3
Issues with Design and Implementation of the Current Longitudinal Study

In an ideal world, the Longitudinal Study of Astronaut Health (LSAH) would provide the answer to a cascading series of questions about long term risks among three populations (see Table 3-1).

TABLE 3-1 LSAH Research Questions and Appropriate Study Populations

Research question

Focus population

1. What are the long-term risks of spaceflight

Astronauts with history of spaceflight.

2. What are the long-term risks of preparing for spaceflight

Above population PLUS astronauts who have never flown in space.

3. What are the long-term risks of working in the environment of JSC

Above populations PLUS employees of JSC.

To get an unbiased estimate of these cascading risks, a series of studies with different populations and comparison groups might be designed and carried out. In each of these the astronauts and their comparison group controls should

  • be equivalent at baseline in all factors that influence risk of disease or adverse health outcomes;

  • have equivalent exposures in day-to-day life except for those related to exposure to spaceflight or preparations for spaceflight;

  • have equivalent monitoring for disease by observers blinded to whether they were exposed to spaceflight or preparations for spaceflight; and

  • participate fully from study entry to the outcome of interest.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 29
Review of Nasa’s Longitudinal Study of Astronaut Health 3 Issues with Design and Implementation of the Current Longitudinal Study In an ideal world, the Longitudinal Study of Astronaut Health (LSAH) would provide the answer to a cascading series of questions about long term risks among three populations (see Table 3-1). TABLE 3-1 LSAH Research Questions and Appropriate Study Populations Research question Focus population 1. What are the long-term risks of spaceflight Astronauts with history of spaceflight. 2. What are the long-term risks of preparing for spaceflight Above population PLUS astronauts who have never flown in space. 3. What are the long-term risks of working in the environment of JSC Above populations PLUS employees of JSC. To get an unbiased estimate of these cascading risks, a series of studies with different populations and comparison groups might be designed and carried out. In each of these the astronauts and their comparison group controls should be equivalent at baseline in all factors that influence risk of disease or adverse health outcomes; have equivalent exposures in day-to-day life except for those related to exposure to spaceflight or preparations for spaceflight; have equivalent monitoring for disease by observers blinded to whether they were exposed to spaceflight or preparations for spaceflight; and participate fully from study entry to the outcome of interest.

OCR for page 29
Review of Nasa’s Longitudinal Study of Astronaut Health In the real world, like many expensive, long-running epidemiological studies, the LSAH has had to make a number of compromises. The remainder of this chapter systematically describes and examines those compromises and their effects on the utility of the LSAH. GOALS OF THE LSAH The published goals of the LSAH appear straightforward at first glance. Hamm et al. (2000) list them as follows: The primary aim of the current LSAH is to investigate and describe the incidence of acute and chronic morbidity and mortality of astronauts and to determine whether the unique occupational exposures astronauts encounter are associated with increased risks of morbidity or mortality. Specifically, the primary a priori hypotheses being tested are: Astronauts are at different risk of total and cause-specific mortality than are ground-based employees; and Astronauts are at different risk of total and specific morbidity than are ground-based employees. The Manual of Procedures for the LSAH explains the purpose of the study in several places. The first lines of Chapter 1 state that the purpose is to “examine the incidence of acute and chronic morbidity and mortality of astronauts and describe the risks of morbidity and mortality associated with the astronauts’ occupational exposures, as compared with the risks for civil service employees of Johnson Space Center.” It goes on to say that assessments of lifetime risk and flight-time risks for specific diseases and disorders “will help to (1) identify health-related problems that require spacecraft medical facilities, and (2) devise and implement methods to reduce risks.” Chapter 2 of the Manual includes a copy of the information handout given to potential comparison participants. Under “goals” it includes the statement: “These assessments will be useful in identifying potential health-related problems resulting from exposure to spaceflight.” The letter of invitation sent to potential comparison participants is even more specific, noting that the LSAH has been established “to evaluate the long-range medical effects from exposure to microgravity.” One study could conceivably encompass all three goals with variations in data analysis, but the committee believes that whatever the goal, two different uses of the data are required, each of which imposes slightly varying demands in design and execution of the study. The first of these potential uses is to provide scientific documentation of the long term effects of space flight and preparation for space flight, and thus serve as an indicator of necessary or desirable modifications to current and future training, spacecraft, or operating procedures for

