national sample of students in the eighth and tenth grades (the National Adolescent Student Health Survey) indicate that approximately 90 percent of these teens knew that the AIDS virus could be transmitted by sexual intercourse or by Shannon needles (see Table 3- 151. Similarly, results of a 1988 survey of a probability sample of 16- to 19-year-olds in Massachusetts (Table 3-16) indicate that virtually every young person in this sample was aware that AIDS could be transmitted by sexual contact or injection of drugs. Yet approximately half of teenagers in the national sample thought that AIDS could be contracted from donating blood (see Table 3-17), and substantial fractions of teenagers in the 1988 Massachusetts sample believed that AIDS could be transmitted by saliva (39 percent) or giving blood (51 percent). Even more disturbing is the number of teens who report misperceptions about ways to reduce the risk of contracting AIDS (see Table 3-171. For example, 27 percent of white males and more than 40 percent of black and Hispanic males in the national sample said that they believed that the risk of AIDS could be reduced by washing after sex. Similarly, 30 percent of males arid 21 percent of females thought that "making sure your partner looks healthy" would help protect them from AIDS. Perhaps most telling, however, are beliefs, and behaviors; the findings can inform the content and planning of education programs and can be used to monitor changes over time. In the survey, students completed a self-administered, anony- mous questionnaire that was developed collaboratively by CDC and 24 state and local departments of education. Core questions provide information on demographic characteristics? HIV beliefs and knowledge, and behaviors associated with transmission. Most sites used a geographically stratified cluster sample, randomly selecting schools within strata, then selecting classes within schools. Other sites used a random sample of schools, then randomly selected students at each school. Response rates of schools varied from 52 to 100 percent. 92There have been many surveys of adolescents' knowledge, attitudes, and beliefs about AIDS, but these "KAB" surveys are of variable scope and quality. Some have surveyed national populations; others have relied on small, local samples. Some surveys have used probability sampling techniques; others have relied on samples of convenience. Variation in the wording of questions and the manner in which questions are posed complicates comparisons of different surveys. In reviewing different KAB instruments, it is important to keep in mind that the populations sampled, the recruitment process, and the methods for collecting data vary. Such variation means that caution is called for in making generalizations and looking for trends.

210 oo - l oo - - ._ C) ._ ~: - C~ C~ ~D - o - C) e~ ~L C~ C) ._ Pt V) 04 ._ o - C) o v _- C~ o C~ _4 Ct ~Ce, ~ oo C~ ~4 - ~: - - C~ - o o C) C) V~ - m ._ Ct ~L cq ._ - ._ - 6 C) ._ - CC s U. ._ - C) Ct C) - - 3 6 U. ;> CQ _. 6 - o o ._ C~ C~ ._ C~ Ct _ ~ oo oo — ', — d — oo ~ oo oo ~ oo o C~ ~ _ ~ o · · · · · · C`3 oo ~o ao oo ~o oo ~ C~ ~ oo o C~ . . . . . ,-~\ ~ ~ ~ ~ o oo Ch oo ~ oo _ oo ~ =, o oo oo [~ ~ ~ ~ oo o ~ O O ~ ~ ~ cr ~ oo OC ~ ~ 00 oo ~n ~ ~ O O O ~ — ~D oo ~ oo ~ 00 V~ 00 ~ ~ 0 ~ . . . . . ,~K '} ~ C~ _ ~ ~ oO =~ 00 oo 00 ~ _ 00 ~ O ~ X ~o ~n ~ d- ~=N —) oo ~ oo t— oo — C) ~ C) ~ O C) C) 3 ~ ~ £ e~ , o o ~ £ ~ ou ._ _ 3 ~ C~ ~2 ~ . _ o u' C) ~ ._ 6 _ C) Ce .~: ~ _ X C~ ~ o — ~ ~ ~ £ ~-! ~— ._ ~ 3 ~ ~ >~ ~, ~o _ ~ _ _ .= ~ ~ ,, o ~ ~ _ _ =, ~^ o o o ~ ~ ~ ~ ~ o ~ ~ ~ ~ ~o ~ o - 3 x C) C~ o d: _ ~ Ct ._ ~ '¢ C~ C~ CO 5 ~ ~ 5 C) S o~ ._ ~ O O C) ~ C) o ~ — ~ .= .= ~ C) 3 ~ ~ _ , °. ~ X — C, ~ ._ s o' ~ ~ ~ ~ ~; C) ~ . _ C _ _ ._ ._ _ C) ~ ~ C) O m C) C~ ec C) C) ~ 0 C,, - - - C~ o C) o - o C~ 3 - C~ ~ ~: o X ~ ~ _ .= .° o ,~ ~ == 3 ,~c D ~ c~ ~ o4 ~ ~ _ .5 o ~ 3 ~ o ~ ~ ~ s _ 3 ~ ~ .c ~ s ,, o, oO c ;^ o o o ~ o-° ~ = = E ~ .c ~ ~ ~ ~ o ~ O 4, D 3 ° .— o o ,; = `= 5 c t~o ° 3 ~ ~ C ~ o ='.^= < C ~ "s 3 ~: '' > 2 c ~ o~ s - ' o CZ ~,¢ o -> ,_ ~ :~: ~ 3 ~ <: Ce ee ~ s ~ 3 ~ ~ o _ ~ .< ._ ° ca = 04 - -o C>' O ;'' ~ ~ > C~ s - CC - - C~ ~ — s C~ ~ C~ o ~; 04 C _ ~ o ~ _ CC 7 O — _ X C X C~ o _ e~ oo - — S ._ _ C o 5 C) - - D . . C~ o cq

AIDS AND ADOLESCENTS I 211 TABLE 3-16 Percentage of 16- to 19-Year-Olds in 1988 Survey of Probability Sample of Massachusetts Teens Who Reported Believing That the AIDS Virus Could be Transmitted by Various Means Percentage Who Believed It Was HIV Transmission Route Means 1986 1988 Sex between two males 99% 98% Sex between a male and female 91 99 Injecting drugs 91 99 Vaginal fluids 69 88 Semen 76 93 Getting a blood transfusion 93 89 Kissing someone on the mouth 58 25 Giving blood 61 51 Tears 19 17 Saliva 60 39 Sharing eating or drinking utensils 38 11 Toilet seats 15 5 N 829 1,762 SOURCE: Hingson, Strunin, and Berlin (l990:Table 2). the results presented in Table 3-17, which indicate that one teenager in every five believed that persons infected with the AIDS virus could only spread the virus if they were sick with AIDS. This substantial disbelief in the risk posed by asymptomatic but infected individuals and the misperception of protection afforded by washing are cause for concern. Furthermore, although many of the basic facts about AIDS appear to be reaching teens, the segments of the adolescent population that are epidemiologically most vulnerable appear to be less well informed about this disease than the majority of youths.93 The data in Table 3-15 provide evidence, for example, that minority youth are less aware than white youth of the behaviors Mat risk HIV transmission. Other studies have found racial differences with respect to beliefs about the benefits afforded 93 For example, a survey of 1,869 students from a community college located in the South Bronx, an epicenter of the epidemic, found that only 69 percent recognized that sexual intercourse without a condom increased risk for HIV infection, and slightly more than half (55 percent) were aware of the risk of vertical transmission if a pregnant woman reported sexual contact with an IV drug user within the past five years (Lesniek and Pace, 1990).

212 of ~ := Cry 5 2, Cal ~ C) D c, O o ~ _ ~ an, ~ _ - ~: o ._ - C) AL ._ e4 ._ o Cal _% U: O ~ _ Cat =5 _ oo Cal Ct so of ~ , U. - C~ C~ C) o o C) C~ _} C) C~ - ~L U: ._ ~: e~ m C) ._ - ~ O & ~ C) a: & ~ '~= - ; ~ ~ _ 00 —. ~ _ ~ _ . . . X — U~ \0 · V~ ~ O \0 0 C~ 00 0 _ ~. —; ~U V~ 00 . . . . . 0 C~ ~ _ — ~ ~ ~ ~ oo ~ r~ C~ 00 00 a0 00 C~ _ ~ ~ U~ ~ ~ ~. 00 ~ U~ _ _ Cx ~i ~ ~ _ ~t _ ~ ~ 0 ~ 00 _ C~\ _ 00 t_ _ ~ _ ~ S ~3 ~ ~ 3 04 & . _ u' ~ £ ° ~ ~o ~ o s~ ~ ~ o o ~ _ ~ 3 ~ <= <: > 0S ° C<,) c,. _ ~ ,¢ o ,¢ X ~ s - ~ - ~ =3 ~ ~ ~ o ~ C~ — ~ o' _ ~C ._ ~ . 1O C) C) — C) ~ ._ ~n C) ~ _ - CQ - ~: U: - o c C~ ._ V' a ~ ~: - c S S ;t C) - o C) - - C) - o - C~ ._ c c) c S C) ~: S - o C) = ~C - - - O ~ ~ :~` ;- e~ ~ D S ~ ~ C) ~ ~ C, — D ~4 S C~ S = o 3 `~ ~ ~ . ~ _ ~ r~ ~ ~ 3 ~ o - _ 3 ~ _ ~ ~ —~ C D · O — . _ ce C _ C: .— _ C¢ O ~ —C C) S = ~ = D ·_ o() ~ ~ . _ ~ _ .. 3 ~ ° c · , = 0 V = = .o x ~ "c - ~ ~ - ~ ~ o - ~ — ee O ·— ~ - ~o - ~ "o ~ - o CD 4-) ·_ S' 2 ' C ',, — .' _ 3 ~ ~— ¢, ~ C - - C, ^ c .' ~ ~ 3 ~ _ .E oX ~ s C _ C~ — ~ ~ C) _ e5 0 _ l_ c, _ C) oc ~ ~ C" o ~ _ =, ~ _ ~ o C ~ ~ ~ _ (e ~: _ C} ~ .D C ~ ;:s C,~ C~ _, — s o I V, o o ~ X <: & — — s s :~ c) D 0) ~ _ ~ ~ ~ O tJ 3 ^_ o ~ ~ -~d ~ C~) 'G _ .— o O ~ C E D ~ 0 = 0 s _ - 3 ~ ° ° C) ~

AIDS AND ADOLESCENTS ~ 213 by condom use.94 However, it should be noted that misconceptions can be corrected. The Massachusetts survey shows considerable improvement of teens' assessment of transmission risks between the 1986 and 1988 cycles (see Table 3-161. There is a relative abundance of survey data on what adolescents know about HIV transmission, but there is very little information about the numbers of teens who may be participating in nsk-associated behaviors. Thus, future information efforts might best be targeted to young teens not yet exposed to information campaigns. More important, these efforts should collect behavioral data. Finally, given the evidence reviewed in earlier sections of this chapter, all intervention programs should seek to motivate teenagers to avoid the use of drugs and to protect themselves against the risks of sexually transmitted HIV . ~ . Injections. Fear REACHING ADOLESCENTS Aspects of Prevention Programs The committee has previously noted95 that STD prevention programs have relied to varying extents on threatening messages that evoke high levels of fear. Similar tactics have been used in programs to prevent AIDS. (For example, such messages as "Bang Bang You're Dead" have been used to call attention to the fatal consequences of sexually trans- mitted HIV infection.) Whether such messages have been effective in changing behavior is not known because there have been no controlled studies or evaluations of their impact. Research suggests, however, that the usefulness of frightening and uninformative messages has particular limitations. Messages designed to evoke high levels of fear or those that rely ex- clusively on threats may be intuitively appealing in the case of preventing a deadly disease, but usually they have been shown to be effective for most people only if coupled with advice about how behavioral change 94DiClemente, Boyer, and Morales (1988) found that minority adolescents were less knowledgeable about protective measures than their white peers. Using data from a 1985 survey of 628 students in the San Francisco school district, they found that 71.7 percent of white students versus 59.9 percent of blacks and 58.3 percent of Hispanic students believed that condoms could reduce the risk of acquiring AIDS. There may be some inconsistency in findings across studies. Siegel and colleagues (1989), for example, reported little variation in factual knowledge of AIDS across racial or ethnic groups in a survey of 1,940 junior high school students in Oakland, California; however, only 5 percent of the Oakland students were white. 95 See Turner, Miller, and Moses (1989:266-268). The text in this section is adapted from that treat- ment.

214 ~ AIDS: THE SECOND DECADE can reduce the threat (Sutton, 1982; Becker, 1985~. For example, educa- tional messages in the syphilis prevention campaigns undertaken by the military early in this century sought to arouse fear in the troops by us- ing STD prevention films (e.g., "Fit to Fight" and "Fit to Win") (Brandt, 19879. Premovie and postmovie measurements of knowledge about STDs revealed that these strategies changed general impressions (e.g., horror and fear were increased and persisted for weeks after the viewing), but knowledge and behavioral changes did not persist (LashIey and Watson, 1922~. In contrast, during World War IT, a prophylaxis program based on condoms and treatment was initiated, and soldiers responded favorably. As many as 50 million condoms were accepted by soldiers each month, and rates of syphilis declined in this population over that time (Brandt, 19871. Nevertheless, the lack of systematic evaluation (a shortcoming of most STD programs) of the message and of prophylaxis programs foils any attempt to draw conclusions as to what factor or factors d~d or did not work in reducing syphilis in this population. Like the military STD prevention programs, early drug prevention programs for adolescents and young adults were largely aimed at pro- viding information and evoking fear (Polich et al., 19841. It is difficult to compare the relative effectiveness of these programs because of vari- ability in the content of their messages and the time devoted to their presentation. Nevertheless, general trends in their findings indicate that increased knowledge alone does not accomplish lasting or widespread change in drug-associated behaviors; neither do messages that are solely dependent on high threat content. School-based programs that rely heav- ily on fear have not been successful, apparently because the fear is associated with a low-probability event and because there is a substantial time lag between risk-associated behavior and adverse outcomes (Des Jariais and Friedman, 19871. The assumption of these programs that teenagers will not use drugs if they are simply informed about the in- herent dangers of drug use does not take into account the social factors involved in initiating and sustaining drug use behavior. Ideally, health promotion messages should heighten an individual's perception of threat and his or her capacity to respond to that threat, thus modulating the level of fear. Job (1988) has proposed five prescriptions for the role of fear in health education messages: 1. Messages containing elements of fear should be introduced before discussing the desired behavior. 2. The audience targeted for a message should perceive the negative behavior or event that is associated with the risk as real and likely to occur.

