Abstract
Cigarette smoking and musculoskeletal pain are prevalent among Department of Veterans Affairs (VA) health care system users. These conditions frequently co-occur; however, there is limited empirical information specific to Afghanistan/Iraq era veterans. The present study sought to examine gender differences in the association between cigarette smoking and moderate to severe musculoskeletal pain in US veterans with Afghanistan/Iraq era service.
A random sample of 5,000 veterans with service after November 11, 2001, participated in a survey assessing health care needs and barriers to care. One thousand ninety veterans completed the survey assessing post-traumatic stress disorder (PTSD) symptoms, depressive symptoms, and current pain severity. Multivariate logistic regression was used to examine the association between gender, cigarette smoking status, and current moderate to severe musculoskeletal pain.
Findings indicated a significant gender by smoking interaction on moderate/severe musculoskeletal pain, adjusting for age, self-reported race/ethnicity and weight status, combat exposure, probable PTSD, depressive symptoms, service-connected injury during deployment, and VA health care service utilization. Deconstruction of the interaction indicated that female veteran smokers, relative to female nonsmokers, had increased odds of endorsing moderate to severe musculoskeletal pain (odds ratio [OR] = 2.73, 95% confidence interval [CI] = 1.16–6.41), whereas this difference was nonsignificant for male veterans (OR = 1.03, 95% CI = 0.69–1.56).
Survey data from Operation Enduring Freedom/Operation Iraqi Freedom veterans suggest an association between current smoking, gender, and moderate to severe musculoskeletal pain. The stronger relationship between smoking and pain in women supports the need for interventional and longitudinal research that can inform gender-based risk factors for pain in veteran cigarette smokers.
Introduction
Lifetime and current cigarette smoking are highly prevalent among Afghanistan and Iraq era veterans [1,2]. Although nicotine has analgesic properties [3,4], epidemiological studies among civilians suggest that cigarette smoking is a risk factor for the development of chronic pain [5,6]. Several prospective cohort studies of civilians suggest that smoking during adolescence and early adulthood is associated with the onset of moderate to severe pain in later adulthood [7–9]. Further, among persons with chronic pain, cigarette smoking is associated with increased pain intensity and functional impairment [10,11]. While studies conducted in civilian samples have documented associations between pain and cigarette smoking, relatively little research has examined whether this relation differs among veterans of the recent conflicts.
Veterans are at increased risk for persistent pain, and pain complaints are common even among recent veterans. Among Afghanistan and Iraq era veterans who seek clinical care, 28–63% report moderate to severe musculoskeletal or chronic pain [12–14]. Comparisons of male and female veterans on pain reporting have been widely mixed. Two studies examining veterans Health Administration (VHA) medical record data of Afghanistan and Iraq era veterans have documented that women have increased rates of musculoskeletal disorders [15] and persistent pain [16] whereas an earlier study of Afghanistan and Iraq era veterans suggested that women were less likely to report the presence of any pain or persistent pain [13,17–19]. Regardless, among those reporting pain, women were more likely to report moderate/severe pain compared with their male counterparts [13]. This topic is of particular importance given the projected increase in the utilization of health care services expected over the next decade. More work is needed to clarify the association between smoking, pain, and gender among veterans.
Although the prevalence of cigarette smoking is generally lower among women than men, women veterans have a higher risk for smoking compared with civilian women [20]. Additionally, female smokers compared with male smokers have higher levels of nicotine dependence [21], increased difficulty quitting smoking [22], and more severe health consequences of smoking [23]. Volkman and colleagues [24] recently evaluated gender-specific effects of cigarette smoking on pain intensity in Afghanistan and Iraq era veterans using VHA health care. Results suggested that both smoking and female gender were independently associated with higher odds of reporting moderate to severe pain [24].
The mechanisms underlying the association between pain and cigarette smoking are likely complex and involve alterations in pain processing, smoking-related physiological damage, both negative reinforcement (e.g., self-medication) and positive reinforcement (e.g., smoking to increase energy/arousal) of smoking behavior, and psychosocial factors (for reviews, see [3,12]) It has been theorized that a mutual maintenance model may explain the relationship between pain and smoking in which pain and smoking exacerbate each other in a destructive positive feedback loop [10,25]. Risk factors for pain among veterans include race, age, obesity, and mental health disorders including PTSD and depression [16,26,27].
