There were 55
021 participants who completed the baseline (July 2001 to June 2003) and the first follow-up (June 2004 to February 2006) questionnaires, of whom 22
208 (40%) deployed in support of the conflicts in Iraq and Afghanistan. The mean time between baseline and submission of the first follow-up questionnaire was 2.7 years (SD 0.5; median 2.8). To study a more homogenous group of deployers who were exposed to one or more traumatic events during deployment, we removed from the analyses those participants whose first deployment did not occur between baseline and follow-up or who did not report combat exposures (n=16
172). To examine deployers with no recorded history of symptoms of PTSD, we also removed those who reported a previous physician diagnosis of PTSD (n=86) or who had symptoms of PTSD (n=98) at baseline. Similarly, we excluded participants who self reported (at follow-up) being diagnosed with PTSD before deployment (n=11) and those with missing baseline or follow-up PTSD, functional health, or covariate data (n=431), leaving 5410 for analysis.
Table 1 shows baseline population characteristics by centile of mental and physical component summary scores. Those whose mental component summary scores were below the 15th centile were proportionately more likely to be women, younger, less educated, not married, non-officers, to have an occupational position other than combat, health care or functional support (such as technical or electrical repair position), and to report current smoking and have a positive CAGE result. Participants below the 15th centile for mental component summary score had a mean score of 37.9, compared with 55.4 and 62.4 for the 15th to 85th and higher than 85 centile groups, respectively. Those whose physical component summary scores were below the 15th centile were proportionately more likely to be less educated, black non-Hispanic, non-officers, active duty, serving in the army, and to have an occupational position other than combat, health care or functional support (such as technical or electrical repair position). They were also proportionately more likely to have a positive CAGE result, report current smoking and have three or more combat exposures. The mean physical component summary score was 41.2 for those below the 15th centile, 55.4 for those in the 15th to 85th centile, and 61.3 for those in the highest centile group.
Table 1 Demographic characteristics of 5410 combat deployed* participants in Millennium Cohort Study by functional status (centile of component score), 2001-6. Figures are numbers (rounded percentage) of participants
Table 2 shows baseline characteristics by percentage of new onset symptoms or diagnosis of PTSD, using both the specific and sensitive criteria. The table also shows odds ratios and 95% confidence intervals representing the odds of new onset symptoms or diagnosis of PTSD. With the specific criteria of the DSM-IV, new onset PTSD was identified in 7.3% (395) of the study group with a sum of 50 points or a physician diagnosis. Using the sensitive criteria of the DSM-IV alone or a physician diagnosis, 8.6% (457) were identified with new onset PTSD. After adjustment for all other variables in the model, individuals whose baseline mental component summary score was below the 15th centile were over three times as likely to have new-onset symptoms or diagnosis of PTSD (odds ratio 3.51, 95% confidence interval 2.74 to 4.50, specific; 3.18, 2.50 to 4.05, sensitive). Individuals whose baseline physical component summary score was below the 15th centile were over twice as likely to have new onset symptoms or diagnosis of PTSD (2.22, 1.71 to 2.89, specific; 2.11, 1.64 to 2.70, sensitive). Participants who reported three or more combat exposures during their deployment were also more than twice as likely to have new onset symptoms or diagnosis of PTSD by follow-up (2.60, 1.89 to 3.58, specific; 2.32, 1.74 to 3.08, sensitive). Additionally, women (2.33, 1.80 to 3.03), current smokers (1.56, 1.21 to 2.01), non-officers (1.77, 1.11 to 2.85), Reserve/National Guard (1.85, 1.47 to 2.33), and army (2.04, 1.39 to 3.02) members had increased odds for new onset symptoms or diagnosis of PTSD (measures are based on sensitive criteria). Participants of other races or ethnicities (excluding black non-Hispanic) were 1.57 times more likely to have new onset symptoms or diagnosis of PTSD by follow-up compared with white non-Hispanics (1.18 to 2.08, sensitive criteria).
Table 2 Percentages and adjusted odds ratios (95% confidence intervals) of new onset PTSD among combat deployed* participants of Millennium Cohort according to specific and sensitive criteria, 2001-6