AIDS AND ADOLESCENTS ~ 215 3. A reasonable, desirable alternative behavior that protects the individual against the undesired health problem should be offered. Attention to short-term benefits is desirable and can reinforce long-term behavioral change. 4. The level of fear invoked by the message should be suf- ficient to create awareness of a potential problem but not so high as to evoke denial. Similarly, the fear level should be low enough that it can be effectively managed by the adoption of the desired behavior. 5. The resulting reduction in fear should be of such magnitude that it will reinforce the desired behavior and confirm its effectiveness. What is not yet known is how to introduce fear in the right way in a particular message intended for a particular audience. Acquinng that knowledge will require planned variations of AIDS education programs that are first carefully executed and then carefully evaluated (see Coyle, Boruch, and Turner, 1990) 96 Personal Vulnerability The committee believes that a particularly crucial message for adolescents in the 1990s pertains to their vulnerability. As discussed in Chapter 1, the segments of the American population being touched by this epidemic are changing. Some adolescents, as well as some adults, still view AIDS as a disease of gay white males (Mays and Cochran, 19871. This belief appears to be especially widespread in areas in which the prevalence of AIDS is currently ioW.97 fit iS CIUCia1 that adolescents (especially those 961n its previous report, the committee recommended that AIDS prevention messages strike a balance in the level of threat they convey. The level should be sufficiently high that it motivates individuals to take action but not so high that it paralyzes them with fear or causes them to deny their susceptibility. The committee also recommended that fear-arousing health promotion messages provide specific in- formation on steps to be taken to protect the individual from the threat to his or her well-being. In that vein, the committee notes that a recent community demonstration project in Dallas, Texas, coupled fear-arousing messages concerning HIV transmission with specific instructions on how individuals can reduce their vulnerability. This project developed a series of AIDS prevention posters for young adults that depicted images related to death: a tombstone, a hearse containing a casket, a body on a stretcher covered by a shroud. Each poster included a slogan and specific instructions about methods for minimizing one's risk of HIV infection (Valdiserri, 1989). Rigorous evaluation research is needed, however, to assess the actual effects of such interventions. 97 Similarly, a study of Mexican-American IV drug users from the American Southwest reports that young people in this population who were beginning to inject drugs tended to deny or minimize the threat posed by AIDS (Mate and Jorquez, 1988). Although similar types of denial occur in adults (see Turner, Miller, and Barker, 1989:Table 4), efforts to dispel these misconceptions would be appropriate wherever they occur (as would research to understand the effectiveness of any such efforts).

216 ~ AIDS: THE SECOND DECADE engaging in high-risk activities) recognize the extent of the epidemic and how it might affect them. Facts and Beliefs Although Tables 3-15 and 3-17 indicate that many of the facts about HIV transmission are reaching most high school students, it cannot be safely assumed that providing information will change behaviors.98 Nonetheless, there is an association between some types of knowledge and beliefs and the likelihood that teens will report engaging in behaviors that reduce the risk of HIV transmission. For example, in a 1988 survey of a prob- ability sample of Massachusetts teenagers,99 those teens who believed condoms were effective in preventing AIDS were more likely to report using them; those teens who reported that condom use was embarrassing to discuss were less likely to report using them (Hingson and Strunin, 19891. Such associations do not provide the strongest of evidence, but it would seem reasonable that pilot studies to change knowledge about condom effectiveness and decrease embarrassment about discussing con- dom use would be worth exploring in small-scale interventions. These interventions should be designed in a manner that allows strong conclu- sions to be drawn about the effectiveness of this technique in inducing sexually active teens to increase their use of condoms. In this regard, the committee also wishes to emphasize the need for a long-term commitment to prevention efforts for adolescents. It is highly unlikely that one lecture, one public service announcement, or exposure to a few brochures will be sufficient to induce and sustain the behavioral changes needed to reduce the risk of HIV transmission among adolescents. AIDS prevention programs should take advantage of mul- tiple venues and formats of communication to deliver clear, consistent messages about behaviors that risk HIV transmission and the ways in which those risks may be reduced or eliminated. Furthermore, the com- mittee would point out that the long-term odds of mounting successful 98Empirical examination of the relationships among knowledge, attitudes, beliefs, and behavior rel- evant to HIV/AIDS has provided mixed evidence. A 1987 survey of California high school students (McKusick, Coates, and Babcock, 1988) found that students with higher AIDS knowledge scores also reported higher rates of unsafe sex, but a survey of high school students from the Bronx (Reuben, Hein, and Drucker, 1988) found that those less knowledgeable about AIDS were more likely to participate in high-risk behaviors, regardless of age, grade, grade point average, race, or gender. In looking at the relationship among knowledge, attitudes, and behavior of gay adolescents, Rotheram-Borus and col- leagues (1989) found that attitudes were more highly correlated with condom use than knowledge. A survey of adolescents in detention found a similar relationship (Temoshok et al., 1989). Among college students, perceived susceptibility was associated more strongly with condom use than was knowledge (DiClemente, Forrest, and Mickler, 1989). 99These teens were 16 to 19 years of age (N = 1,762).

AIDS AND ADOLESCENTS ~ 217 prevention programs will be enhanced by-prudent investments in the research required to identify successful (and unsuccessful) strategies for inducing behavioral change. Skills In addition to information, adolescents may require new skills to apply what they have learned to actual situations. Broaching difficult topics in conversations, resisting peer pressures to have unprotected sex or to use drugs, and negotiating less risky activities may be more difficult than learning the facts about transmission routes of HIV. More must be known about providing effective skills training for adolescents and how to combine this effort with the delivery of infor- mation. Selected sex education programs have included skills training, and some programs were able to increase knowledge about sex and contraception and change attitudes while demonstrably improving the participants' ability to use contraceptives (Gilchrist and Schinke, 1983; Schinke, 1984~. The number of school-based sex education programs that have included skills training in their cumculum is small, however, and students generally graduate from these programs with little change in their ability to use contraceptives effectively or to resist pressure to have unprotected sex (Flora and Thoresen, 19881. The available data indicate that, for the most part, drug use is learned and practiced in the context of friendship group s.~°° The social nature of early drug use points to the need for broad-scale interventions that take into account the social network of the adolescent and peer influences that may support the initiation and continuation of drug use. Thus, recent ef- forts to prevent adolescent drug use have gone beyond traditional didactic methods to include strategies that develop social skills and skills to resist peer pressure; other efforts include approaches that correct mispercep- tions of social norms and values (Severson, 1984~. These programs show some promise, but they have not been uniformly successful (Howard et al., 1988; Ellickson and Bell, 19901. Teaching students skills to resist peer pressure and He lure of unhealthy models has had some success in smoking intervention efforts (Battles, 1985),~°~ and assertiveness training for early adolescents has been associated with decreased use of alcohol Iowan exception to this generalization involves alcohol use, which adolescents often report they tried for the first time at home with parents or with other relatives (Severson, 1984). iOIIt is important to note, however, that success rates for smoking programs have not been the same for all subgroups of adolescents, and much of the available research involves studies of adolescents who are enrolled in school. Furthe~.ore, these programs have been more effective among females than among males and more effective among white students than among Asian Americans, native Americans, or black students. The programs also appear to be more effective in preventing smoking

218 ~ AIDS: THE SECOND DECADE and marijuana (Horan and Williams, 19821. Finally, programs that com- bine assertiveness training plus social skills training have reported some success in decreasing rates of cigarette smoking among some young teens and pre-teens (e.g., Botvin and Eng, 19821. Community Norms In its recent report,~02 the committee noted the powerful role that social norms may play in modifying behaviors that risk HIV transmission. This phenomenon follows from the fact that the individual actions that risk HIV transmission (sexual behaviors and sharing of injection equipment) occur in a social environment. Thus, if AIDS prevention programs are to be successful, they must take into account several important aspects of the social factors that affect individual change. For example, people are less likely to behave in ways that will incur the disapproval of others in their social group; people tend to conform to the "shoulds" and "oughts" of behavior specified in the norms of their community. Behavioral change reported by homosexual men has been influenced by changes in the accepted standards and expectations for sexual behavior (i.e., normative shifts) in this group. Identifying the social factors that support behavioral change among adolescent social groups could lead to the development of important components of AIDS prevention programs. Furthermore, programs that seek to change behavior must inevitably confront the diverse and complex social forces that motivate and shape the behaviors at issue. There is increasing evidence that social support has an effect on the health status of individuals. Such support appears to be especially important In the management of chronic disease or in situations in which long-term behavioral change is required to prevent or ameliorate disease (Becker, 19851. For example, family support has been shown to be important in cuing and reinforcing appropriate behaviors for the control of obesity, hypertension, arthritis, and coronary heart disease (Becker, 1985; Morisky et al., 1985~. Families and other groups can affect an individual's adherence to prescribed behaviors by providing material, cognitive, and psychological support. The greater the compatibility of family roles and beliefs, the greater the support for health behaviors and the greater the likelihood that the individual will initiate and sustain them. Furthermore, it is possible that modifying social norms to make risk- associated behavior inconsistent with the prevailing social norms in the among adolescents who have not yet begun to smoke than they are in helping those who already use cigarettes to stop smoking. 102Tu~ner, Miller, and Moses (1989:290-291). The following text is adapted from this source.

AID S AND ADOLESCENTS ~ 2 1 9 adolescent community may be an efficient and effective strategy for AIDS prevention in the long term. In this regard, the committee notes two small surveys of high school students in California. These studies found that the strongest predictor of a teenager's use of condoms during intercourse was the perception that the norms of his or her peers supported safer sex practices and condom use (McKusick, Coates, and Babcock, 1988; Greenblatt et al., 1989~. If norms in the adolescent community could be shifted, the proba- bility that any given individual would indulge n high-risk behavior is likely to be reduced (Turner, Miller, and Moses, 1989:Chapter 4~. If, for example, new norms specified the use of condoms for sexual intercourse, individuals might anticipate that their partners would expect them to behave accordingly. For adolescents who have reported difficulties dis- cussing AIDS prevention measures with their sexual partners (Fisher, 1988), such changes In expectations may relieve some of the pressure of negotiating the use of safer sexual practices dunng difficult situations. It should also be noted that the norms that shape gender and sex roles may have profound influences on sexual behavior. "Macho" gender roles for males, for example, may be at odds with risk reduction (Fisher, 1988), and female gender roles may be inconsistent with the assertive behaviors needed to negotiate sexual behavior or condom use. It may be possible, nonetheless, to design messages Cat would be consistent with both male roles and risk reduction. The committee notes, for example, attempts to disseminate the message that "real men wear condoms." The success of these efforts remains to be measured. Sources of Messages An important pnnciple of human health behavior indicates that people are more likely to act on information if they perceive the source of the message to be credible (Office of Technology Assessment, 1988; Turner, Miller, and Moses, l989:Chapter 41. Therefore, the individuals chosen to deliver AIDS messages should be determined in part by who has credibility with the targeted population. In the case of adolescents, adherence to this principle may mean using peers to present risk reduction . . interventions. Indeed, although one often hears about the barriers to risk reduction posed by peer pressure, peer pressure can have a positive influence on adolescent risk taking. Positive effects of peer influence have been noted in a variety of studies of health problems such as smoking (Evans et al., 1979, Banks, Bewley, and Bland, 1981), and alcohol use (R. lessor, Chase, and Donovan, 19801. Moreover, peer-led intervention

220 ~ AIDS: THE SECOND DECADE programs have shown some promise in smoking prevention programs for adolescents. However, because many smoking prevention programs have also included other important elements for example, skills training to resist peer pressure- it is difficult to isolate the effects of peer leadership (Flay, 1985; McCaul and Glasgow, 1985~. Another potential benefit of a peer-led strategy is that it is quite possible that teenagers will be more comfortable discussing sexual ex- periences with peers rather than with adults. Peer counselors have often been used in sex education programs on the assumption that teens are less embarrassed about discussing sensitive issues among themselves and that they are more likely to follow the advice of other teens (Hayes, 19871. Because these programs have not been evaluated, however, it is not possible to determine whether they have been effective in achieving their goals. Encouraging teen participation as counselors and program leaders may also be beneficial in that it includes adolescents in the development and execution of prevention programs and makes them part of the solution to a problem rather than mere targets of interventions.~03 A program at the University of Illinois at Chicago, for example, seeks to promote adolescent health behavior and to increase teenagers' involvement in the community by training inner-city high school students as peer counselors. Teens have also created audio and visual materials to communicate the facts of HIV transmission and the impact of AIDS on their community. Similarly, in Bethesda, Maryland, specially trained high school student volunteers staff an AIDS hotline for teens, answering questions and making referrals to other sources of information. Hotlines set up and staffed by teens can also be found in Kansas City and Baltimore. In addition, peer counselors have been used on a variety of college campuses (Hein, 1989a). Unfortunately, there is little clear evidence on which to judge the overall effectiveness of such programs although it does appear that at a minimum the peer counselors benefit from participating (Hayes, 1987). Although peer counseling has attractive features, families and other adult social institutions have a major responsibility for educating adoles- cents about health risks. For most adolescents, the family In particular is the social group from which they seek care when they are ill; it is also the source of insurance or money for professional medical help. As part of its information campaign, "Amenca Responds to AIDS," CDC has produced and distributed a brochure to assist parents in discussing 103 Program descriptions were provided by the National AIDS Information Clearinghouse.

AIDS AND ADOLESCENTS ~ 221 HIV and AIDS-related issues with their children.l04 Indeed, more than half (62 percent) of participants in the April-June 1988 wave of the National Health Interview Survey who had children between the ages of 10 and 17 reported discussing AIDS with their children (Hardy, 19891. Yet very little is known about the amount of time spent on this issue or the precise nature of the information delivered to children by parents. Surveys indicate that most parents feel responsible for providing sex education to their children, but few actually offer specific information (Alan Guttmacher Institute, 1981; Roberts, Kline, and Gagnon, 1981~. ~deed, as another National Research Council committee noted, there are many reasons to worry about the effectiveness of parents' efforts at sex education, given the fact that more than 400,000 pregnancies occur each year among teenage girls aged 15 to 17 (Hayes, 1987~. That committee noted: First, in many cases, less parent-child communication takes place than is commonly assumed; second, such communication, whether to provide information or to prescribe behavior, may not be fully heard by the child; and third, communication about sexual behavior frequently does not occur until after initiation of sexual activity (Newcomer and Udry, 1983; Inazu and Fox, 1980~. Fox (1981) points out that parents' (especially mothers') roles in sex education are relatively minor, and that the more traditionally oriented mothers are on matters of sexual morality, sex roles, etc., the less likely they are to initiate discussions of these topics with their children. Unfortunately, however, as Hofferth (1987, Vol. II:Ch. 1) points out, there is little research to specify the context of communication or to distinguish the effects of communication before and after initiation of sexual activity. (Hayes, 1987:103) Such findings may indicate a need for interventions to motivate and assist parents in the difficult and important task of educating their children about the dangers of HIV transmission. These findings also suggest that it would be a mistake to rely exclusively on parents as the source of AIDS education for teenagers. Venues for Program Delivery There are many venues through which AIDS prevention programs for adolescents can be delivered. Some venues (e.g., schools) include most youth, but they miss some of the teens who are at highest risk of becoming infected (e.g., runaways). Other approaches, however, such as use of the media or outreach to special populations, can provide access to 104This element of CDC's AIDS information campaign targets 10- to 20-year-old youths and their parents. It also provides information through workshops, lectures, conferences, and various media events.

