Close
Data in this report were obtained from the Centers for Disease Control and Prevention’s (CDC) Behavioral Risk Factor Surveillance System (BRFSS), the National Health Interview Survey (NHIS), the Substance Abuse and Mental Health Services Administration’s National Survey on Drug Use and Health (NSDUH) and the U.S. Census Bureau’s American Community Survey (ACS).
Data are from 2011-2020. To ensure adequate sample size to detect differences between those who have served and those who have not, in most cases two years of data were combined into three reporting periods used for this report: 2011-2012 (baseline), 2017-2018 (prior period) and 2019-2020 (current). The few exceptions include several BRFSS variables for which data is collected in even-numbered years only, as well as the “able-bodied” measure from ACS, which has a sufficiently large sample size to allow using only one year of data. Note: The COVID-19 pandemic affected data collection for all data sources, resulting in smaller sample sizes in 2020 than in prior years.
Data were weighted according to each survey’s design and weighting methodology to correct for selection bias and enable the development of nationally representative estimates by demographic variables. To reflect the differing age distributions of those who have and have not served, all estimates, with the exception of those presented for the Able-Bodied indicator and its disability-related components, were age-adjusted to the 2000 U.S. Standard Population. Nationally representative, age-adjusted point estimates were calculated and are included in this report for those who have and have not served overall and by gender, age, race/ethnicity, income, education and metropolitan status.
Confidence intervals for all estimates were used to assess whether differences between those who have served and civilians were significant overall and for subpopulations, whether estimates differ significantly between service member subpopulations (e.g., gender, ethnicity) and the significance of change over time. All group differences and changes over time highlighted in the report are statistically significant at the 95% confidence level.
Subpopulation categories were reported consistently across all data sources where possible. However, in some instances, categories were not comparable. Differences include the following: (1) data on the Hawaiian/Pacific Islander racial/ethnic group are not collected in the NHIS; (2) the “other” and “multiple race” categories are combined in the NHIS, whereas the other data sources separate these two groups; and (3) cutoff points for the lowest income group differ across data sources. The lowest income cutoffs are as follows:
- ACS and BRFSS: $25,000
- NSDUH: $30,000
- NHIS: $35,000
In order to ensure that estimates presented are reliable, data were suppressed according to guidance provided by each data source. For the ACS, BRFSS and NHIS, estimates were suppressed if the relative standard error was greater than 30% or if the unweighted group sample size was less than 50. For the NSDUH, estimates were suppressed where estimates approximated 0 or 1 (were less than 0.00005 or greater than 0.999995), if the unweighted sample size was less than 100, or if the relative standard error was greater than 0.175 (using calculation methods suggested by the NSDUH) and if the effective sample size was less than 68.
Limitations
The data presented in this report are reliable, nationally representative, age-adjusted estimates due to the large sample sizes of the pooled years of data, the sampling designs of the data sources and the utilization of age-standardization methods. Further, the sampling designs of these surveys ensure representation by multiple demographic variables. However, a few considerations limit interpretation of the data.
First, each of the four sources of data analyzed for this report asks different questions about military service. Both ACS and BRFSS ask whether a respondent has served on active duty in the U.S. Armed Forces. The NSDUH and NHIS ask whether respondents have served in the U.S. Armed Forces in the past, and exclude those on active duty. Thus, the definition of “those who have served” for ACS and BRFSS measures in this report includes those who have served in the past, and those currently on active duty. The NHIS and NSDUH exclude those who are actively serving. None of the surveys allows analysis by the nature of discharges, specific field of service or involvement in active combat. Only the ACS includes data on the era in which one served. Thus, changes over time could be influenced by cohort effects and may confound the interpretation of age-specific results and comparisons.
Second, while we adjusted in the analyses, samples of those who have served and those who have not may differ from one another on other demographic characteristics. Observed differences between groups may be explained not solely by military service but by other characteristics that vary across the two populations.
Third, readers should be aware that all health outcome data are self-reported, with many health outcome measures asking respondents whether a health care professional has ever told them that they have a disease or condition. This method of collecting data excludes those who may have a condition but are unaware of it due to not having sought treatment or obtained a diagnosis.
Finally, the COVID-19 pandemic affected data collection for all four data sources. In most cases, national surveys halted in-person data collection for part of 2020. When collection resumed, it was more likely to be conducted virtually than in the past. This resulted in higher non-response rates and smaller sample sizes in general. In addition, some items on potentially sensitive subjects (e.g., mental health topics, opioid misuse) experienced higher levels of non-response. This limited our ability to produce reliable estimates for those who have served overall as well as for subpopulations. In addition, due to documented uncertainties around estimates generated from the 2020 ACS, we used ACS data from 2019, even though it was not the most recent year available.