The primary goal of the current study is to further examine the association between pain, smoking, and gender among Afghanistan/Iraq era veterans. It was predicted that female gender and current and lifetime cigarette smoking would be related to reports of musculoskeletal pain after accounting for other known risk factors.
Methods
Participants and Procedures
The present study (N = 1,090) is a secondary analysis of a larger survey conducted as part of the Operation Enduring Freedom/Operation Iraqi Freedom (OEF/OIF) Veterans Health and Needs Study, which was designed to assess the health care needs, preferences, and barriers to care for veterans returning from Iraq and Afghanistan [28–31]. The parent study identified a random sample of 5,000 Afghanistan/Iraq era veterans with a last known address in the VA mid-Atlantic region catchment area through a data use agreement with the VA Environmental Epidemiology Service. Criteria for study participation included being eligible for VA health care and having a valid US mailing address. Basic eligibility for VA health care was defined as having former active duty military service with a favorable discharge or federally activated military service as a reservist or national guardsmen.
Of the 5,000 veterans identified in the parent study, 72 (1.4%) were determined to be ineligible (e.g., deceased, deployed) and 924 (18.5%) surveys were undeliverable (returned to sender). Of the 4,004 surveys that were delivered, 1,161 were completed and returned, resulting in a response rate of 29%, which is consistent with other published mail surveys involving Afghanistan/Iraq era veteran samples (e.g., 22–33%) [32,33]. A modified Dillman procedure was utilized [34] in which all participants received a pre-alert letter, the survey, and, if needed, a follow-up letter and duplicate survey. The survey package included a cover letter, a 60-question survey, and a postage-paid business reply return envelope.
As previously reported [28,29,31], demographic characteristics were compared between early responders (i.e., responders to the first survey wave; N = 978) and later responders (second wave; N = 183) in order to assess possible nonresponse bias following the continuum of the resistance model [35,36]. These analyses revealed little difference between early and late responders [29]. Respondents to the first wave were slightly older (M = 39 years) than later respondents (M = 38 years) and were more likely to be married (73% vs 63%). There were no differences in the proportion of women, minorities, enlisted soldiers, veterans who served in the Reserves/National Guard, veterans who screened positive for PTSD or depression, the proportion of participants who suffered a service-connected injury, or the proportion of veterans who used VA health care.
Measures
Demographic Characteristics and Military History
Participants completed multiple survey items assessing demographic characteristics, including age, gender, race, and marital status. Participants were asked multiple questions about their military careers, including whether they suffered an injury during their deployment that resulted in a service-connected disability rating by VA.
Pain
Pain was measured through single-item questions assessing the presence or absence of seven pain-related symptoms during the last month. Pain symptoms included stomach pain, back pain, joint pain, muscle aches or cramps, pain during sexual intercourse, headaches, and chest pain. Participants indicated the presence of current pain symptoms on a three-point Likert scale indicating no, mild, or severe pain. Mild pain was defined as “just aware, not slowed down by the symptoms, or sufficient to take nonprescription drugs to relieve symptoms (aspirin, Tums, etc.).” Severe pain was characterized as “sufficient to seek medical advice, take nonprescription drugs, miss work, or limit routine activities.”
A visual analog scale (VAS) ranging from 0–10 was used to assess the severity of musculoskeletal pain. Participants were asked, “Please rate the severity of your muscle and joint pain AT THIS MOMENT on the following scale.” Zero indicated no pain, and 10 indicated extreme pain. We chose to model moderate/severe pain a priori in order to be consistent with prior work in this area [37]. Scores of 4 or greater indicated moderate to severe pain [38–40]. Note, this cutoff is consistent with previous recommendations considering the differential impact of pain levels on the VAS pain severity measure.
Cigarette Smoking
Smoking status was assessed via a single question item in which respondents were asked: “Have you ever smoked cigarettes?” Response choices included: “yes, still smoking every day”; “yes, still smoking some days”; “yes, but no longer smoke at all”; and “no, never smoked.” The first two responses (“yes, still smoking every day” and “yes, still smoking some days”) were combined to represent current smoking.
Combat Exposure
The Combat Experiences Scale [41] was used to assess combat exposure. Participants indicated whether or not they had experienced 17 different combat situations during their deployment(s). The total number of reported experiences was summed to result in a total score ranging from 0 to 17.