222 ~ AIDS: THE SECOND DECADE these groups. In the following sections the committee reviews four types of efforts to reach adolescents with AIDS prevention information: media and other programs designed for all adolescents, school-based programs, programs that target out-of-school youth, and interventions directed toward teenage IV drug users a group at particularly high risk for HIV infection. Programs for All Adolescents Since 1983 CDC has operated a toll-free AIDS hotline (1-800-342-AIDS) whose capacity to deal with requests for information and to answer specific questions has quadrupled since the service began; contractors responsible for the hotline can now handle up to 8,000 calls per day (Mason et al., 1988), and in late 1988 and early 1989 the hotline was responding to approximately 90,000 calls per month (Coyle, Boruch, and Turner, l990:Chapter 31. This service is available to anyone who has access to a telephone, but there are no data on the extent to which adolescents use this program. In its recent report (Coyle, Boruch, and Tumer, 1990:Chapter 3), the committee's Pane! on the Evaluation of AIDS Interventions recommended that a variety of data be collected on users of the national AIDS hotline—including caller age and a small number of other demographic statistics, as well as the services provided to the caller. Such data would allow estimates of the extent of teen use of this program in the future. Limited survey data indicate that the media are an important source of information on AIDS prevention and contraception for teens (PearI, Bouthilet, and Lazar, 1982; Price, Desmond, and Kukulka, l9SS; Ziffer et al., 19891. Adolescents are estimated to spend an average of 17 hours per week watching television (Lawrence et al., 1986), and television thus represents an efficient medium for reaching teens, perhaps before r~sk- associated behaviors begin. Several media-based activities have been initiated for the adolescent population, including AIDS education and public service announcements. CDC has also provided support for the development of AIDS public service announcements for the national adolescent Hispanic population.~05 Research designs to assess the effects of these programs are described elsewhere (Coyle, Boruch, and Turner, 1990). In additior~, less conventional venues have also been used to reach 1°SSee, for example, the description of the CDC-supported efforts of Hispanic Designers, IncoIporated (Organization No. S~7896) and Hispanic AIDS Forum (Organization No. S 04611 ) in the Resource Database, National AIDS Intonation Clearinghouse.

AIDS AND ADOLESCENTS ~ 223 out to teens with information about AIDS.~06 In such cities as Cincinnati, Los Angeles, New York, and Washington, D.C., teenagers participate in AIDS theater groups and rap sessions, performing for peer audiences. Performances are given in English and Spanish in some communities, and in many instances presentations are followed by a question-and-answer period, distribution of educational materials, and occasionally one-on-one conversations with members of the audience. Various sites are used to reach teens with dramatized presentations of AIDS-related information, including public schools, community centers, and even Job Corps sites. Although these activities are imaginative and intuitively appealing, their impact has not yet been ascertained. The committee finds that carefully designed and implemented research efforts are needed to determine which venues and formats are most effective in reaching teens and informing them about the AIDS epidemic in their communities. Printed materials have been developed in several languages for the different age groups that constitute the adolescent population. Read- ing skills may vary considerably in this population, however, and very young teens and adolescents who have dropped out of school may be at a disadvantage when information is provided in written form. Small surveys of adolescents in detention centers (Rolf et al., in press) find that poor reading skills compromise the use of written AIDS prevention information (as well as questionnaires for evaluation activities). Indeed, problems in understanding written AIDS information are not confined to the adolescent population. An analysis of 16 AIDS prevention brochures for the general population found that, on average, they were written at a fourteenth grade (second year of college) reading level (Hochhauser, 19871. Many intervention activities rely at least in part on printed matter to deliver the facts about AIDS, yet very little is known about how the adolescent and young adult populations perceive these materials. A survey of 37 universities found that providing brochures was the extent of the action most universities took regarding AIDS prevention (Caruso and Haig, 19871. With such an emphasis on this avenue of information delivery, it will be necessary to find out more about the opinions and problems of the audience to be reached by such materials, and to incorporate these findings into new editions of the brochures as a way to make them appeal to specific audiences. One comparison of two brochures using a sample of 223 undergraduate university students, for example, found significant gender differences in the evaluation of the materials. Women placed a higher value than did men on information i06 Program descriptions were provided by the National AIDS Information Clearinghouse.

224 ~ AIDS: THE SECOND DECADE about specific safer sex practices and descriptions of strategies that might be used to discuss AIDS and sex with dating partners (D 'Augelli and Kennedy, 19891. Further research including experimental studies differences in the needs of various segments of younger populations will be crucial in developing more effective brochures in the future. School-Based Programs Because schools have the potential to reach 45.5 million students annually (Allensworth and Symons, 1989) and because most adolescents under the age of 18 are enrolled in school,~07 school-based programs are an efficient way to reach a substantial portion of the adolescent population. Schools are long-standing agents of socialization for American youth, and they have played important roles in trying to solve other health problems of adolescents through school-based clinics and health education classes. It is thus logical that schools should become involved in AIDS education and prevention efforts. CDC has developed and funded a multimillion-doliar project for pre- vention programs in schools and other organizations that serve youth.~°8 With input from a wide range of governmental and pnvate-sector organi- zations, the agency has also recommended guidelines for AIDS education (CDC, l98Sd) to help school personnel set the scope and content of their programs. The specifics of such programs, however, are detertnined locally in consultation with the health department, parent groups, and community leaders to ensure that programs are consistent with commu- nity values and needs (Mason et al., 1988~. There are very few descriptions in the scientific literature of precisely how schools are implementing CDC's guidelirles on AIDS program con- tent. As of May 1989, only half (25) of the states required that students receive HIV/AIDS education, and 22 states had no HIV/AIDS education requirements (Voelker, 1989~. For states that have implemented AIDS education programs, process and outcome evaluations are limited, mak- ing it difficult to draw conclusions about the adequacy or effectiveness twin 1984,virtuallyal1 U.S. 14-and 15-year-olds(98percent)and92 percent of lowland 17-year-olds were in school, although by ages 18 and 19 the proportion of students fell to approximately 50 percent (Hayes, 1987). Although there appears to be no difference in male and female enrollment for any age group, Hispanic adolescents are less likely than white or black teens lo be enrolled in school; among 16- to 17-year-olds, 91.2 percent of whites, 92.4 percent of blacks, and 85.7 percent of Hispanics are in school (Hayes, 1987). twin 1987, $1 1 .1 million was allocated to the program; in 1988, this amount increased to $29.9 million (Tolsma, 1988). In 1987, CDC set up cooperative agreements with several organizations that serve schools and youth. Five of the agreements involve organizations that will address the specific needs of black and Hispanic youth, and seven involve organizations that target out-of-school teens (Mason et al., 1988).

AIDS AND ADOLESCENTS ~ 225 of the programs. Workers in the field have suggested that some adoles- cent AIDS prevention programs are not given sufficient classroom time and are not viewed as part of an ongoing educational activity but rather one-time events.~09 The committee cautions that the time and resources currently being devoted to adolescent AIDS prevention efforts may be in- sufficient to effectively block or alter nsk-associated patterns of behavior before they become established. The committee is also concerned that the subset of very young teens who report intercourse and drug use may be initiating these behaviors before they receive information about the hazards associated with them and the means to protect themselves against these hazards. For example, CDC guidelines for AIDS curricula targeting late elementary and middle school youth (i.e., early adolescence) recommend providing the following information on HIV transmission: "The AIDS virus can be transmitted by sexual contact with an infected person; by using needles and other in- jection equipment that an infected person has used; and from an infected mother to her infant before or during birth" (CDC, l98Sd:6~. The com- mittee notes that no information is offered on protective measures that can be taken to reduce the risks of HIV transmission. Although it is true that a majority of young teens will not have initiated risk-associated be- haviors during elementary or middle school (see the preceding sections), a small and particularly vulnerable group will have done so. Indeed, it appears that by the age at which most communities are prepared to accept the notion of providing explicit AIDS education for their youth, a sub- stantial portion of their teens may already be engaging in the behaviors that transmit the virus. If school-based programs are to reach and educate students effec- tively, an informed and supportive group of adults must be developed in local communities to implement these programs.~° The CDC guide- lines (198Sd:3) recommend that "e teem of representatives, including the local school board members, parent-teacher associations, school admin- ~strators, school physicians, school nurses, teachers, educational support personnel, school counselors, and other relevant school personnel should receive general training about: 109An AIDS education program in a medium-sized metropolitan area involved a"minimum of two hours" of instruction by trained health teachers (Copello et al., 1989). A separate educational activity for runaway youth who were admitted into a school system constituted "3 to 4 instructional classes about AIDS and prevention information" (Hudson et al., 1989). 1 1~ ' kin addition to schools, there are a variety of other community-based organizations that serve the adolescent population. Organizations as diverse as the Girls Clubs, Boys Clubs, Girl Scouts, Boy Scouts, Camp Fire, 4-H, the YWCA, and the YMCA report reaching 25 million youths each year (J. Quinn, 1988).

226 ~ AIDS: THE SECOND DECADE · the nature of the AIDS epidemic and means for controlling its spread, · the role of the school in providing education to prevent transmission of HTV, · methods and materials to accomplish effective programs of school health education about AIDS, and · school policies for students and staff who may be infected." Among programs that have been implemented in schools to date, the quality vanes greatly from place to place. Commentators on these programs (e.g., Valdiserri, 1989; Mantell and Schinke, in press) have pointed to the limited amounts of time and resources dedicated to them, the lack of consensus on program content and degree of explicitness of teaching materials, disagreements over the age at which AIDS education should begin, undue emphasis on lectures rather than skills training, and inadequate coordination with community organizations. At present, the effectiveness of school-based programs is unknown, but the committee applauds CDC's efforts to mount systematic evaluations of these activities (Kolbe et al., 1988~. Programs for Out-of-Schoo] Youth Some of the youth who are at highest risk for HIV infection drop out of school before receiving any AIDS education. In 1984, there were more than half a million 14- to 17-year-olds who had dropped out of high school prior to graduation (Hayes, 1987~. Hispanics who were 14 to 15 years old were twice as likely as blacks or whites to have dropped out of school; 13.2 percent of Hispanics between the ages of 16 and 17 had dropped out of school. Indeed, some risk behaviors, such as injecting illicit substances, are likely to cause teenagers to have increasing problems with absenteeism and to hasten their exit from the school system. Obviously, efforts to reach such youth need to go beyond the schools. Community-based organizations have some unique and important char- actenstics that make them particularly promising vehicles for reaching this important segment of the adolescent population. These character- istics include credibility within their community, knowledge of local cultural values and beliefs, and knowledge of local channels of commu- nication. A subset of these organizations are now involved in prevention activities for AIDS, developing and disseminating educational materials, sharing expertise, and establishing referral systems. For example, the National Coalition of Hispanic Health and Human Services Organiza- tions (COSSMHO), with support from CDC and the Metropolitan Life

AIDS AND ADOLESCENTS ~ 227 Insurance Foundation, has begun a major AIDS education initiative for out-of-school Hispanic youths (COSSMHO, 19891. Blacks and Latinos have a tradition of using community-based and other local organizations to address social problems. The organization Hispanos Unidos Contra e] SIDA/AIDS has taken responsibility for train- ing trainers in the Hispanic community and is developing AIDS education curncula for outreach workers and educators who work with Hispanics in churches and teen and community centers in New Haven, Connecticut. For those teens who cannot be reached through the institutions men- tioned above, other mechanisms will be needed to deliver both prevention and social services. Particularly vulnerable teens, some of whom live without families or homes, can often be found on the streets. The pre- cise number of homeless youth is not known, but the dimensions of the adolescent runaway and homeless problem exceed available resources. In one year, for example, 210 programs provided at least one night of shelter to teens on 50,354 occasions, and thousands of teens were turned away from shelters because they were filled to capacity (Bucy, 1985:111. During 1984, Congress appropriated $23.25 million to support 260 shelters for this population run by community-based organizations, as well as other selected services. Some of these monies were used to sup- port a toll-free hotline and communication channel for youth who were either thinking about running away or were already on the street and looking for counseling or referral services. Other support went to family reunification strategies, independent living programs for older homeless teens, suicide prevention, training and employment services, programs to rehabilitate teenage prostitutes, and drug and alcohol counseling. Simi- larly, in September 1987, the National Network of Runaway and Youth Services, Inc., with support from CDC, began a five-year Safe Choices Program, which includes outreach and intervention activities to prevent the spread of HIV infection. Although the effects of such programs have not been ngorously assessed, the recent report of the committee's Panel on the Evaluation of AIDS Interventions describes strategies that can be used to conduct such evaluation (Coyte, Boruch, and Tumer, 1990:Chapter 4~. Reaching Teenage Drug Users Clearly, the most desirable and efficacious goal for AIDS intervention ARIA pilot study of 11- to 18-year-olds contacted on the streets of Newark, New Jersey, between 10 p.m. and 2 a.m. resulted in 27 interviews with 14 males and 13 females (Hummer et al., 1989). Seven did not live with their parents' and four already had children of their own. Half (13 of 27) had dropped out of school, 11 (of 27) reported a history of STDs, and 7 (of 27) said they had a history of drug use.

