Statistics from Altmetric.com
Editor,—In the United States, smoking rates continue to be higher in male and female active duty military personnel than among their civilian counterparts.1 2 Other countries have similarly reported higher smoking rates among their military personnel.3 Several studies have investigated whether the United States navy “attracts” or “produces” smokers—that is, whether the higher prevalence of smoking among navy active duty personnel is due to self selection of smokers into the navy, or by navy institutional norms that promote smoking.1 4-9 Studies to date have had methodological limitations—for instance, small samples, limited age ranges, use of active duty personnel rather than new enlistees, or no adjustments for differences in civilian–military sociodemographic factors—and results have been conflicting. The present analysis addressed the “attracting” aspect of the question by studying a large sample of women as they entered the navy. The smoking rate of the navy recruit sample was compared with that of a large, representative civilian sample equated to the navy population in terms of age, race/ethnicity, and education using a direct standardisation procedure.10
During a one-year period, all female recruits (n = 5503) completed smoking surveys as they entered the navy. Refusals were virtually nonexistent. Civilian data were obtained from the tobacco use supplement (TUS) to the 1992–1993 United States Bureau of the Census’ current population survey.11 Almost 63 000 unweighted cases were extracted from the TUS for women between the ages of 17 and 35 years to correspond with the complete age range of the navy recruit sample. Software for survey data analysis (SUDAAN)12 was used to weight and standardise the civilian data to the joint distribution of the navy recruit sample in terms of education and race/ethnicity, and then comparisons were made within three age strata. Estimates for civilians can be interpreted as those that would be obtained if the civilian population had had the same sociodemographic distribution as the navy recruit population. For recruits and civilians, current smoking was defined as having smoked 100 cigarettes in one’s life and being an everyday or someday smoker.
The comparison of current smoking between navy women recruits and civilian women, stratified by age category, is presented in the figure. Comparisons for women aged 17–18 years and those aged 19–23 years were statistically significant, with navy women recruits having higher rates of current smoking in both of these age strata. Navy women recruits who were 17–18 years old had over 2½ times the smoking rate of civilians, and women aged 19–23 had over 1½ times the rate of civilians. Smoking rates were not significantly different for recruits and civilians in the 24–35 age range.
Although this analysis cannot rule out the role that the navy environment may play in “producing” smokers, it provides more definitive evidence that the navy attracts young female smokers from the outset. Further, this high rate of smoking cannot be accounted for by sociodemographic characteristics. The specific factors that might account for the high rate of smoking among women entering the navy are not known. Certain “unconventional” personality factors and behaviours including risk taking, sensation seeking, rebelliousness, and self confidence have been associated with smoking in young women,13 14 and perhaps these same characteristics are associated with enlisting in the military. In addition, there may be differences in geographical location and family/peer patterns of tobacco use between women who choose to join the navy and the general population. Additional work is needed to explore reasons for the high rate of smoking among incoming military recruits to inform strategies for effective cessation.
This research was supported by the Department of Defense (DoD) Women’s Health Research Program (DAMD17-95-5075). This project was approved by the Human Use committees of the United States Army Medical Research and Material Command and San Diego State University. The authors gratefully acknowledge the contributions of Suzanne L Hurtado, Linda K Hervig, Shu-Hong Zhu, and Kathleen B Weaver.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.