OCR for page 29
Review of Nasa’s Longitudinal Study of Astronaut Health future flights that would enhance the safety of spaceflight for future astronauts. To be useful in this regard, the data would need to be queried regularly but especially when changes in the space program are contemplated (e.g., transition from short-duration shuttle missions to much longer duration missions on the International Space Station, or from the shuttle to a second-generation reusable launch vehicle). The second potential use of the LSAH is as a surveillance instrument to facilitate rapid prevention, detection, and treatment of occupation-related health problems in the group of current and former astronauts. To accomplish this mission, data entry would have to be far more rapid than it has been, and analysis far more frequent. The committee believes that insofar as participation in the LSAH is mandatory for active duty astronauts, NASA is ethically bound to include this surveillance mission along with the research mission. In practice that means that not only should test results and other measures be passed on to individual astronauts and their physicians for individualized evaluation and clinical evaluation, but any group-related information that could influence individual health care must be fully and quickly shared with all affected parties, including payload specialists and astronauts from other countries. The two missions described present different and sometimes conflicting challenges, and striking a balance is no mean feat, but it appears to the committee that, at this point, the LSAH is serving only the first of the two purposes. A case could be made that LSAH data on thyroid function were useful in the decision to reduce the amount of bactericidal iodine used in the shuttle’s potable water, but as noted in the previous chapter, the LSAH seems to have played only a confirmatory role in the discovery of iodine-induced thyroid dysfunction in shuttle astronauts. There is certainly no evidence that the LSAH has ever been queried regularly enough and with sufficiently powerful pattern detection methods to serve as the basis for an effective prevention program for current and former astronauts. That is not to minimize the salutary effects of the annual physicals and other testing of astronauts and former astronauts on the health of individual participants, but those benefits have resulted primarily from individual history and physical and laboratory examinations rather than the study findings. An exemplary surveillance effort requires more frequent reviews and analyses of the accumulating data than a pure research project, and inclusion of additional data as well. No mental health data are included in the current LSAH dataset, for example, although there is no reason to believe that astronauts are immune from psychiatric problems. Such data might also prove useful in the research aspect of the LSAH as well, because many of the other measures that are being collected in this study can be influenced by mental disorders and personality variables. For example, the presence of an eating disorder can have a dramatic impact on measures of bone density. Depression is also increasingly recognized as having a strong negative effect on the morbidity and mortality of many other comorbid conditions if not recognized and treated promptly.

OCR for page 29
Review of Nasa’s Longitudinal Study of Astronaut Health A second difficulty related to the goals and purposes of the LSAH is whether the LSAH is primarily a study of the health consequences of being an astronaut or a study of the long term effects of space flight. Analyses of the data by NASA to date have been consistent with the declared purpose of assessing the occupational risks of a career as an astronaut, but they have not always been suited to the goal of “identifying potential health-related problems resulting from exposure to spaceflight” (Manual of Procedures, Section 2.2). The recent analysis of cancer prevalence (Fischer, 2003), for example, includes astronauts who had not as yet flown a space mission, even though risk for cancer due to exposure to the unique radiation environment of space has been a major concern since the beginning of the space program (Peterson et al., 1993). The database is clearly sufficient to investigate both hypotheses (i.e., that members of the astronaut corps are at no greater risk for cancer than Johnson Space Center [JSC] civil servants, and that spaceflight does not increase the risk of cancer), but the analysis required would be different in each case. If the long-term risks of spaceflight are truly a concern, not only should analyses exclude astronauts who have yet to fly a mission, but some additional power would be gained by including the 27 payload specialists who have flown in space but are not included in the LSAH. Some of these specialists have logged more hours in space than many career astronauts. STATISTICAL POWER An important question for the design of any study is whether it has sufficient power to detect differences among the subject groups in the variables of interest. In the case of the LSAH, this means asking how likely we are to detect a difference between the astronaut and comparison groups in the prevalence of a specific health problem given that astronauts really are at higher risk for that health problem. The two primary pieces of information needed to answer that question are the prevalence of the problem and the number of subjects in the study sample. The number of subjects in the LSAH (312 astronauts, 928 comparisons) is substantial, but small for an epidemiological study and not likely to increase substantially, at least on the astronaut side. More problematic is the fact that the LSAH will potentially be queried about a large and unspecified number of health conditions with widely varying prevalence. That means that the study’s power to detect group differences will vary widely as well. Table 3-2 shows in a readily understandable way how those differences limit the ability of the study to demonstrate increased risk of relatively uncommon health conditions. The table shows the minimum detectable relative risk (for astronauts relative to the comparison participants) for health conditions varying in prevalence in the comparison group from 0.1 percent to 10 percent. Power (the probability of finding an effect given that there is one) was held constant at 0.80, and minimum detectable relative risks were computed for three different significance