228 ~ AIDS: THE SECOND DECADE efforts related to drug use is to prevent the initial use of drugs. However, given the questionable success of programs to prevent drug use among teens and the fact that a subset of youth already report drug use, the se- rious public heals threat posed by HIV infection requires that additional strategies be explored. One strategy for decreasing the likelihood that teens will use the "harder" illicit drugs, including injectable dogs, attempts to postpone experimentation with so-called gateway Mugs (alcohol, cigarettes, and marijuana). The use of gateway drugs typically precedes the use of more dangerous substances, and recent studies have demonstrated that it is in fact possible to reduce or postpone initial experimentation with such substances (Botvin et al., 1984; Botvin, 1986; Pentz et al., l989a,b). Suc- cessfu] drug prevention programs, such as the STAR program in Kansas City, have reduced the percentage of adolescents who drink alcohol or smoke mar~juana.~3 The design of this and similar programs uses a "comprehensive" approach, that is, one that goes beyond information delivery on the hazards of drug use and fear arousal to teach the social skills needed to refuse offers of drugs and to find sources of personal satisfaction that do not involve drug use. Such programs often reach beyond the individual adolescent to involve teachers, parents, and the community. Unfortunately, these interventions have not been followed for suf- ficiently Tong periods to know what effects, if any, they will ultimately have on injection practices. Although decreasing the number of teens who use "soft" drugs such as marijuana seems consistent with He notion of fewer individuals injecting drugs, it is possible that even the most suc- cessful programs will have only a small effect in deterring those youths who are at highest risk for progression to injection. Furthermore, because many drug prevention programs work through local social systems, they may miss youths who are at greatest risk for injecting drugs because this MacProject STAR (Students Taught Awareness and Resistance) is a comprehensive, com~nunity-based drug and alcohol abuse prevention program that has been implemented in 15 contiguous communities that together constitute Kansas City. The program is implemented through a school-based curriculum that teaches resistance skills lo middle and junior high school students and includes expanded media and community program elements that target parents, community organizations, and health policy groups (Pentz et al., 1986). 1 13 Data from schools one year after the initiation of Project STAR found that prevalence rates for the use of gateway drugs were significantly lower than rates reported by a delayed intervention cohort: 17 versus 24 percent for cigarette use, 11 versus 16 percent for alcohol use, and 7 versus 10 percent for marijuana use in the previous month (Pentz et al., 1989a). The net increase in drug use prevalence in STAR schools was half that of the delayed intervention schools.

AIDS AND ADOLESCENTS ~ 229 highest risk segment of the adolescent population is not well integrated into youth-oriented institutions (e.g., schools). To conduct the research needed to determine whether a program that reduces experimentation with gateway drugs during early or middle adolescence will also reduce the number of late adolescents and young adults who inject drugs would require a very complex research design and considerable resources. It would also require following participants (who may not wish to stay in contact with any authorities) for a long period of time. Furthermore, a range of factors would need to be measured to determine the effects of any particular type of program on the rate of drug injection. In the absence of such research, it is probably safe to say that it is overly pessimistic to assume that there is no effect of such . . . . . . . . . programs on t le progression to injection, anc it IS over y optimistic to assume that such programs will be sufficient to reduce the likelihood of injection to zero. One extreme of the range of strategies to prevent injection by reduc- ing experimentation with gateway drugs in the broad teenage population comprises programs that focus attention on youths who are at greatest risk for the use of injectable drugs. A New York City research project (Des Jariais et al., 1989) has attempted to determine the impact of inter- vening during latter stages of drug use progression. The study recruited persons who were using heroin intranasally (so-called sniffing) and paid them to participate in a four-session AIDS intervention program. One- half of the subjects were randomly assigned to an intervention that taught skills needed to refuse offers of injectable drugs, negotiate safer sex, and negotiate entry into drug abuse treatment. The other half were assigned to a control group that did not receive the four-session intervention but was otherwise treated identically. At the end of the follow-up period, the rate of injection among subjects who had received the intervention was half that reported by the control group. (Approximately 15 percent of those who received the intervention injected an illicit drug during the n~ne-month per~od.~il4 There remains, however, the question of whether less seasoned, less practiced youths would respond as favorably to such an intervention effort. TV drug use prevention among adolescents must also take into ac- count the diverse social factors that may affect needle use. One study from Baltimore (see VIahov et al., 1989), for example, found evidence of less risk taking among new injectors. Yet studies conducted in the New York area present a more complex picture. New injectors, who were not 1 l4Don C. Des Jarlais, Chemical Dependency Institute, Beth Israel Medical Center, personal commu- nication, May 9, 1990.

230 ~ AIDS: THE SECOND DECADE fully integrated into the drug subculture, were less likely to engage in very high risk behavior, such as the use of shooting galleries (Friedman et al., 19891. Unfortunately, they were also less likely to practice deliberate AIDS risk reduction (Kleinman et al., in press). New injectors may also be influenced by older, more practiced users. Studies of heroin use among Mexican-Amer~can adolescents found that punitive attitudes toward these youths resulted in their arrest and incar- ceration in jails, forestry camps, and prisons (Mate and Jorquez, 19881. This policy led to the unfortunate and unintended consequence of putting these youth in direct contact with older and more practiced users of IV drugs. One outcome of this experience was the development of a common perspective and code of conduct among the youths and older addicts, which in turn fostered norms that supported the sharing of drugs, injection equipment, and even sexual partners. Intervening to prevent the spread of HIV infection among drug-using adolescents presents considerable challenges. Some have hoped that the fear of contracting AIDS would serve as a "natural" intervention tool, preventing would-be injectors from progressing to more hazardous pat- terns of behavior. (The hope for such a spontaneous event underlies much of the opposition to such controversial AIDS prevention programs as sy- nnge exchanges, which are misperceived as encouraging Mug injection.) It has also been hypothesized that the fear of a deadly disease might lead some people who were predisposed to intensive Mug use to choose routes of administration other than injection. Heroin, cocaine, and am- phetamines, the most commonly injected drugs in this country, can all be "smoked" to provide a drug effect that approximates the effect achieved by injecting.ll5 Indeed, there has been a dramatic increase in the smoking of cocaine (as "crack") since the beginning of the AIDS epidemic, but there does not appear to be a concomitant decrease in the amount of cocaine that is injected (Des lariats and Friedman, 19881. Rather than substituting for cocaine injection, smoking cocaine appears to occur in addition to existing drug use; frequently, it is also associated with in- creased sexual risk taking (see Chapter 1~. Moreover, interviews with heroin users who report intrallasal administration of this drug indicate that few cite concern about AIDS as affecting their drug use. Teens who have already begun to inject drugs and who resist the notion of treatment or cessation of use require other strategies to protect ~ is Users report that snorting or intranasal administration of heroin and cocaine produces a less intense effect than intravenous administration. 116A small study of 102 persons who were using heroin intranasally found that only one person was sufficiently concerned about AIDS lo report that it affected his drug use (Des Jarlais et al., 1989).

AIDS AND ADOLESCENTS ~ 231 them from HIV infection. Strategies to deliver information and teach skills related to safer injection practices have been developed for adults; modifications of these approaches may be useful for adolescents. For example, older peers (e.g., more experienced injectors who no longer use drugs) may be credible sources for information about drug use. Street outreach efforts in New York, Chicago, and San Francisco have trained ax-addicts to reach out to current users to disseminate information on the risk reduction options available to injectors: treatment for those who wish it and bleach sterilization techniques for those who are unable to accept or locate available treatment, and condom use to prevent sexually transmitted HTV infection (Des Jarlais et al., 19881. Unfortunately, many drug treatment programs are not well adapted to help adolescent users, and some are not even open to individuals younger than IS years of age (Polich et al., 1984~. Preventing AIDS by preventing the initiation of injection of illicit drugs is an appropriate public health goal whose achievement would not only reduce the transmission of HIV infection but would also reduce many of the other significant health and social problems associated with injection. The level and extent of prevention needed, however, could consume extensive resources. Implementing effective interventions is likely to require a combination of expensive programs: programs that focus on the individual and on the psychological and social problems that make individuals likely to inject illicit Hugs, and programs that address social problems that have supported risky behaviors and prevented the establishment of effective intervention strategies. Program Needs for High-Risk and HIV-Seropositive Youth The range of services that may be helpful to youth who participate in risk- associated behaviors is extensive. Such services include family planning clinics, drug use and prevention programs, teenage pregnancy programs, STD clinics, and comprehensive service programs that target a variety of social and physical problems of adolescents. The current demands on these programs, most of which involve non-HIV-related problems, are considerable; the feasibility of adding responsibility for HIV-related prevention strategies, testing, counseling, and follow-up services for at- risk teens, their partners, and families is not known. As the epidemic continues, there may well be more infected teens who will require a wide range of services for their care. Three basic strategies can be used to address the needs of infected adolescents: (1) expand existing services for other adolescent health problems to include

232 ~ AIDS: THE SECOND DECADE HIV infection and AIDS, (2) expand adult HIV and AIDS programs to include adolescents, or (3) develop new programs specifically designed to treat HIV-infected teens. As an example of the first strategy, regional hemophilia centers have expanded their services to include HIV- and AIDS-related diagnosis, treatment, and care, and a few hospital-based clinics have developed the capacity to evaluate and care for infected adolescents. In the New York City area, examples of all three strategies have appeared over the past several years. For example, the Adolescent Health Service at Mt. Sinai Hospital has integrated HIV services into preexisting comprehensive health care clinics for teenagers. Similarly, the Adolescent AIDS Program at Montefiore Medical Center comprises a multidisciplinary team of professionals who have established protocols for counseling and testing teens, evaluating and canny for infected ado- lescents, and conducting research on the epidemiology of HIV infection and behavioral prevention strategies in this population. Preexisting models for providing multiple services for adolescents include such comprehensive service programs as E1 Puente and The Door in New York City, programs for Hispanic youth and their families (Hen- ncks, 1985; Paget, 19881. These programs have a tradition of dealing with a spectrum of adolescent problems and offer a wide range of ser- vices: family plating and sex counseling, medical services, psychiatric counseling, drug and alcohol treatment, workshops, and other social ser- vices. In addition, E] Puente pairs successful members of the community with teens to provide role models and a bridge to more successful and empowered lifestyles (Pages, 19881. The resources required to provide necessary services are likely to be extensive, but the precise nature and quantity of these resources have not been determined. It is likely, however, that staffing and training needs for prevention, as well as care for at-nsk and infected teens, will be substantial. DOING BETTER IN THE SECOND DECADE In considering, how to improve future efforts to retard the spread of HIV among teenagers and young adults, the committee returns to several issues that are discussed in the opening pages of this chapter. As previously noted, the diversity of opinion in this country concerning matters of sexual behavior has contributed to a lack of consensus on appropriate educational messages. This lack of consensus, in tum, presents substantial baITiers to AIDS prevention activities. The committee recognizes the difficulties posed for some individuals by frank discussions of sensitive issues such as sexual behavior, contraceptive use, and prevention of STDs. It believes,

AIDS AND ADOLESCENTS ~ 233 however, that clear, appropnate, and effective educational programs can and must be crafted and delivered. Such a task is likely to be discomfiting and controversial. Yet for the sake of the nation's 35 million teens,~7 national and community leaders must shoulder the responsibility for such education and be held accountable for any failure to provide it. As former Surgeon General C. Everett Koop noted, "many young people in the United States are not receiving information that is vital to their future health because of our reticence in dealing with the subjects of sex, sexual practices, and homosexuality. As parents, educators, and community leaders we must assume our responsibility to educate our young. The need is critical and the price of neglect is high." In the task of AIDS education, the committee believes that prevention efforts should involve adolescents themselves in the design and execution of these programs. This policy not only allows the program to benefit from the counsel of members of the targeted audience but it empowers adolescents by including them in the processes that will affect their lives. For many adolescents, inclusion in such activities will provide another stimulus for thinking about AIDS and planning for their futures. It is clear, nonetheless, that it is not yet known how best to educate adolescents about the behavioral changes required to retard the spread of HIV. The committee believes that much is to be gained from the systematic study of planned variations of intervention strategies, and it regrets the persisting lack of understanding regarding the types of behav- ioral interventions that will be most effective in containing the spread of this disease. The committee believes that the Public Health Service would realize substantial returns in practical and scientific knowledge from careful investments in research to determine the effects of planned variations of those behavioral interventions to be implemented in the fu- ture. This strategy offers hope that, if the epidemic enters a third decade, our understanding of how to curb the spread of HIV infection will be far greater than it is today. Finally, the committee wishes to reiterate that the diversity of the adolescent population means that a multiplicity of venues and formats will be needed to deliver AIDS prevention messages, and a variety of strategies will be required to intervene effectively. For some adolescents, effective intervention may be provided by AIDS prevention programs in 1l7The U.S. Bureau of the Census estimates that in 1980 there were 37,174,000 individuals between the ages of 10 and 19 years residing in the United States (U.S. Bureau of the Census, 1987:17). ll8C. Everett Koop, public statement at press conference on release of Surgeon General's Report on AIDS, Washington, D.C., October 22, 1986.

234 ~ AIDS: THE SECOND DECADE the school systems at an early age—before students initiate the behaviors that can threaten their future. Yet reaching Marty of the adolescents at highest risk for HIV infection will require going beyond the schools. Most important, the committee notes that there is a small segment of the teen population that at this moment is at relatively high risk for HIV infection.~9 This segment includes runaway and "throwaway" children, teens who exchange sex for survival needs, and out-of-home and homeless youths. The committee believes that AIDS prevention programs should focus attention on these youths in a manner commensurate with the elevated risks they face. In addition, although programs to prevent HIV transmission can and should be deployed for this segment of the teen population, the needs of these high-risk youth extend far beyond the scope of AIDS prevention and include shelter, medical care, education, and a reason to believe that the future will be better than the present. The committee believes that efforts to prevent HIV transmission should be complemented by interventions that seek to satisfy the broader needs of these youth. At a time when the AIDS virus has been introduced into the adoles- cent population but is not evenly distributed throughout it, the near-term future presents unique and important opportunities to prevent the acquisi- tion of infection. As stated in the first chapter of this report, the greatest opportunities for getting ahead of the AIDS epidemic lie in geographic areas or populations that currently have a low prevalence of AIDS and HIV infection. These opportunities must not be overlooked or squan- dered, for once lost, they are gone forever. The future course of the epidemic will be determined by actions that are taken now. If the youth of this nation are taught how to engage in healthy behaviors and protect themselves against HIV infection, these efforts may not only prevent further spread of He AIDS epidemic but may also prevent future health problems in their adult years. REFERENCES Alan Guttmacher Institute. (1981) Teenage Pregnancy: The Problem That Hasn't Gorze Away. New York: Alan Guttmacher Institute. Alexander-Rodriguez, T., and VeITnund, S. H. (1987) Gonorrhea and syphilis in incarcerated urban adolescents: Prevalence and physical signs. Pediatrics 80:561- 564. Allensworth, D. D., and Symons, C. W. (1989) A theoretical approach to school-based HIV prevention. Journal of School Health 59:59~5. 1l9It is important to remember, however, that relative risk is a function of the prevalence of the virus and the behaviors that transmit it. Thus, as the epidemic evolves, so the relative risk for adolescents will change.

