Sleep difficulties as a predictor of attrition in United States Air Force recruits

The objective of the study was to estimate the rates of "difficulty sleeping at night" in newly enlisted United States Air Force (USAF) recruits and determine if these sleep difficulties predicted 1-year attrition (discharge for any reason) independently and after controlling for all other Lackland Behavioral Questionnaire (LBQ) predictors.

The LBQ was administered to 202,339 active duty, enlisted USAF trainees completing basic military training at Joint Base San Antonio-Lackland, Texas, from January 2006 to December 2012. Attrition data were obtained through the Defense Manpower Data Center.

Fifty percent of the sample reported at least occasional sleep difficulties in the past year, with 9% reporting frequent (6% = often and 3% = most of the time) sleep difficulties. Twelve percent of trainees had been discharged within one year, and sleep difficulties were the second strongest significant predictor of this attrition after accounting for all other predictor variables in the LBQ, Wald χ² = 254.19, p < .0001. Trainees with frequent sleep difficulties were 2.7 times more likely to be discharged than those without sleep difficulties based on the odds ratio statistic.

Rates of self-reported frequent "difficulty sleeping…," before basic training started, were similar to civilian populations and were the second strongest predictor of one-year attrition. "Difficulty sleeping…" encompasses many sleep problems (e.g., insomnia, sleep deprivation, circadian misalignment). Future studies should determine what specific sleep difficulties pose the greatest risk for attrition and then determine if they can be remediated, thus decreasing attrition risk, or if they should be a focus of screening.