Weight
Self-reported weight was assessed with a single item from the Department of Defense Survey of Health-Related Behaviors [42]: “In thinking about your weight, do you consider yourself to be…?” Response options included overweight, about the right weight, or underweight.
PTSD and Depression
The PTSD Checklist (PCL-M) was administered to assess PTSD symptoms [43]. The PCL is a well-validated self-report measure consisting of 17 items that correspond to the DSM-IV diagnostic symptoms of PTSD [44]. For the current study, a cut-point score of 50 was used to determine probable PTSD. The Patient Health Questionnaire–2 (PHQ-2) [45] was used to screen for the presence of depression. Participants with total scores of three or higher were considered to have screened positive for depression. Criteria and construct validity of scores from the PHQ-2 have been established [46].
VA Health Care Service Use
Use of any health care services was based on a single item, which queried use of VA health care services, use of non-VA health care services, dual use (both VA and non-VA health care), or no health care use since their last deployment. A dichotomous variable of any VA health care use vs no VA health care use was utilized in statistical analyses.
Analyses
Prior to data analysis, missing data were examined for systematic missingness. Missingness was unrelated to any of the study measures or covariates. We chose to drop 41 cases (3.5%) that were missing data on a key predictor or outcome (i.e., gender, pain intensity, or smoking status) or that were missing more than three items on the PCL or more than one of the PHQ-2 (30 cases, 2.5%), resulting in an analysis sample of 1,090 veterans. Missing data on other covariates were imputed using multiple imputation using the Monte Carlo Markov chain method with 10 iterations. Consistent with previous research in this area, the primary dependent variable was moderate to severe musculoskeletal pain (4 or more on the VAS) [22,30]. Logistic regression was used to evaluate both the unadjusted and adjusted associations between smoking, gender, and moderate to severe musculoskeletal pain. To explore whether the association between smoking and pain was moderated by gender, a gender by smoking interaction term was included in adjusted models. Adjusted models controlled for age, gender, self-reported race, VA health care use, presence of service-connected injury during deployment, severity of combat exposure, probable PTSD, probable depression, and self-reported weight. All analyses were conducted using SAS software version 9.4.
Results
Demographic characteristics are presented in Table 1. The mean age of the sample was 38 years. Compared with male veterans, female veterans were younger, more likely to identify as a minority race, less likely to be married, and reported exposure to fewer combat experiences during service (see Table 1). Findings that women were younger, less white, and less married are consistent with the demographics of female veterans currently in the military [47]. The proportion of veterans using VA health care (36%) did not differ significantly by gender. Rates of probable PTSD and probable depression were similar between men and women. Lifetime smoking (46%) was prevalent in the sample and did not differ by gender. Eighteen percent of the sample identified as current smokers.
Sample characteristics, gender (N = 1,090)
| Variable | Total sample | Men (N = 901) | Women (N = 189) | Test statistic |
|---|---|---|---|---|
| Age, M (SD) | 38.85 (9.83) | 39.42 (9.80) | 36.13 (9.53) | t = −4.22** |
| Race, No. (%) | χ2 = 19.20** | |||
| White | 775 (71) | 662 (73) | 113 (60) | |
| Black or African American | 182 (17) | 140 (16) | 42 (23) | |
| Other | 92 (9) | 64 (7) | 28 (15) | |
| Married or living as married, No. (%) | 766 (71) | 668 (75) | 98 (52) | χ2 = 37.29** |
| Military duty status†, No. (%) | χ2 = 0.62 | |||