AIDS AND ADOLESCENTS | 235 American Medical Association Council on Scientific Affairs (AMA-GSA). (1990) Health status of detained and incarcerated youth. Journal of the American Medical Association 263:987-991. American School Health Association, Association for the Advancement of Health Education, and the Society for Public Health Education, Inc. (ASHA, AAHE, and SPHE). (1989) The National Adolescent Student Health Survey. Oakland, Calif.: Third Party Publishing Company. Austin, G. A., and Prendergast, M. L., eds. (1984) I9 rug Use and Drug Abuse: A Guide to Research Findings. Vol. 2: Adolescents. Denver: ABC-Clio Information Services. Bachman, J. G., and O'Malley, P. M. (1981) When four months equal a year: Inconsistencies in student reports of drug use. Public Opinion Quarterly 45:53 548. Bachman, J. G., Johnston, L. D., and O'Malley, P. M. (1981) Smoking, drinking and drug use among American high school students: Correlates and trends, 1975-1979. American Journal of Public Health 71:59~9. Bailey, S. L., and Hubbard, R. L. (1990) Developmental variation in the context of marijuana initiation among adolescents. Journal of Health and Social Behavior 31:58-70. Baker, C. J., Huscroft, S., Morris, R., Re, O., Roseman, J., and Shultz, B. (1989) HIV seroprevalence and behavior survey of incarcerated adolescents. Unpublished paper. Los Angeles County Department of Health Services, Juvenile Court Health Services. Banks, M. H., Bewley, B. R., and Bland, J. M. (1981) Adolescent attitudes to smoking Their influence on behavior. International Journal of Health Education 24:39~4. Battjes, R. J. (1985) Prevention of adolescent drug abuse. The International Journal of the Addictions 20:1113-1134. Becker, M. H. (1985) Patient adherence to prescribed therapies. Medical Care 23:539- 555. Bell, T. A., and Hein, K. (1984) Adolescents and sexually transmitted diseases. In K. Holmes, P. Mardh, P. F. Sparling, and P. J. Wiesner, eds., Sexually Transmitted Diseases. New York: McGraw Hill. Billy, J. O. G., and Udry, J. R. (1985) The influence of male and female best friends on adolescent sexual behavior. Adolescence 20:21-32. Billy, J. O. G., Rodgers, J. L., and Udry, J. R. (1984) Adolescent sexual behavior and friendship choice. Social Forces 62:653~78. Blum, R. W. and Resnick, J. F. (1982) Adolescent decision making: Contraception, pregnancy, abortion and motherhood. Pediatric Annals 11:797-805. Botvin, G. J. (1986) Substance abuse prevention research: Recent developments and future directions. Journal of School Health 56:369-374. Botvin, G. J., and Eng, A. (1982) The efficacy of a multicomponent approach to the prevention of cigarette smoking. Preventive Medicine 11:199-211. Botvin, G. J., Baker, E., Renick, N. L., Filazzola, A. D., and Botvin, E. M. (1984) A cognitive-behavioral approach to substance abuse prevention. Addictive Behaviors 9:137-147. Brandt, A. M. (1987) lVo Magic Bullet. New York: Oxford University Press.

236 ~ AIDS: THE SECOND DECADE Brooks-Gunn, J., and Furstenberg, F. F., Jr. (1989) Adolescent sexual behavior. American Psychologist 44:249-257. Brooks-Gunn, J., Boyer, C., and Hein, K. (1988) Preventing HIV infection and AIDS in children and adolescents: Behavioral research and intervention strategies American Psychologist 43:958-964. Brundage, J. F., Burke, D. S., Gardner, L. I., Visintine, R., Peterson, M., and Redfield, R. R. (1988) HIV infection among young adults in the New York City area: Prevalence and incidence estimates based on antibody screening among civilian applicants for military service. New York State Journal of Medicine 88:232-235. Brunswick, A. F. (1969) Health needs of adolescents: How the adolescent sees them. American Journal of Public Health 59:173~1745. Brunswick, A. F. (1988) Young black males and substance use. In J. T. Gibbs, ea., Young, Black, and Male in America: An Endangered Species. Dover, Mass.: Auburn House. Brunswick, A. F., Merzel, C. R., and Messen, P. A. (1985) Drug use initiation among urban black youth: A seven-year follow-up of developmental and secular influences. Youth and Society 17: 189-216. Bucy, J. (1985) To Whom Do They Belong? A Profile of America's Runaway and Homeless Youth and the Programs That Help Them. Washington, D.C.: The National Network of Runaway and Youth Services, Inc. Buning, E., Coutinho, R. A., and van Brussell, G. H. A. (1986) Preventing AIDS in drug addicts in Amsterdam. Lancet 1:1435-1436. Buning, E., Hartgers, C., Verster, A. D., van Santen, G. W., and Coutinho, R. A. (1988) The evaluation of the needle/syringe exchange in Amsterdam. Presented at the Fourth International Conference on AIDS, Stockholm, June 12-16. Burke, D. S., Brundage, J. F., Herbold, J. R., Berner, W., Gardner, L. I., et al. (1987) Human immunodeficiency virus infections among civilian applicants for United States military service, October 1985 to March 1986. New England Journal of Medicine 317:131-136. Burke, D. S., Brundage, J. F., Goldenbaum, M., Gardner, L. I., Peterson, M., et al. (1990) Human immunodeficiency virus infections in teenagers: Seroprevalence among applicants for U.S. military service. Journal of the American Medical Association 263:207~2077. Campbell, B. (1984) National Longitudinal Survey of Labor Force Behavior: Technical Report on Interviewing Submitted to the Center for Human Resource Research. National Opinion Research Center, Chicago. Caruso, B. A., and [Iaig, J. R. (1987) AIDS on campus: A survey of college health service priorities and policies. Journal of American College Health 36:32-36. Catania, J. A., Kegeles, S. M., and Coates, T. J. (1988) Towards an understanding of risk behavior: The CAPS' AIDS risk reduction model (AARM). Center for AIDS Prevention Studies, University of California at San Prancisco, January. Centers for Disease Control (CDC). (1986) HTLV III/LAY. Antibody prevalence in U.S. military recruit applicants Morbidity and Mortality Weekly Report 35:421~24. Centers for Disease Control (CDC). (1987) Sexually Transmitted Disease (STD) Statis- tiCS: 1985. Atlanta, Gal: Centers for Disease Control. Centers for Disease Control (CDC). (1988a) HIV-related beliefs, knowledge, and behaviors among high school students. Morbidity and Mortality Weekly Report 37:717-721.

AIDS AND ADOLESCENTS ~ 237 Centers for Disease Control (CDC). (1988b) AIDS Weekly Surveillance ReportMJnited States. ALDS Program, Center for Infectious Diseases, Centers for Disease Control, Atlanta, Gal, January 18. Centers for Disease Control (CDC). (1988c) Trends in human immunodeficiency virus infection among civilian applicants for military service United States, October 1985 March 1988. Morbidity and Mortality Weekly Report37:677~79. Centers for Disease Control (CDC). (1988d) Guidelines for effective school health education to prevent the spread of AIDS. Morbidity and Mortality Weekly Report 37(Suppl. Six: 1-14. Centers for Disease Control (CDC). (1988e) Sexually Transmitted Disease Statistics: 1987. Atlanta, Gal: Centers for Disease Control. Centers for Disease Control (CDC). (1989) HIVIAIDS Surveillance: AIDS Cases Reported Through September 1989. Atlanta, Gal: Centers for Disease Control. Chiasson, R. E., Osmond, D., Moss, A., Feldman, H., and Bernacki, P. (1987) HIV, bleach and needle sharing (letter). Lancet 1:1430. Clayton, R. R., and Ritter, C. (198S) The epidemiology of alcohol and drug abuse among adolescents. In B. Stimmel, ea., Alcohol and Substance Abuse in Adolescence. New York: Haworth Press. Connell, D. B., Turner, R. R., and Mason, E. F. (1985' Summary of findings of school health education evaluation: Health promotion effectiveness, implementation and costs. Journal of School Health 55:317-321. Consumer Reports. (1989) Can you rely on condoms? Consumer Reports March:135- 142. Copello, A. G., Sheets, R., Ross, S., and Curvin, M. (1989) Evaluation of a targeted AIDS/HIV education program for secondary school students in a medium-size USA metropolitan area. Presented at the Fifth International Conference on AIDS, Montreal, June =9, 1989. COSSMHO (The National Coalition of Hispanic Health and Human Services Organi- zations). (1989) AIDS education for youth enters second year. The COSSMHO AIDS Update 3:1-2, May-June. Coyle, S. L., Boruch, R. F., and Turner, C. F., eds. (1990) Evaluating AIDS Prevention Programs, Expanded Edition. Washington, D.C.: National Academy Press. D'Angelo, L., Getson, P., Luban, N., Stallings, E., and Gayle, H. (1989) HIV infection in adolescents: Can we predict who is at risk. Presented at the Fifth International Conference on AIDS, Montreal, June =9. Darrow, W. W. (1989) Condom use and use-effectiveness in high-risk populations. Sexually Transmitted Diseases 16:157-160. O'Augelli, A. R., and Kennedy, S. P. (1989) An evaluation of AIDS prevention brochures for university women and men. AIDS Education and Prevention 1:13~140. Davies, M., and Kandel, D. B. (1981) Parental and peer influence on adolescents' educational plans: Some further evidence. American Journal of Sociology 87:363- 387. Davis, J. A., and Smith, T. W. (1989) General Social Surveys, 1972-1989: Cumulative Codebook. Chicago: National Opinion Research Center, University of Chicago. Des Jarlais, D. C., and Friedman, S. R. (1987) HIV infection among intravenous drug users: Epidemiology and risk reduction (editorial review). AIDS 1:67-76.

238 ~ AIDS: THE SECOND DECADE Des Jarlais, D. C., arid Friedman, S. R. (1988) Intravenous cocaine, crack, and HIV infection (letter). Journal of the American Medical Association 259:1945-1946. Des Jarlais, D. C., Friedman, S. R., Sotheran, J. L., and Stoneburner, R. (1988) The sharing of drug injection equipment and the AIDS epidemic in New York City: The first decade. In R. J. Battjes, and R. W. Pickens, eds., Needle Sharing Among Intravenous Drug Abusers: National and International Perspectives. National Institute for Drug Abuse Research Monograph 80. Washington, D.C.: U.S. Government Printing Office. Des Jarlais, D. C., Casriel, C., Friedman, S. R., Rosenblum, A, Rodriguez, R., et al. (1989) AIDS education and the transition from non-injecting to injecting drug use. Presented at the Fifth International Conference on AIDS, Montreal, June =9. DiClemente, R. J. (1989) Prevention of human immunodeficiency virus infection among adolescents: The interplay of health education and public policy in the development and implementation of school-based AIDS education programs. AIDS Education and Prevention 1:7(~78. DiClemente, R. J., and DuNah, R. (1989) A comparative analysis of risk behaviors among a school-based and juvenile detention facility sample of adolescents in San Francisco. Presented at the Fifth International Conference on AIDS, Montreal, June =9. DiClemente, R. J., Boyer, C. B., and Morales, E. S. (1988) Minorities and AIDS: Knowledge, attitudes and misconceptions among black and Latino adolescents. American Journal of Public Health 78:55-57. DiClemente' R., Forrest, K., and Mickler, S. (1989) Differential effects of AIDS knowl- edge and perceived susceptibility on the reduction of high-risk sexual behaviors among college adolescents. Presented at the Fifth International Conference on AIDS, Montreal, June 09. DiClemente, R. J., Zone, J., and Temoshok, L. (1986) Adolescents and AIDS: A survey of knowledge, attitudes and beliefs about AIDS in San Francisco. American Journal of Public lIealth 76:1443-1445. DiClemente, R. J., Zorn, J., and Temoshok, L. (1987) The association of gender, ethnicity, and length of residence in the Bay area to adolescents' knowledge and attitudes about acquired immune deficiency syndrome. Journal of Applied Social Psychology 17:21~230. Duncan, M. E., Tibaux, G., Pelzer, A., Reimann, K, Pentherer, J. F., et al. (1990) First coitus before menarche and risk of sexually transmitted disease. Lancer 335:338-340. Dusher, R. W., and Mills, C. A. (1963) The adolescent looks at his health and medical care. American Journal of Public Health 53:1928-1936. Elkind, D. (1985) Cognitive development and adolescent disabilities. Journal of Adolescent Health Care 6:8089. Ellickson, P. L., and Bell, R. M. (1990) Drug prevention in junior high: A multi-site longitudinal test. Science 247:1299-1305. Elliott, D. S., and Morse, B. J. (1989) Delinquency and drug 1lse as risk factors in teenage sexual activity. Youth and Society 21:32~0. Ensminger, M. E. (1987) Adolescent sexual behavior as it relates to other transition behaviors in youth. In S. L. Hofferth and C. D. Hayes, eds., Risking the Future: Adolescent Sexuality, Pregnancy, and Childbearing. Vol. 2, Working Papers and Statistical Appendixes. Washington, D.C.: National Academy Press.