OCR for page 29
Review of Nasa’s Longitudinal Study of Astronaut Health levels (α, the probability of falsely concluding that there is an effect when there is none). For health conditions with a 10 percent prevalence in the comparison group, for example, the astronaut group would have to have a prevalence of 14.9 percent (1.49 times the prevalence in the comparison group) to demonstrate an effect at the 0.05 level of statistical significance. Demonstrating significant differences for less common health conditions demands much larger effects. For disorders with a prevalence of only 0.1 percent in the comparison group would require a prevalence 11.15 times higher among the astronauts for the group difference to reach conventional (α = .05) statistical significance. Under these circumstances, it will be especially important to remember that absence of evidence is not evidence of absence. TABLE 3-2 Minimum Detectable Relative Risk (Astronauts versus Comparisons) at Different Criteria for Statistical Significance (two-tailed α) with Power = 0.80 Disease Prevalence in Comparison Group Minimum Detectable Relative Risk   α = 0.05 α = 0.10 α = 0.15 0.1 percent 11.15 9.27 8.18 0.5 percent 3.89 3.44 3.17 1.0 percent 2.81 2.55 2.39 2.0 percent 2.18 2.02 1.92 5.0 percent 1.70 1.61 1.55 10.0 percent 1.49 1.43 1.39 COMPOSITION OF THE COMPARISON GROUP The procedures used for selecting comparison participants from among JSC civil servants have been successful in producing a cohort of individuals closely matched to the astronauts in gender, race, age, and body mass index. The comparison group is less well-matched in education level, but it too is a well-educated group that is much closer to the astronauts in this respect than is the general population. Like the astronauts, they were all gainfully employed at selection and have received regular preventive medical care since they were hired. Like many of the astronauts, they have spent a considerable portion of their adult lives living in Houston. The Flight Medicine Clinic and Occupational

OCR for page 29
Review of Nasa’s Longitudinal Study of Astronaut Health Medicine Clinic at JSC provide physical exams and treatment of job-related injuries and illnesses for the astronauts and JSC employees, respectively. In all these respects, JSC employees make a far better comparison group for the astronauts than the general U.S. population. On the other hand, JSC employee comparison participants may differ from the astronauts in some important unmeasured characteristics. Their physical activity and endurance may be less, and they may be less motivated to achieve and maintain excellent physical conditioning. They may perceive that the reporting of illness will be less threatening to the security of their jobs and thus be more compliant in such reporting. Other unmeasured psychological variables such as risk-taking, harm-avoidance, and competitiveness are both relevant to morbidity and mortality, and they are likely to be different in the two groups. For many of the variables that might affect detection of, susceptibility to, or recovery from specific diseases of interest, the match between the comparison participants and the astronauts is simply unknown. For example, a longitudinal study focused exclusively on cancers would match comparisons on the basis of family cancer history, exposure to known carcinogens other than tobacco (e.g., sunlight, industrial chemicals, diagnostic and therapeutic X-rays), or genetic predisposition (e.g., BRCA1 and BRCA2). Some of this information may well be available in the LSAH database, and it may be possible to adjust statistically for it during data analysis, but the point remains that trying to create an all-purpose comparison group ironically produced one that is less than ideal for any specific health problem. ASCERTAINMENT BIAS It is a truism that the harder one looks for something the more likely one is to find it. Yet this is a serious flaw in the current design of the LSAH. Budgetary shortfalls over the years since the LSAH began have led to increasing disparities in the searches for health problems in the astronaut and comparison groups. The latter now receive physical exams half as frequently as the astronauts; some laboratory tests and physical evaluations that the astronauts receive yearly are provided to comparison participants only at four-year intervals; some procedures (Dual energy X-ray absorptiometry [DEXA] scans, dental exams, comprehensive visual exams) are provided only to astronauts; and there is less follow-up between routine examinations for the employees than for the active duty astronauts (see Appendix B). An additional, though perhaps minor and unavoidable, source of ascertainment bias is that the data collectors, the examining physicians in particular, are not blind to the status of the participants. The flight surgeons examining the astronauts may well look harder for health problems they believe to be associated with spaceflight, while the occupational health clinic physicians examining the comparisons pay more attention to problems that have been identified or that they suspect are associated with ground-based employment at JSC.