| Reserves or National Guard | 426 (39) | 355 (39) | 71 (36) | |
| Active duty | 664 (61) | 546 (60) | 118 (62) | |
| Combat exposure, M (SD) | 4.05 (4.03) | 4.33 (4.18) | 2.79 (2.97) | t = −4.84** |
| SC injury during deployment, No. (%) | 392 (37) | 329 (37) | 63 (34) | χ2 = 0.61 |
| Probable depression, No. (%) | 178 (16) | 142 (16) | 36 (19) | χ2 = 0.70 |
| Probable PTSD, No. (%) | 149 (14) | 125 (14) | 24 (13) | χ2 = 0.34 |
| Smoking status, No. (%) | χ2 = 0.51 | |||
| Current smoker | 194 (18) | 157 (18) | 37 (20) | |
| Former smoker | 306 (28) | 259 (29) | 47 (25) | |
| Never smoker | 590 (54) | 485 (54) | 105 (56) | |
| Use of VA health care service, No. (%) | 387 (37) | 321 (37) | 66 (37) | χ2 = 0.0002 |
| Generalized muscle aching or cramps | ||||
| Any | 531 (50) | 426 (49) | 105 (56) | χ2 = 3.08 |
| Severe | 85 (8) | 61 (7) | 24 (13) | χ2 = 7.00* |
| Arm, leg, or joint pain | ||||
| Any | 720 (67) | 600 (68) | 120 (64) | χ2 = 0.29 |
| Severe | 209 (19) | 174 (20) | 35 (19) | χ2 = 0.74 |
| Back pain | ||||
| Any | 766 (71) | 629 (71) | 137 (72) | χ2 = 0.68 |
| Severe | 247 (23) | 200 (23) | 47 (25) | χ2 = 0.50 |
| Headaches | ||||
| Any | 628 (59) | 498 (57) | 130 (69) | χ2 = 9/88* |
| Severe | 141 (13) | 101 (12) | 40 (21) | χ2 = 12.89 |
| Pain during sexual intercourse | ||||
| Any | 152(14) | 101(12) | 51(27) | χ2 = 29.57** |
| Severe | 34 (3) | 26 (3) | 8 (4) | χ2 = 0.35 |
| Pain severity | ||||
| None (0 vs any pain) | 387 (35) | 308 (34) | 79 (42) | χ2 = 3.96 |
| Moderate to severe pain | 329 (30) | 280 (31) | 49 (26) | χ2 = 0.22 |
| Pain score for those with any pain, M (SD) | 3.73 (2.17) | 3.73 (2.16) | 3.71 (2.28) | t = −1.48 |
| Variable | Total sample | Men (N = 901) | Women (N = 189) | Test statistic |
|---|---|---|---|---|
| Age, M (SD) | 38.85 (9.83) | 39.42 (9.80) | 36.13 (9.53) | t = −4.22** |
| Race, No. (%) | χ2 = 19.20** | |||
| White | 775 (71) | 662 (73) | 113 (60) | |
| Black or African American | 182 (17) | 140 (16) | 42 (23) | |
| Other | 92 (9) | 64 (7) | 28 (15) | |
| Married or living as married, No. (%) | 766 (71) | 668 (75) | 98 (52) | χ2 = 37.29** |
| Military duty status†, No. (%) | χ2 = 0.62 | |||
| Reserves or National Guard | 426 (39) | 355 (39) | 71 (36) | |
| Active duty | 664 (61) | 546 (60) | 118 (62) | |
| Combat exposure, M (SD) | 4.05 (4.03) | 4.33 (4.18) | 2.79 (2.97) | t = −4.84** |
| SC injury during deployment, No. (%) | 392 (37) | 329 (37) | 63 (34) | χ2 = 0.61 |
| Probable depression, No. (%) | 178 (16) | 142 (16) | 36 (19) | χ2 = 0.70 |
| Probable PTSD, No. (%) | 149 (14) | 125 (14) | 24 (13) | χ2 = 0.34 |
| Smoking status, No. (%) | χ2 = 0.51 | |||
| Current smoker | 194 (18) | 157 (18) | 37 (20) | |
| Former smoker | 306 (28) | 259 (29) | 47 (25) | |
| Never smoker | 590 (54) | 485 (54) | 105 (56) | |
| Use of VA health care service, No. (%) | 387 (37) | 321 (37) | 66 (37) | χ2 = 0.0002 |
| Generalized muscle aching or cramps | ||||
| Any | 531 (50) | 426 (49) | 105 (56) | χ2 = 3.08 |
| Severe | 85 (8) | 61 (7) | 24 (13) | χ2 = 7.00* |
| Arm, leg, or joint pain | ||||
| Any | 720 (67) | 600 (68) | 120 (64) | χ2 = 0.29 |
| Severe | 209 (19) | 174 (20) | 35 (19) | χ2 = 0.74 |
| Back pain | ||||
| Any | 766 (71) | 629 (71) | 137 (72) | χ2 = 0.68 |
| Severe | 247 (23) | 200 (23) | 47 (25) | χ2 = 0.50 |
| Headaches | ||||
| Any | 628 (59) | 498 (57) | 130 (69) | χ2 = 9/88* |
| Severe | 141 (13) | 101 (12) | 40 (21) | χ2 = 12.89 |
| Pain during sexual intercourse | ||||
| Any | 152(14) | 101(12) | 51(27) | χ2 = 29.57** |
| Severe | 34 (3) | 26 (3) | 8 (4) | χ2 = 0.35 |
| Pain severity | ||||
| None (0 vs any pain) | 387 (35) | 308 (34) | 79 (42) | χ2 = 3.96 |
| Moderate to severe pain | 329 (30) | 280 (31) | 49 (26) | χ2 = 0.22 |
| Pain score for those with any pain, M (SD) | 3.73 (2.17) | 3.73 (2.16) | 3.71 (2.28) | t = −1.48 |
PTSD = post-traumatic stress disorder; SC = service-connected injury; VA= Veterans Affairs.