AIDS AND ADOLESCENTS ~ 239 Ensminger, M. E., and Kane, L. P. (1985) Adolescent drug and alcohol use, delinquency and sexual activity: Patterns of occurrence and risk factors. Presented at the National Institute for Drug Abuse Technical Review on Drug Abuse and Adolescent Sexual Activity, Pregnancy and Parenthood, March. Ernst, J. A., Bauer, S., Amaral, L., St. Louis, M., and Falco, I. (1989) HIV seropreva- lence at the Bronx Lebanon Hospital Center, a CDC sentinel hospital. Presented at the Fifth International Conference on AIDS, Montreal, June =9. Evans, R. I., Henderson, A., Hill, P., and Raines, B. (1979) Smoking in children and adolescents: Psychosocial determinants and prevention strategies. In N. A. Krasnegor, ea., The Behavioral Aspects of Smoking. National Institute on Drug Abuse Research Monograph No. 26. Washington, D.C.: U.S. Government Printing Office. Faltz, B., and Madoover, S. (1987) Substance abuse as a co-factor for AIDS. In Womerz and AIDS Clinical Resource Guide, 2nd ed. San Francisco: San Francisco AIDS Foundation. Pay, R. E., Tumer, C. F., Klassen, A. D., and Gagnon, J. H. (1989) Prevalence and patterns of same-gender sexual contact among men. Science 243:338-348. Feldblum, P. J., and Fortney, J. A. (1988) Condoms, spennicides, and the transmission of human immunodeficiency virus: A review of the literature. American Journal of Public Health 78:52-54. Fisher, J. D. (1988) Possible effects of reference group-based social influence on AIDS-nsk behavior and AIDS prevention. American Psychologist 43:914-920. Flavin, D. K., and Prances, R. J. (1987) Risk taking behavior, substance abuse disorders, and the acquired immunodeficiency syndrome. Advances in Alcohol and Substance Abuse 6:23-31. Flay, B. R. (1985) Psychosocial approaches to smoking prevention: A review of findings. Health Psychology 4:449~88 Flora, J. A., and Thoresen, C. E. (1988) Reducing the risk of AIDS in adolescence. American Psychologist 43:965-970. Fox, G. L. (1981) The family's role in adolescent sexual behavior. In T. Ooms, ea., Teenage Pregnancy in a Family Context. Philadelphia: Temple University Press. Freudenberg, N., Lee' J., and Silver, D. (1989) How black and Latino community organizations respond to the AIDS epidemic: A case study in one New York City neighborhood. AIDS Education and Prevention 1:12-21. Friedman, H. L. (1989) The health of adolescents: Beliefs and behavior. Social Science and Medicine 29:309-315. Friedman, I. M., and Litt, I. F. (1987) Adolescents' compliance with therapeutic regimens: Psychological and social aspects and intervention. Journal of Adolescent Health Care 8:52~7. Friedman, S. R., Des Jarlais, D. C., Neaigus, A., Abdul-Quader, A., Sotheran, J. L., et al. (1989) AIDS and the new drug injector. Nature 339:333-334. Fullilove, R. E., Fullilove, M. T., Bowser, B. P., and Gross, S. A. (1989) Crack use and risk for AIDS among black adolescents. Presented at the Fifth International Conference on AIDS, Montreal, June =9. Fullilove, R. E., Fullilove, M. T., Bowser, B. P., Gross, S. A. (1990) Risk of sexually transmitted disease among black adolescent crack users in Oakland and San Francisco, Calif. Journal of the American Medical Association 263:851-855.

240 ; AIDS: THE SECOND DECADE Furstenberg, F. F., Morgan, S. P., Moore, K. A., and Peterson, J. L. (1987) Race differences in the timing of adolescent intercourse. American Sociological Review 52:511-518. Gayle, H., Manoff; S., and Rogers, M. (1989) Epidemiology of AIDS in adolescents, U.S.A. Presented at the Fifth International Conference on AIMS. Montreal, June =9. Gayle, H., Rogers, M., Manoff, S., and Starcher, E. (1988) Demographic and sexual transmission differences between adolescent and adult AIDS patients, U.S.A. Presented at the Fourth International Conference on AIDS, Stockholm, June 12-16. Gayle, H., Keeling, R., Garcia-Tunon, M., Kilbourne, B., and Narkunas, J. (1989) HIV seroprevalence on university campuses, U.S.A. Presented at the Fifth International Conference on AIDS, Montreal' June =9. Gellert, G. A., and Durfee, M. J. (1990) HIV infection and child abuse. New England Journal of Medicine 321:685. Gerstein, D. R., Judd, L. L., And Rovner, S. A. (1979) Career dynamics of female heroin addicts. American Journal of Drug and Alcohol Abuse 6:1-23. Gilchrist, L. D., and Schinke, S. P. (1983) Coping with contraception: Cognitive and behavioral methods with adolescents. Cognitive Therapy and Research 7:379-388. Ginzburg, H. M. (1988) Acquired immune deficiency syndrome (AIDS) and drug abuse. In R. P. Galea? B. F. Lewis, and L. A. Baker, eds., AIDS and IV Drug Abusers: A Current Perspective. Owings Mills, Md.: National Health Publishing. Girodo, M., Dotzenroth, S., and Stein, S. (1981) Causal attribution bias in shy males: Implications for self-esteem and self-confidence. Cognitive Therapy and Research 5:325-338. Goedert, J. J., Kessler, C. M., Aledort, L. M., Biggar, R. J., Andes, W. A., et al. (1989) A prospective study of human immunodeficiency virus type 1 infection and the development of AIDS in subjects with hemophilia. New England Journal of Medicine 32 1:114 1-1148. Gold, M., and Petronio, R. J. (1980) Delinquent behavior in adolescence. In J. Adelson, ea., Handbook of Adolescent Psychology. New York: John Wiley and Sons. Greenblatt, R. M., Catania, J. A., Kegeles, S. M., Schachter, Y., Miller, J., and Coates, T. J. (1989) Predictors of condom use and STDs in a group of sexually active adolescent women. Presented at the Fifth International Conference on AIDS, Montreal, June ~9 Haignere, C., Rotheram-Borus, M., Bradley, J., Koopman, C., and Harden, N. (1989) Stressful life events and social supports as mediators of safe behaviors among runaway and gay youths. Presented at the Fifth International Conference on AIDS, Montreal, June =9. Hamburg, B. A., Kraemer, H. C., and Jahnke, W. (1975) A hierarchy of drug use in adolescence: Behavioral and attitudinal correlates of substantial drug use. American Journal of Psychiatry 132: 1 155-1 163. Hamburg, D. (1986) Preparing for life: The critical transition of adolescence. In The 1986 Annual Report of the Carnegie Corporation of New York. New York: The Carnegie Corporation. Hardy, A. M. (1989) AIDS knowledge and attitudes for Apnl-June 1989: Provisional data from the National Health Interview Survey. NCHS Advance Data 179:1-12.

AIDS AND ADOLESCENTS ~ 241 Hartgers, C., Buning, E. C., van Santen, G. W., Verster, A. D., and Coutinho, R. A. (1989) The impact of the needle exchange programme in Amsterdam on injecting risk behavior. AIDS 3:571-577. Haverkos, H. W., Bukoski, W. J., and Amsel, Z. (1989) The initiation of male homosexual behavior. Journal of the American Medical Association 262:501. Hayes, C. D., ed. (1987) Risking the Future: Adolescent Sexuality, Pregnancy, and Childbearing. Vol. 1. Washington, D.C.: National Academy Press. Hein, K. (1989a) AIDS in adolescence: Exploring the challenge. Journal of Adolescent Health Care lO(Number 3 Supplement):lOS-35S. Hein, K. (1989b) Commentary on adolescent acquired immunodeficiency syndrome: The next wave of the human immunodeficiency virus epidemic? Journal of Pediatrics 114:14~149. Hein, K., and DiGeronimo, T. F. (1989) AIDS: Trading Fears for Facts. Mount Vernon, N.Y.: Consumers Union. Hein, K., Marks, A., and Cohen, M. (1977) Asyrnptomatic gonorrhea: Prevalence in a population of urban adolescents. Journal of Pediatrics 90:634~35. Hein, K., Cohen, M. I., Marks, A., Schonberg, S. K., Meyer, M., and McBride, A. (1978) Age at first intercourse among homeless adolescent females. Journal of Pediatrics 93:147-148. Henricks, L. E. (1985) Establishment of accessible and relevant services for adolescents. In D. Shaffer, A. A. Eh~hardt, and L. L. Greenhill, eds., The Clinical Guide to Child Psychiatry. New York: The Free Press. Hingson, R., and Strunin, L. (1989) Do health belief model beliefs about HIV infection and condoms predict adolescent condom use? Presented at the Fifth International Conference on AIDS, Montreal, June =9. Hingson, R., Strunin, L., and Berlin, El. (1990) Acquired immunodeficiency syn- drome transmission: Changes in knowledge and behaviors among teenagers, Massachusetts statewide surveys, 1986 to 1988. Pediatrics 85:2~29. Hingson, R. W., Strunin, L., Berlin, B. M., and Heeren, T. (1990) Beliefs about AIDS, use of alcohol and drugs, and unprotected sex among Massachusetts adolescents. American Journal of Public Health 80:295-299. Hochhauser, M. (1987) Readability of AIDS Educational Materials. Presented at the Annual Meeting of the American Psychological Association, New York, August 30. Hofferth, S. L. (1987) Factors affecting initiation of sexual intercourse. In S. L. Hofferth, and C. D. Hayes, eds., Risking the Future. Vol. 2, Working Papers and Statistical Appendixes. Washington, D.C.: National Academy Press. Hofferth, S. L., Kahn, J. R., and Baldwin, W. (1987) Premarital sexual activity among U.S. teenage women over the past three decades. Family Planning Perspectives 19:4~53. Horan, J. J., and Williams, J. M. (1982) Longitudinal study of assertion training as a drug abuse prevention strategy. American Educational Research Journal 19:341-35 1. Horton, J., Alexander, L., and Brundage, J. (1989) HIV prevalence among military women: An examination of military applicant, active duty and reserve testing data. Presented at the Fifth International Conference on AIDS, Montreal, June =9.

242 ~ AIDS: THE SECOND DECADE Howard, J., Taylor, J. A., Ganikos, M. L., Holder, H. D., Godwin, D. F., and Taylor, E. D. (1988) An overview of prevention research: Issues, answers, and new agendas. Public Health Reports 103:674 683. Hudson, R. A., Petty, B. A., Freeman, A. C., Haley, C. E., and Krepcho, M. A. (1989) Adolescent runaways' behavioral risk factors, knowledge about AIDS and attitudes about condom usage. Presented at the Fifth International Conference on AIDS, Montreal, June ~9. Hummel, R., Rodriguez, G., Brandon, D., and Wells, D. (1989) Outreach model for HIV positive adolescents and adolescents currently at high risk for HIV infection. Presented at the Fifth International Conference on AIDS, Montreal, June =9. Inazu, J. K., and Fox, G. L. (1980) Maternal influence on the sexual behavior of teenage daughters. Journal of Family Issues 1:81-102. Institute of Medicine (IOM). (1988) [1omelessness, Health, and Human Needs. Wash- ington, D.C.: National Academy Press. Institute of Medicine/National Academy of Sciences (IOM/NAS). (1986) Confronting AIDS: New Directions in Public Health, Health Care, and Research. Washington, D.C.: National Academy Press. Institute of Medicine/National Academy of Sciences (IOM/NAS). (1988) Confronting AIDS: Update 1988. Washington, D.C.: National Academy Press. Jackson, J., and Neshin, S. (1986) New Jersey community health project: Impact of using ex-addict education to disseminate infonnation on AIDS to intravenous drug users. Presented at the Second International Conference on AIDS, Paris, June 25-26. Jackson, J., and Rotl~iewicz, L. (1987) A coupon program: AIDS education and drug treatment. Presented at the Third International Conference on AIDS, Washington, D.C., June 1-5. Jaffe, L. R., Seehaus, NJ., Wagner, C., and Leadbeater, B. J. (1988) Anal intercourse and knowledge of acquired immunodeficiency syndrome among minonty-group female adolescents. Journal of Pediatrics 112:1005-1007. Jessor, R., Chase, J. A., and Donovan, J. E. (1980) Psychosocial correlates of marijuana use and problem drinking in a national sample of adolescents~merican Journal of Public Health 70:6~613. Jessor, R., Costa, F., Jessor, S. L., and Donovan, J. E. (1983) Time of first intercourse: A prospective study. Journal of Personality and Social Psychology 44:608~26. Jessor, S. L., and Jessor R. (1975) Transition from virginity to nonvirginity: A social-psychological study over time. Developmental Psychology 11:473~84. Job, R. F. S. (1988) Effective and ineffective use of fear in health promotion campaigns. American Journal of Public Health 78:163-167. Johnston, L. D., Bachman, J. G., and O'Malley, P. M. (1984) Monitoring the Future: Questionnaire Responses from the Nation's High School Seniors, 1983. Ann Arbor, Mich.: Institute for Social Research. Johnston, L. D., Bachman, J. G., and O'Malley, P. M. (1989) Results of the 1988 National High School Senior Survey (press release). University of Michigan, Ann Arbor, Mich., Februaly 28. Johnston, L. D., O'Malley, P. M., and Bachman, J. G. (1988) Illicit Drug Use, Smoking, and Drinking by America's High School Students, College Students, and Young Adults 1975-1987. Rockville, Md.: National Institute on Drug Abuse

AIDS AND ADOLESCENTS ~ 243 Jones, E. F., Forrest, J., Goldman, N., Henshaw, S. K., Lincoln, R., et al. (1985) Teenage pregnancy in developed countries: Determinants and policy implications. Family Planning Perspectives 17:53~3. Kahn, J. R., Kalsbeck, W. D., and Hofferth, S. L. (1988) National estimates of teenage sexual activity: Evaluating the comparability of three national surveys. Demography 25:189-204. Kahn, J. R., Rindfuss, R. R., and Guilkey, D. K. (in press) Adolescent contraceptive method choices. Demography. Kandel, D. B. (1975) Stages in adolescent involvement in drug use. Science 190:912- 914. Kandel, D. B. (1978) Similarity in real-life adolescent friendship pairs. Journal of Personality and Social Psychology 36:30~312. Kandel, D. B. (1980) Drug and drinking behavior among youth. In I. Coleman, A. Inkeles, and N. Smelser, eds., Annual Review of Sociology 6:235-285. Kandel, D. B. (1984) Marijuana users in young adulthood. Archives of General Psychiatry 41:20~209. Kandel, D. B. (1985) On processes of peer influences in adolescent drug use: A developmental perspective. In B. Stimmel, ea., Alcohol and Substance Abuse in Adolescence. New York: Haworth Press. Kandel, D. B., and Davies, M. (In press) Cocaine use in a national sample of U.S. youth (NLSY): Epidemiology, predictors and ethnic pattems. In C. Schade and S. Scholer, eds., The Epidemiology of Cocaine Use and Abuse, National Institute on Drug Abuse Research Monograph. Rockville, Md.: National Institute on Drug Abuse. Kandel, D. B., and Logan, J. A. (1984) Patterns of drug use from adolescence to young adulthood: I. Periods of risk for initiation, continued use, and discontinuation. American Journal of Public Health 74 660466. Kandel, D. B., Kessler, R. C., and Margulies, R. Z. (1978) Antecedents of adolescent initiation into stages of Snug use: A developmental analysis. Journal of Youth and Adolescence 7:13~0. Kandel, D. B., Murphy, D., and Kams, D. (1985) Cocaine use in young adulthood: Patterns of use and psychosocial correlates. In N. J. Kozel, and E. H. Adams, eds., Cocaine Use in America: Epidemiologic and Clinical Perspectives. National Institute on Drug Abuse Research Monograph 61. Rockville, Md.: National Institute on Drug Abuse. Keating, D. P. (1980) Thinking processes in adolescence. In J. Adelson, ea., Handbook of Adolescent Psychology. New York: John Wiley and Sons. Kegeles, S., Greenblatt, R., Catania, J., Cardenas, C., Gottlieb, J., and Coates, T. (1989) AIDS risk behavior among sexually active Hispanic and Caucasian adolescent females. Presented at the Fifth International Conference on AIDS, Montreal, June 09. Kegeles, S. M., Adler, N. E., and Irwin, Jr., C. E. (1988) Sexually active adolescents and condoms: Changes over one year in knowledge, attitudes and use. American Journal of Public Health 78:460~61. Kelley, P. W., Miller, R. N., Pomerantz, R., Wann, F., Brundage, J. F., and Buck, D. S. (1990) Human immunodeficiency virus seropositivity among members of the active duty U.S. Army 1985-1989. American Journal of Public Health 80:405~10.