OCR for page 29
Review of Nasa’s Longitudinal Study of Astronaut Health In this era of increased sensitivity to patient confidentiality, it cannot be expected that the participating physicians will routinely share their suspicions without a mandate and mechanism to do so. PARTICIPATION Closely related to the ascertainment bias discussed in the previous section is the problem of participant drop out. Wear (2003) provided the committee with data on the percentage of participants appearing for scheduled physical exams each year from 1993 through 2001. Annual physical exams are required of active astronauts, so their participation rate is consistently near 100 percent. Table 3-1 shows the participation rates of active civil servants, comparison participants who are no longer working at JSC, and former members of the astronaut corps. TABLE 3-3 Percentage of Active JSC Civil Servants, Ex-JSC Civil Servants, and Ex-Astronaut LSAH Participants Appearing for Scheduled Physical Exams, 1993-2001   1993 1994 1995 1996 1997 1998 1999 2000 2001 Active JSC Civil Servants 93 79 87 79 79 84 84 82 69 Former JSC Civil Servants 72 53 61 67 59 65 59 63 50 Former Astronauts 71 69 61 64 61 88 63 74 68 Between 70 percent and 90 percent of active JSC civil servants receive their scheduled physical evaluations each year, but neither former JSC civil servants nor former astronauts regularly achieve a 70 percent return rate. Former astronauts have been returning at a slightly higher rate than former civil servants, at least in the 1998 to 2001 period. Nine comparison participants have officially dropped out of the study, and the LSAH staff members have been unable to locate an additional five comparison participants. As noted briefly in the preceding chapter, it is possible that the rest of the study participants are all still “in the study” and a varying percentage simply miss an occasional physical exam. LSAH staff members were unable to provide data to evaluate that hypothesis, but it is their impression that this is not the case. Rather, the same individuals return over and over. The most obvious explanation for the lower participation rates for the former civil servants and former astronauts is that many of them may have left the Houston area and find it inconvenient to return. No study of the reasons for non-participation has been performed, nor have any studies assessed remediable bar-

OCR for page 29
Review of Nasa’s Longitudinal Study of Astronaut Health riers to consistent participation. It may be that participants are more likely to return when healthy (and are able to travel easily) or when sick (and the physical exam may be of most use to them), but the travel requirement could introduce a significant bias. A further source of bias, which may explain why the former astronauts return somewhat more reliably than the former civil servants, is that the former astronauts are reimbursed for their travel to Houston, but the former civil servants are not. If travel is a significant factor, participation might be increased by arranging for or allowing participants to receive their examinations at a more convenient local site. The resulting increase in participation, if there is one, will have to be weighed against the additional variance introduced by inclusion of additional physician examiners. An unrelated concern of the committee that may bear on participation stems from the copy of the informed consent form contained in the LSAH Manual of Procedures provided to the Institute of Medicine (IOM) staff. The committee recognizes, like the IOM committee which authored Safe Passage (IOM, 2001c), that the compulsory health surveillance of the active astronaut corps is occupational in nature and therefore not dependent on approval by the NASA institutional review board (IRB). However, inclusion of the data from active astronauts in the LSAH crosses the line into research activity, which does require the informed consent of the active astronauts. A vigorous explanatory effort should be part of the consent process. Participation in the LSAH by former astronauts and all members of the comparison group is clearly covered by the Common Rule, and the LSAH Manual of Procedures has an appropriate section dealing with obtaining informed consent from these participants. However, the committee is concerned that the content and format of the consent form provided to them is not up to currently acceptable standards, specifically in regard to varying statements about the goals of the study, possible benefits, risks to confidentiality, compensation for participating, and procedures for encouraging participation. The committee believes that the risks to both confidentiality and full participation in the space program may not fall into the minimal risk category and encourages both the LSAH scientists and the NASA IRB to review the current informed consent process and make appropriate changes. INTERNATIONAL PARTNERS The charge to this committee included a request for recommendations on the inclusion of astronauts from NASA’s international partners (Russia, Europe, Japan, Canada). The number of U.S. astronauts is still quite small by the standards of most epidemiological studies, especially when the health problems of greatest concern have a low incidence in the general population. For that reason, inclusion of all space travelers in the LSAH would seem advantageous. Inclusion of cosmonauts who spent months on the Mir space station would seem es-