P < 0.05.
P < 0.001.
Represents military duty status during OEF/OIF deployment.
Sample characteristics, gender (N = 1,090)
| Variable | Total sample | Men (N = 901) | Women (N = 189) | Test statistic |
|---|---|---|---|---|
| Age, M (SD) | 38.85 (9.83) | 39.42 (9.80) | 36.13 (9.53) | t = −4.22** |
| Race, No. (%) | χ2 = 19.20** | |||
| White | 775 (71) | 662 (73) | 113 (60) | |
| Black or African American | 182 (17) | 140 (16) | 42 (23) | |
| Other | 92 (9) | 64 (7) | 28 (15) | |
| Married or living as married, No. (%) | 766 (71) | 668 (75) | 98 (52) | χ2 = 37.29** |
| Military duty status†, No. (%) | χ2 = 0.62 | |||
| Reserves or National Guard | 426 (39) | 355 (39) | 71 (36) | |
| Active duty | 664 (61) | 546 (60) | 118 (62) | |
| Combat exposure, M (SD) | 4.05 (4.03) | 4.33 (4.18) | 2.79 (2.97) | t = −4.84** |
| SC injury during deployment, No. (%) | 392 (37) | 329 (37) | 63 (34) | χ2 = 0.61 |
| Probable depression, No. (%) | 178 (16) | 142 (16) | 36 (19) | χ2 = 0.70 |
| Probable PTSD, No. (%) | 149 (14) | 125 (14) | 24 (13) | χ2 = 0.34 |
| Smoking status, No. (%) | χ2 = 0.51 | |||
| Current smoker | 194 (18) | 157 (18) | 37 (20) | |
| Former smoker | 306 (28) | 259 (29) | 47 (25) | |
| Never smoker | 590 (54) | 485 (54) | 105 (56) | |
| Use of VA health care service, No. (%) | 387 (37) | 321 (37) | 66 (37) | χ2 = 0.0002 |
| Generalized muscle aching or cramps | ||||
| Any | 531 (50) | 426 (49) | 105 (56) | χ2 = 3.08 |
| Severe | 85 (8) | 61 (7) | 24 (13) | χ2 = 7.00* |
| Arm, leg, or joint pain | ||||
| Any | 720 (67) | 600 (68) | 120 (64) | χ2 = 0.29 |
| Severe | 209 (19) | 174 (20) | 35 (19) | χ2 = 0.74 |
| Back pain | ||||
| Any | 766 (71) | 629 (71) | 137 (72) | χ2 = 0.68 |
| Severe | 247 (23) | 200 (23) | 47 (25) | χ2 = 0.50 |
| Headaches | ||||
| Any | 628 (59) | 498 (57) | 130 (69) | χ2 = 9/88* |
| Severe | 141 (13) | 101 (12) | 40 (21) | χ2 = 12.89 |
| Pain during sexual intercourse | ||||
| Any | 152(14) | 101(12) | 51(27) | χ2 = 29.57** |
| Severe | 34 (3) | 26 (3) | 8 (4) | χ2 = 0.35 |
| Pain severity | ||||
| None (0 vs any pain) | 387 (35) | 308 (34) | 79 (42) | χ2 = 3.96 |
| Moderate to severe pain | 329 (30) | 280 (31) | 49 (26) | χ2 = 0.22 |
| Pain score for those with any pain, M (SD) | 3.73 (2.17) | 3.73 (2.16) | 3.71 (2.28) | t = −1.48 |
| Variable | Total sample | Men (N = 901) | Women (N = 189) | Test statistic |
|---|---|---|---|---|
| Age, M (SD) | 38.85 (9.83) | 39.42 (9.80) | 36.13 (9.53) | t = −4.22** |
| Race, No. (%) | χ2 = 19.20** | |||
| White | 775 (71) | 662 (73) | 113 (60) | |
| Black or African American | 182 (17) | 140 (16) | 42 (23) | |
| Other | 92 (9) | 64 (7) | 28 (15) | |
| Married or living as married, No. (%) | 766 (71) | 668 (75) | 98 (52) | χ2 = 37.29** |