244 ~ AIDS: THE SECOND DECADE Kilbourne, B. W., Buehler, J. W., and Rogers, M. F. (1990) AIDS as a cause of death in children, adolescents, and young adults. American Journal of Public Health 80:499-500. Kilbourne, B. W., Rogers, M. F., and Bush, T. J. (1989) The relative importance of AIDS as a cause of death in pediatric and young adult populations in the U.S. 198~1987. Presented at the Fifth International Conference on AIDS, Montreal, June =9. Kiselica, M. S. (1988) Helping an aggressive adolescent through the "Before, during and after program." The School Counselor 35:299-306. Kleinman, P. H., Goldsmith, D. S., Friedman, S. R., Mauge, C. E., Hopkins, W., and Des Jarlans, D. C. (In press.) Knowledge about and behaviors affecting the spread of AIDS. International Journal of the Addictions. Kolbe, L., Jones, J., Nelson, G., Daily, L., Duncan, C., et al. (1988) School health education to prevent the spread of AIDS: Overview of a national program. Hygie 7: 1(~13. Kreipe, R. E., and Strauss, J. (1989) Adolescent medical disorders, behavior, and development. In G. R. Adams, R. Montemayor, and T. P. Gullotta, eds., The Biology of Adolescent Behavior and Development. Newbury Park, Calif.: Sage Publications. Lamke, L. K., Lujan, B. M., and Showalter, J. M. (1988) The case for modifying adolescents' cognitive self-statements. Adolescence 92:967-974. Lashley, K. S., and Watson, J. B. (1922) A Psychological Study of Motion Pictures in Re- lation to Venereal Disease Campaigns. Washington, D.C.: U.S. Interdepartmental Social Hygiene Board. Lawrence, F. C., Tasker, G. E., Daly, C. T., Orhiel, A. L., and Wozniak, P. H. (1986) Adolescent's time spent viewing television. Adolescence 21:431~36. Lemp, G. F., Payne, S. F., Neal, D., Temelso, T., and Rutherford, G. W. (1990) Survival trends for patients with AIDS. Journal of the American Medical Association 263:402~06. Lesnick, H., and Pace, B. (1990) Knowledge of AIDS risk factors in South Bronx minority college students. Journal of Acquired Immune Deficiency Syndromes 3:173-176. Lewis, C. E., and Lewis, M. A. (1984) Peer pressure and risk-taking behaviors in children. American Journal of Public Health 74:58() 584. Longini, Jr., I. M., Clark, W. S., Horsburgh, C. R., Lemp, G. F., Byers, R. H., et al. (1989) Statistical analysis of the stages of HIV infection using a Markov model. Presented at the Fifth Intemational Conference on AIDS, Montreal, June =9. Lui, K., Darrow, W. W. and Rutherford, G. W. (1988) A model-based estimate of the mean incubation penod for AIDS in homosexual men. Science 240:1333-1335. Mantell, J. E., and Schinke, S. P. (In press) The crisis of AIDS for adolescents: The need for preventive risk-reduction interventions. In A. R. Roberts, ea., Crisis Intervention Handbook. New York: Springer. Marks, A., Malizio, J., Hoch, J., Brody, R., and Fisher, M. (1983) Assessment of health needs and willingness to utilize health care resources of adolescents in a suburban population. Journal of Pediatrics 102:456~60. Marlatt, G. A. (1982) Relapse prevention: A self-control program for the treatment of addictive behaviors. In R. B. Stuart, ea., Adherence, Compliance and Generaliza- tion in Behavioral A1edicine. New York: Brunner/Mazel.

AIDS AND ADOLESCENTS 245 Martin, A. D., and Hetrick, E. S. (1987) Designing an AIDS risk reduction program for gay teenagers: Problems and proposed solutions. In D. G. Ostrow, ea., Biobehavioral Control of AIDS. New York: Irvington Publishers, Inc. Mason, J. O., Noble, G. R., Lindsey, B. K., Kolbe, L. J., Van Ness, P., et al. (1988) Current CDC efforts to prevent and control human immunodeficiency virus infection and AIDS in the United States through information and education. Public Health Reports 103:255-260. Mata, A. G., and Jorquez, J. S. (1988) Mexican-A:nencan intravenous drug users' and needle-sharing practices: Implications for AIDS prevention. In R. J. Battjes, and R. W. Pickens, eds., Needle Sharing Among Intravenous Drug Abusers: National and International Perspectives. National Institute on Drug Abuse Research Monograph 80. Washington, D.C.: U.S. Government Printing Office. Mays, V. M., and Cochran, S. D. (1987) Acquired immunodeficiency syndrome and black Americans: Special psychosocial issues. Public Health Reports 102:22= 231. McCaul, K. D., and Glasgow, R. E. (1985) Preventing adolescent smoking: What have we learned about treatment construct validity? Health Psychology 4:361-387. McKusick, L., Coates, T. J., and Babcock, K. (1988) Knowledge and attitudes about AIDS and sexual behavior in California high school students. Presented at the Fourth International Conference on AIDS, Stockholm, June 12-16. Mensch, B. S., and Kandel, D. B. (1988a) Dropping out of high school and drug inv olv ement. Soc to logy of Education 6 1: 95 - 11 3 . Mensch, B. S., and Kandel, D. B. (1988b) Underreporting of substance use in a national longitudinal youth cohort: Individual and interviewer effects. Public Opinion Quarterly 52:10~124. Meyer-Bahlburg, H. F. L. (1980) Sexuality in early adolescence. In B. B. Wolman, and J. Money, eds., Handbook of Human Sexuality. Englewood Cliffs, N.J.: Prentice-Hall. Miller, P. Y., and Simon, W. (1974) Adolescent sexual behavior: Context and change. Social Problems 22:58-76. Mitchell, F., and Brindis, C. (1987) Adolescent pregnancy: Ihe responsibilities of policy makers. Health Services Research 22:399~37. Moore, D., and Schultz, N. (1983) Loneliness at adolescence: Correlates, attnbutions, and coping. Journal of Youth and Adolescence 12:187-196. Moore, K. A., Nord, C. W., and Peterson, J. L. (1989) Nonvoluntary sexual activity among adolescents. Family Planning Perspectives 21:11~114. Moore, K. A., Wenk, D., Hofferth, S. L., and Hayes, C. D., eds. (1987) Statistical appendix: Trends in adolescent sexual and fertility behavior. In S. L. Hofferth, and C. D. Hayes, eds., Risking the Future. Vol. 2, Working Papers and Statistical Appendixes. Washington, D.C.: National Academy Press. Morris, R., Huscroft, S., Roseman, J., Re, O., Baker, C. J., and Iwakoshi, K. A. (1989) Demographic and high-risk behavior study of incarcerated adolescents. Presented at the Fifth International Conference on AIDS, Montreal, June =9. Morisky, D. E., DeMuth, N. M., Field-Fass, M., Green, L. W., and Levine, D. M (1985) Evaluation of family health education to build social support for long-term control of high blood pressure. Health Education Quarterly 12:35-50. Mosher, W. D., and Bachrach, C. A. (1987) First premarital contraceptive use: United States, 196~82. Studies in Family Planning 18:83-95.

246 ~ AIDS: THE SECOND DECADE Mott, F. L., and Haunn, R. J. (1988) Linkages between sexual activity and alcohol and drug use among American adolescents. Family Planning Perspectives 20:128-136. Murray, D. M., and Petty, C. L. (1985) The prevention of adolescent drug abuse: Implications of etiological, developmental, behavioral, and environmental models. In C. L. Jones, and R. J. Battjes, eds., Etiology of Drug Abuse: Implications for Prevention. National Institute on Drug Abuse Research Monograph 56. Rockville, Md.: U.S. Department of Health and Human Services. Nahmias, A., Corey, L., Lee, F., Clumeck, N., Cannon, R., and Holmberg, S. (1989) Genital herpes as a possible risk factor for HIV transmission. Presented at the Fifth International Conference on AIDS, Montreal, June =9. National Institute on Drug Abuse (NIDA). (1988) National Household Survey on Drug Abuse: Main Findings 1985. Rockville, Md.: National Institute on Drug Abuse. Newcomb, M. D., and gentler, P. M. (1988) Consequences of Adolescent Drug Use: Impact on the Lives of Young Adults. Newbu~ Park, Calif.: Sage. Newcomb, M. D., and gentler, P. M. (1989) Substance use and abuse among children and teenagers. American Psychologist 44:242-248. Newcomer S. F., and Udry, J. R. (1983) Adolescent sexual behavior and popularity. Adolescence 18:515-522. Newmeyer, J. A. (1988) Why bleach? Development of a strategy to combat HIV contagion among San Francisco intravenous drug users. In R. J. Battjes, and R. W. Pickens, eds., Needle Sharing Among Intravenous Drug Abusers: National and International Perspectives. NIDA Research Monograph 80. Washington, D.C.: U.S. Government Printing Office. New York State Department of Health. (1989) AIDS in New York State through 1988. Albany, N.Y.: New York State Department of Health. Novello, A. C. (1988) Secretary's Work Group on Pediatric HIV Infection and Disease. Washington, D.C.: Department of Health and Human Services. Novick, L. F., Beans, D., Stricof, R., Stevens, R., Pass, K., and Wethers, J. (1989a) HIV seroprevalence in newborns in New York State. Journal of the American Medical Association 261:1745-1750. Novick, L. F., Glebatis, D., Stricof, R., and Berns, D. (1989b) HIV infection in adolescent childbearing women. Presented at the Fifth International Conference on AIDS, Montreal, June 1~. O'Donnell, J. A. (1985) InteIpreting progression from one drug to another. In L. N. Robins, ea., Studying Drug Abuse. New Brunswick, N.J.: Rutgers University Press. O'Donnell, J. A., and Clayton, R. R. (1982) The stepping-stone hypothesis: MaIijuana, heroin, and causality. Chemical Dependencies 4:229-241. Office of Technology Assessment (OTA). (1988) How Effective is AIDS Education? Washington, D.C.: Office of Technology Assessment. Osgood, D. W., Johnston, L. D., O'Malley, P. M., and Bachman, J. G. (1988) The generality of deviance in late adolescence and early adulthood. American Sociological Review 53:81-93. Paget, K. D. (1988) Adolescent pregnancy: Implications for prevention strategies in educational settings. School Psychology Review 17:57~580.

AIDS AND ADOLESCENTS ~ 247 Parcel, G. S., Nader, P. R., and Meyer, M. P. (1977) Adolescent health concerns, problems and patterns of obligation in a triethnic urban population. Pediatrics 66:157-164. Pearl, D., Bouthilet, L., and Lazar, J., eds. (1982) Television and Behavior: Ten Years of Scientif c Progress and Implications for the Eighties. Vol. 1: Summary Reports. Rockville, Md.: U.S. Department of Health and Human Services. Pentz, M. A., Cormack, C., Flay, B., Hansen, W. B., and Johnson, C. A. (1986) Balancing program and research integrity in community drug abuse prevention: Project STAR approach. Journal of School Health 56:389-393. Pentz, M.A., Dwyer, J. H., MacKinnon, 1:). P., Flay, B. R., Hansen, W. B., et al. (1989a) A mu' ticommunity trial for primary prevention of adolescent drug abuse. Journal of the American Medical Association 261:3259-3266. Pentz, M. A., MacKinnon, D. P., Flay, B. R., Hansen, W. B., Johnson, C. A., and Dwyer, J. H. (1989b) Primary prevention of chronic diseases in adolescence: Effects of the Midwestern prevention project on tobacco use. American Journal of Epidemiology 130:713-724. Perlman, J. A., Kelaghan, J., Wolf, P. H., Baldwin, W., Coulson, A., and Novello, A. (1990) HIV risk difference between condom users and nonusers among U.S. heterosexual women. Journal of Acquired Immune Deficiency Syndromes 3:155- 165. Perry, C. L., Klepp, K. I., and Schultz, J. M. (1988) Primary prevention of cardiovascular disease: Community-wide strategies for use. Journal of Consulting and Clinical Psychology 56:358-364. Peterson, M. R., Mumm, A. H., Mathis, R., Kelley, P. W., White S. L., et al. (1988) Prevalence of HIV antibody in U.S. active-duty military personnel, April 1988. Morbidity and Mortality Weekly Report 37:461~63. Polich, J. M., Ellickson, P. L., Reuter, P., and Kahan, J. P. (1984) Strategies For Controlling Adolescent Drug Use. Santa Monica, Calif.: The Rand Corporation. Polit, D. F., and Kahn, J. R. (1986) Early subsequent pregnancy among economically disadvantaged teenage mothers. American Journal of Public Health 76:167-171. Pratt, W. F., Mosher, W. D., Bachrach, C. A., and Horn, M. C. (1984) Understanding U.S. fertility: Findings from the National Survey of Family Growth, Cycle III. Population Bulletin 39:3~1. Price, J. H., Desmond, S., and Kukulka, G. (1985) High school students' perceptions and misperceptions of AIDS. Journal of School Health 55:107-109. Quinn, J. (1988) Natural allies: Youth organizations as partners in AIDS education. In M. Quackenbush, M. Nelson, and K. Clark, eds., The AIDS Challenge: Prevention Education for Young People. Santa Cruz, Calif.: Network Publications. Quinn, T. C., Glasser, D., Cannon, R. O., Matuszak, D. L., Dunning, R. W., et al. (1988) Human immunodeficiency v~rus infection among patients attending clinics for sexually transmitted diseases. New England Journal of Medicine 318:197-203. Radius, S. M., Dielman, T. E., Becker, M. H., Rosenstock, I., and Horvath, W. J. (1980) Health beliefs of the school-aged child and their relationship to risk-taking behaviors. International Journal of Health Education 23:3-11. Remafedi, G. (1988) Preventing the sexual transmission of AIDS during adolescence. Journal of Adolescent Health Care 9:139-143.

