are at maintaining weight and fitness standards, and the long-term health consequences of these behaviors. A DoD-wide evaluation system is recommended.

Several research projects have been conducted by the services on the health-related behaviors of servicemembers. In addition to the wide variety of demographic and personnel data maintained in the Defense Manpower Data Center database, health outcome data are maintained in several medical cost accounting databases.

A combination of the survey instruments that have been used in the past would be suitable for collecting most of the information needed (including longitudinal data). The personnel and medical databases are capable of producing much of the remaining information needed. However, the committee finds that there are two problems with this method of data collection. First, some of the survey data were collected anonymously (with no identification numbers of any type), precluding any attempt to examine the data longitudinally or merge the databases with existing personnel and medical databases that contain the demographic and health outcome data needed for a comprehensive analysis of the data. Second, the personnel and medical databases were not designed to be linked to each other or to survey databases. Thus, although much potentially worthwhile information is collected, little meaningful analysis can be performed.

The committee recommends that the military survey a representative sample of active-duty personnel individually and review the individuals' personnel and medical records during the course of the interview. This method would enable the investigator to obtain all the data needed in a single effort, ensure quality control of the data, build a database that would preserve the anonymity of the individual, and obviate the need to merge automated information systems with highly sensitive data. However, the need to create a system that will obtain information from several large and representative samples of the entire DoD over the course of several years may make this choice cost-prohibitive.

An alternative recommendation is to expand the triennial Survey of Health-Related Behaviors among Military Personnel to include the demographic, medical, nutrition, fitness, and pregnancy data needed. Changing the questionnaire to include social security number, as was done with the Navy's Perceptions of Wellness and Readiness Assessment survey and the Army's Health Risk Appraisal survey, would permit a longitudinal and potentially integrated database to be developed. The practice of using questions from federal surveys of health and fitness-related

Front Matter (R1-R16)

Executive Summary (1-24)

1 Introduction (25-32)

2 Body Composition (33-60)

3 Physical Fitness Policies and Programs (61-86)

4 Weight Management (87-108)

5 Nutritional Concerns of Military Women (109-126)

6 Pregnancy and Lactation and Postpartum Return-To-Duty Fitness (127-156)

7 Conclusions and Recommendations (157-170)

Appendixes (171-172)

A Workshop Summary, Agenda, Participants, and Abstracts (173-252)

B Practices and Policies Tables (253-262)

C Military Occupational Specialty Classification Tables (263-280)

D Search Strategy for Literature Review (281-284)

E Civilian Inquiry Letter and Table of Responses (285-290)

F Abbreviations (291-294)

G Bibliography (295-334)

H Biographical Sketches (335-338)

Index (339-352)