| Military duty status†, No. (%) | χ2 = 0.62 | |||
| Reserves or National Guard | 426 (39) | 355 (39) | 71 (36) | |
| Active duty | 664 (61) | 546 (60) | 118 (62) | |
| Combat exposure, M (SD) | 4.05 (4.03) | 4.33 (4.18) | 2.79 (2.97) | t = −4.84** |
| SC injury during deployment, No. (%) | 392 (37) | 329 (37) | 63 (34) | χ2 = 0.61 |
| Probable depression, No. (%) | 178 (16) | 142 (16) | 36 (19) | χ2 = 0.70 |
| Probable PTSD, No. (%) | 149 (14) | 125 (14) | 24 (13) | χ2 = 0.34 |
| Smoking status, No. (%) | χ2 = 0.51 | |||
| Current smoker | 194 (18) | 157 (18) | 37 (20) | |
| Former smoker | 306 (28) | 259 (29) | 47 (25) | |
| Never smoker | 590 (54) | 485 (54) | 105 (56) | |
| Use of VA health care service, No. (%) | 387 (37) | 321 (37) | 66 (37) | χ2 = 0.0002 |
| Generalized muscle aching or cramps | ||||
| Any | 531 (50) | 426 (49) | 105 (56) | χ2 = 3.08 |
| Severe | 85 (8) | 61 (7) | 24 (13) | χ2 = 7.00* |
| Arm, leg, or joint pain | ||||
| Any | 720 (67) | 600 (68) | 120 (64) | χ2 = 0.29 |
| Severe | 209 (19) | 174 (20) | 35 (19) | χ2 = 0.74 |
| Back pain | ||||
| Any | 766 (71) | 629 (71) | 137 (72) | χ2 = 0.68 |
| Severe | 247 (23) | 200 (23) | 47 (25) | χ2 = 0.50 |
| Headaches | ||||
| Any | 628 (59) | 498 (57) | 130 (69) | χ2 = 9/88* |
| Severe | 141 (13) | 101 (12) | 40 (21) | χ2 = 12.89 |
| Pain during sexual intercourse | ||||
| Any | 152(14) | 101(12) | 51(27) | χ2 = 29.57** |
| Severe | 34 (3) | 26 (3) | 8 (4) | χ2 = 0.35 |
| Pain severity | ||||
| None (0 vs any pain) | 387 (35) | 308 (34) | 79 (42) | χ2 = 3.96 |
| Moderate to severe pain | 329 (30) | 280 (31) | 49 (26) | χ2 = 0.22 |
| Pain score for those with any pain, M (SD) | 3.73 (2.17) | 3.73 (2.16) | 3.71 (2.28) | t = −1.48 |
PTSD = post-traumatic stress disorder; SC = service-connected injury; VA= Veterans Affairs.
P < 0.05.
P < 0.001.
Represents military duty status during OEF/OIF deployment.
Female veterans were more likely to report severe headaches, severe generalized muscle aching or cramps, and pain during sexual intercourse. There were no differences between male and female veterans in the proportion reporting back or joint pain. Examining the severity of musculoskeletal pain, females were less likely to report having any musculoskeletal pain (58% vs 66%, χ2 (1df) = 3.96, P < 0.05) but failed to differ in the proportion reporting moderate to severe pain (see Table 1).
In unadjusted analyses, veterans with moderate to severe pain intensity were more likely to be of minority race, have suffered a service-connected injury during service, experienced more combat exposure, and report utilizing VA health care. Moderate to severe pain intensity was also positively associated with screening positive for PTSD and depression and describing oneself as overweight.