248 ~ AIDS: THE SECOND DECADE Reuben, N., Hein, K., Drucker, E., Bauman, L., and Lauby, J. (1988) Relationship of high-risk behaviors to AIDS knowledge in adolescent high school students. Presented at the Annual Research Meeting of the Society for Adolescent Medicine, New York City, March. Rizvi, M. H. (1983) An empirical investigation of some item nonresponse adjustment procedures. In W. G. Madow, H. Nisselson, and I. Olkin, eds., Incomplete Data in Sample Surveys, Vol. 1, Report and Case Studies (Report of the National Research Council Panel on Incomplete Data). New York: Academic Press. Roberts, E. S., Kline, D., and Gagnon, I. (1981) Family Life and Sexual Learning of Children. Vol. 1. Cambridge, Mass.: Population Education, Inc. Robins, L. N., and Wish, E. (1977) Childhood deviance as a developmental process: A study of 223 urban black men from birth to 18. Social Forces 56:448~73. Rolf, J., Nanda, J., Thompson, L., Mamon, J., Chandra, A., et al. (1989) Issues in AIDS prevention among juvenile offenders. In J. O. Woodruff, D. Doherty, and J. G. Athey, eds., Troubled Adolescents and HIV Infection. Washington, D.C.: Child and Adolescent Service System Program (CASSP), Georgetown University Child Development Center. Rolf, J., Nanda, J., Baldwin, J., Chandra, A., and Thompson, L. (In press) Substance abuse and HIV/AIDS risk among delinquents: A prevention challenge. Inter- national Journal of Addictions, Silver Anniversary Issue on Prevention (Special Issue No. 3~. Rosenbaum, E., and Kandel, D. B. (In press) Early onset of adolescent sexual behavior and drug involvement. Journal of Marriage and the Family. Rosenberg, M. (1965) Society and the Adolescent Self Image. Princeton, N.J.: Princeton University Press. Rotheram-Borus, M. J., Selfndge, C., Koopman, C., Haignere, C., Meyer-Bahlburg, H., and Ehrhardt, A. (1989) The relationship of knowledge and attitudes toward AIDS to safe sex practices among runaway and gay adolescents. Presented at the Fifth International Conference on AIDS, Montreal, June 09. Rothman, J., and David, T. (1985) Status offenders in Los Angeles County: Focus on runaway and homeless youth. Los Angeles: School of Social Welfare, University of California at Los Angeles. Santrock, J. W. (1981) Adolescence: An Introduction. Dubuque, Iowa: William C. Brown. Schinke, S. P. (1984) Preventing teenage pregnancy. In M. Hersen, R. M. Eisler, and P. M. Miller, eds., Progress in Behavior Modification. San Prancisco: Academic Press. Severson, H. H. (1984) Adolescent social drug use: School prevention program. School Psychology Review 13:15~161. Shaffer, D., and Caton, D. (1984) Runaway and homeless youth in New York City: A report to the Ittleson Foundation. New York: The Ittleson Foundation. Sharp, E., Cowan, D., Goldenbaum, M., Brundage, J., and McNeil, J. (1989) Epidemi- ology of HIV infection among young adults in the U.S.: Regional variations and trends. Presented at the Fifth International Conference on AIDS, Montreal, June =9.

AIDS AND ADOLESCENTS ~ 249 Siegel, D., Lazarus, N., Durbin, M., Krasnovsky, F., Chesney, M., and Kakimoto, D. (1989) AIDS prevention in junior high school students in an AIDS epicenter: Results of a baseline survey. Presented at the Fifth International Conference on AIDS, Montreal, June =9. Single, E., Kandel, D. B., and Johnson, B. D. (1975) The reliability and validity of drug use responses in a large-scale longitudinal swvey. Journal of Drug Issues 5:426~43. Smith, E. A., and Udder, J. R. (1985) Coital and non-coital sexual behaviors of white and black adolescents. American Journal of Public Health 75:120~1203. Sobol, J. (1987) Health concerns of young adolescents. Adolescence 22:739-750. Sonenstein, F. L., Fleck, J. H., and Ku, L. C. (1989a) At risk of AIDS: Behaviors, knowledge and attitudes among a national sample of adolescent males. Presented at the Annual Meeting of the Population Association of America, Baltimore, Md., March 31. Sonenstein, F. L., Pleck, J. H., and Ku, L. C. (1989b) Sexual activity, condom use and AIDS awareness among adolescent males. Family Planning Perspectives 21:152-158. Sorenson, R. C. (1973) Adolescent Sexuality in Contemporary America. New York: World Publishing Co. Spencer, B. D. (1989) On the accuracy of current estimates of the numbers of intravenous drug users. In C. F. Turner, H. G. Miller, and L. E. Moses, eds., AIDS, Sexual Behavior, and Intravenous Drug Use. Washington, D.C.: National Academy Press. St. Louis, M. E., Hayman, C. R., Miller, C., Anderson, J. E., Peterson, L. R., and Dondero, T. J. (1989) HIV infection in disadvantaged adolescents in the U.S.: Findings from the Job Corps screening program. Presented at the Fifth International Conference on AIDS, Montreal, June =9. Stall, R., McKusick, L., Wiley, J., Coates, T. J., and Ostrow, D. G. (1986) Alcohol and drug use during sexual activity and compliance with safe sex guidelines for AIDS: The AIDS Behavioral Research Project. Health Education Quarterly 13:359-371. Stark, E. (1986) Young, innocent, and pregnant. Psychology Today 20:28-30,32-35. Sternleib, J. J., and Muncan, L. (1972) A survey of health problems, practices and needs of youth. Pediatrics 49:177-186. Stimson, G. V. (1988) Injecting Equipment Exchange Schemes: Final Report. London: Monitoring Research Group, Sociology Department, Goldsmith's College. Stimson, G. V. (1989) Syringe exchange programmer for injecting drug users. AIDS 5:253-261. Strecher, V. J., DeVellis, B. M., Becker, M. H., and Rosenstock, I. M. (1986) The role of self-efficacy in achieving health behavior change. Health Education Quarterly 13:73-91. Strunin, L., and Hingson, R. (1987) Acquired immunodeficiency syndrome and adoles- cents: Knowledge, beliefs, attitudes and behavior. Pediatrics 79:825-828. Sutton, S. R. (1982) Fear-arousing communications: A critical examination of theory and research. In J. R. Eiser, ea., Social Psychology and Behavioral Medicine. New York: John Wiley and Sons.

250 ~ AIDS: THE SECOND DECADE Telzak, E. E., Chiasson, M. A., Stoneburner, R. L., Rivera, J., Jaffee, H. W., and Schultz, S. (1989) A prospective cohort study of HIV-1 seroconversion in patients with genital ulcer disease in New York City. Presented at the Fifth International Conference on AIDS, Montreal, June k9. Temoshok, L., Moulton, J. M., Elmer, R. M., Sweet, D. M., Baxter, M., and Shalwitz, J. (1989) Youth in detention at high risk for HIV: Knowledge, attitudes and behaviors regarding condom use. Presented at the Fifth International Conference on AIDS, Montreal, June =9. Tolsma, D. D. (1988) Activities of the Centers for Disease Control in AIDS education. Journal of School Health 58: 133-136. Trussell, J. (1988) Teenage pregnancy in the United States. Family Planning Perspec- tives 20:262-272. Turner, C. F., Miller, H. G., and Barker, L. (1989) AIDS research and the behavioral and social sciences. In R. Kulstad, ea., AIDS, 1988 Washington, D.C.: American Association for the Advancement of Science. Turner, C. F., Miller, H. G., arid Moses, L. E., eds. (1989) AIDS, Sexual Behavior, and Intravenous Drug Use. Washington, D.C.: National Academy Press. Udry, J. R., and Billy, J. O. G. (1987) Initiation of coitus in early adolescence. American Sociological Review 52:841-855. U.S. Bureau of the Census. (1985) Statistical Abstract of the United States: 1985. 105th ed. Washington, D.C.: U.S. Government Printing Office. U.S. Bureau of the Census. (1987) Statistical Abstract of the United States: 1988. 108th ed. Washington, D.C.: U.S. Government Printing Office. Valdiserri, R. O. (1989) Preventing AIDS: The Design of Effective Programs. New Brunswick, N.J.: Rutgers University Press. van den Hoek, I. A. R., Coutinho, R. A., van Haastrecht, H. J. A., van Zadelhoff, A. W., and Goudsmit, J. (1988) Prevalence and risk factors of HIV infection among drug users and drug-using prostitutes in Amsterdam. AIDS 2:55~0. van den Hock, J. A. R., van Haastrecht, H. J. A., and Coutinho, R. A. (1990) Risk reduction among intravenous drug users in Amsterdam under the influence of AIDS. American Journal of Public Health 79:1355-1357. Vermund, S. H., Hein, K., Gayle, H. D., Caky, J. M., Thomas, P. A., and Drucker, E. (1989) Acquired irnmunodeficiency syndrome among adolescents. American Journal of Diseases in Children 143:122~1225. Vemon, M. E. L., Green, J. A., and Frothingham, T. E. (1983) Teenage pregnancy: A prospective study of self-esteem and other sociodemographic factors. Pediatrics 72:632~35. Vincent, M. L., Clearie, A. F., and Schluchter, M. D. (1987) Reducing adolescent preg- nancy through school and community-based education. Journal of the American Medical Association 257:3382-3386. Vlahov, D., Anthony, J. C., Celentano, D. D., Solomon, L., Choudhury, N., and Mandell, W. (1989) Trends of risk reduction among initiates into intravenous drug use 1982-1987. Presented at the Fifth International Conference on AIDS, Montreal, June =9. VoeLker, R. (1989) No uniform policy among states on HIV/AIDS education. American Medical News September 15:3.

AIDS AND ADOLESCENTS ~ 251 Walker, D. K., Cross, A. W., Heyman, P. W., Ruch-Ross, H., Benson, P., and Tuthill, J. W. G. (1982) Comparisons between inner-city and private school adolescents' perception of health problems. Journal of Adolescent Health Care 3:82-90. Watters, J. K., (1987) Preventing human immunodeficiency virus contagion among intravenous drug users: The impact of street-based education on risk behavior. Presented at the Third International Conference on AIDS, Washington, D.C., June 1-5. Wetzel, J. R. (1987)American Youth: A Statistical Snapshot. Washington, D.C.: The William T. Grant Foundation. Wiebel, W. W. (1988) Combining ethnographic and epidemiologic methods in targeted AIDS interventions: The Chicago model. In R. J. Battjes, and R. W. Pickens, eds., Needle Sharing Among Intravenous Drug Abusers: National and Interna- tional Perspectives. National Institute on Drug Abuse Research Monograph 80. Washington, D.C.: U.S. Government Printing Office. Wigersma, L., and Oud, R. (1987) Safety and acceptability of condoms for use by homosexual men as a prophylactic against transmission of HIV during anogenital sexual intercourse. British Medical Journal 295:94. Yamaguchi, K.? and Kandel, D. B. (1984a) Patterns of drug use from adolescence to young adulthood: II. Sequences of progression. American Journal of Public Health 74:668-672 Yamaguchi, K., and Kandel, D. B. (1984b) Patterns of drug use from adolescence to young adulthood: III. Predictors of progression. American Journal of Public Health 74:673~81. Yates, G., MacKenzie, R., Pennbndge, J., and Cohen, E. ( 1988) A risk profile comparison of runaway and non-runaway youth. American Journal of Public Health 78:82~821. Zabin, L. S., and Clark, S. D. (1983) Institutional factors affecting teenagers' choice and reasons for delay in attending a family planning clinic. Family Planning Perspectives 15:25-29. Zabin, L. S., Kantner, J. F., and Zelnik, M. (1979) The risk of adolescent pregnancy in the first months of intercourse. Family Planning Perspectives 11:215-222. Zabin, L. S., Hardy, J. B., Smith, E. A., and Hirsch, M. B. (1986) Substance use and its relation to sexual activity among inner-city adolescents. Journal of Adolescent Health Care 7:32~331. Zelnik, M. (1983) Sexual activity among adolescents: Perspective of a decade. In E. R. McAnarney, ea., Premature Adolescent Pregnancy and Parenthood. New York: Grune and Stratton. Zelnik, M., and Kantner, J. F. (1977) Sexual and contraceptive experience of young unmamed women in the United States, 1976 and 1971. Family Planning Per- spectives 9:55-71. Zelaik, M., and Kantner, J. (1979) Reasons for nonuse of contraception by sexually active women aged 15-19. Family Planning Perspectives 11:289-296. Zelnik, M., and Kantner, J. (1980) Sexual activity, contraceptive use and pregnancy among metropolitan-area teenagers: 1971-1979. Family Planning Perspectives 12:23~237. Zelnik, M., and Shah, F. K. (1983) First intercourse among young Americans. Family Planning Perspectives 15:64 70.

252 ~ AIDS: THE SECOND DECADE Zelnik, M., Kanmer, J. F., and Ford, K. (1981) Sex and Pregnancy in Adolescence. Beverly Hills, Calif.: Sage Publications. Zenilman, J. (1988) Sexually transmitted diseases in homosexual adolescents. Journal of Adolescent Health Care 9:129-138. Zewdie, D., Abdurahman, M., Ayhunie, S., Adal, G., Tadesse, M., and Yemane, B. T. (1989) High prevalence of V-1 antibodies in STD patients with genital ulcers. Presented at the Fifth International Conference on AIDS, Montreal, June =9. Ziffer, J., Ziffer, A., Bywater, M., and Bywater, L. (1989) Knowledge of HIV transmissions and adolescent sexual behavior. Presented at the Fifth International Conference on AIDS, Montreal, June =9.