OCR for page 29
Review of Nasa’s Longitudinal Study of Astronaut Health pecially valuable. On the other hand, given the problems with the study already outlined above, as well as the reported differences among the partner countries in lifestyles, health care, mortality and morbidity, and expert opinion about what health measures and medical procedures are considered valuable or necessary in any circumstance, it is difficult to envision how space travelers from other countries could be enrolled in the LSAH without creating further difficulties in interpretation. Would JSC civil servants be appropriate comparisons for those astronauts and cosmonauts or would they need comparison participants from their own country and culture? If the latter, would the numbers ever be large enough to achieve the statistical power to detect any but the most obvious effects of space flight? Negotiations among the international partners are already underway concerning the content, protocols, testing hardware and software, and data-sharing policies for the Clinical Status Evaluation (CSE). The CSE is to be a standardized battery of clinical, physiological, and psychological tests performed on each long-duration crew member from all International Space Station agencies. Pre-, in-, and post-flight evaluations will be used to guide clinical interventions, identify negative health or fitness trends in preclinical stages, plan optimal postflight rehabilitation, and assess the efficacy of countermeasures. NASA might find value in obtaining historical data from international partner astronauts and cosmonauts that is similar to that in the LSAH database. When and if CSE standards are finalized, the possibility of continuing regular CSE measurements in former International Space Station crews could be entertained. FOLLOW-UP The general question of NASA’s responsibilities when the analysis of LSAH data suggests that astronauts may be at risk for a specific health problem is addressed in Chapter 5. The LSAH and the staff conducting it should not be generally held responsible for implementing preventive or therapeutic countermeasures, but it has and is playing major and very different roles in the two instances reported to date, cataracts and thyroid disorders. The committee therefore comments here on the actions taken in those two cases. When available LSAH data showed an association between cataract incidence in astronauts (eye examinations for comparison participants is limited to visual acuity) and dose of radiation received during space flight radiation, the JSC scientific staff recognized the limitations of the study and contracted for a new study with more objective observation and classification of cataract severity and a comparison/control group of current and former military pilots age and gender matched to the astronauts. They also called for documenting lens opacities with digital photography for both astronauts and comparison participants in future LSAH physical exams and incorporation of cataract susceptibility into the debate on career radiation dose limits. The committee considers this approach a

OCR for page 29
Review of Nasa’s Longitudinal Study of Astronaut Health good general model, that is, whenever LSAH data show even a suggestion of a risk to astronauts (see the section earlier in this chapter on statistical power), the observation should be followed up with a more focused study using within-group comparisons based on such variables as time in space or radiation dose or by comparisons with a hypothesis-specific control group (brand new or a subset of the current comparison group). In contrast, the committee believes that despite taking rapid and effective action to remediate the actual and potential problems of iodine-induced thyroid dysfunction by adding a filter to the space shuttle’s drinking water taps, a proposed change in the medical standards restricting long duration space flight is unwise and premature. The proposal would exclude astronauts with thyroid autoantibodies or thyroid peroxidase antibodies from participating in long duration missions, including the International Space Station, because of concern about increased risk of thyroid dysfunction or even clinical thyroid disease with extended exposure to even low levels of iodine during long duration flight. There are insufficient data in the LSAH at present to fully assess these risks, but elevated thyroid autoantibodies occur about three times more frequently in women than in men and increasing in age (Eheman, 2003). The proposed exclusion would thus disproportionately limit the opportunity for female astronauts to participate on long duration missions. The 16-fold reduction in drinking water iodine concentration produced by the anion exchange resin filters should eliminate or drastically reduce the risk of thyroid dysfunction in all future crew members. Additional filtering or adoption of an alternate water treatment regimen are obvious engineering solutions to that would protect all astronauts. Thyroxin (T4) replacement therapy would control these problems should they occur, but the International Board has previously argued against daily medications on missions, and the board, not JSC, controls the standards for international expeditions. Because prevention and treatment are so simple and effective, the IOM committee believes that it would be a mistake to adopt the proposed policy of excluding astronauts with thyroid autoantibodies or thyroid peroxidase antibodies from participating in long duration missions.