After controlling for demographic and other risk factors including probable PTSD and service connected disability, results of adjusted multivariate logistic regression indicated that gender and current smoking status were significantly associated with increased odds of moderate to severe musculoskeletal pain (see Table 2). These main effects, however, were qualified by a significant interaction between gender and smoking. As shown in Figure 1, decomposition of the interaction revealed that male veteran nonsmokers were more likely than female veteran nonsmokers to report moderate to severe pain (odds ratio [OR] = 1.68, 95% confidence interval [CI] = 1.06–2.66); however, there were no gender differences in the endorsement of moderate to severe pain among veteran smokers (OR = 0.65, 95% CI = 0.28–1.48). Further, female veteran smokers, relative to female nonsmokers, had increased odds of endorsing moderate to severe musculoskeletal pain (OR = 2.73, 95% CI = 1.16–6.41), whereas this difference failed to be significant among male veterans (OR = 1.03, 95% CI = 0.69–1.56).
Logistic regression model: Variables associated with moderate-severe muscle and/or joint pain (N = 1,090)
| Unadjusted | Adjusted† | |||
|---|---|---|---|---|
| Variable | OR | (95% CI) | OR | (95% CI) |
| Age (>30 y) | 0.97 | (0.72–1.31) | 1.06 | (0.75–1.50) |
| Male gender‡ | 1.29 | (0.90–1.84) | 1.68* | (1.06–2.66) |
| Minority Race§ | 1.49** | (1.11–1.99) | 1.40 | (1.00–1.95) |
| Use of VA health care | 2.74** | (2.08–3.62) | 1.59** | (1.16–2.18) |
| Service-connected injury | 4.16** | (3.16–5.48) | 2.81** | (2.05–3.83) |
| Combat exposure | 1.10** | (1.06–1.13) | 1.01 | (0.97–1.05) |
| Overweight | 1.34* | (1.03–1.74) | 1.23 | (0.92–1.66) |
| Probable PTSD | 6.72** | (4.63–9.74) | 2.93** | (1.81–4.76) |
| Probable depression | 3.88** | (2.80–5.38) | 1.53 | (1.00–2.36) |
| Current smoking | 1.45* | (1.06–2.03) | 2.71* | (1.15–6.37) |
| Current smoking × gender‡ | – | – | 0.39* | (0.15–0.99) |
| Unadjusted | Adjusted† | |||
|---|---|---|---|---|
| Variable | OR | (95% CI) | OR | (95% CI) |
| Age (>30 y) | 0.97 | (0.72–1.31) | 1.06 | (0.75–1.50) |
| Male gender‡ | 1.29 | (0.90–1.84) | 1.68* | (1.06–2.66) |
| Minority Race§ | 1.49** | (1.11–1.99) | 1.40 | (1.00–1.95) |
| Use of VA health care | 2.74** | (2.08–3.62) | 1.59** | (1.16–2.18) |
| Service-connected injury | 4.16** | (3.16–5.48) | 2.81** | (2.05–3.83) |
| Combat exposure | 1.10** | (1.06–1.13) | 1.01 | (0.97–1.05) |
| Overweight | 1.34* | (1.03–1.74) | 1.23 | (0.92–1.66) |
| Probable PTSD | 6.72** | (4.63–9.74) | 2.93** | (1.81–4.76) |
| Probable depression | 3.88** | (2.80–5.38) | 1.53 | (1.00–2.36) |
| Current smoking | 1.45* | (1.06–2.03) | 2.71* | (1.15–6.37) |
| Current smoking × gender‡ | – | – | 0.39* | (0.15–0.99) |
CI = confidence interval; OR = odds ratio; PTSD = post-traumatic stress disorder; VA= Veterans Affairs.
P < 0.05.
P < 0.01.
Adjusted for age, gender, race, use of VA health care, service-connected injury, combat exposure, self-reported weight, PTSD, depression, and smoking status.
Females comprised the reference group.
Race was modeled as a dichotomous variable (white vs other).
Logistic regression model: Variables associated with moderate-severe muscle and/or joint pain (N = 1,090)
| Unadjusted | Adjusted† | |||
|---|---|---|---|---|
| Variable | OR | (95% CI) | OR | (95% CI) |
| Age (>30 y) | 0.97 | (0.72–1.31) | 1.06 | (0.75–1.50) |
| Male gender‡ | 1.29 | (0.90–1.84) | 1.68* | (1.06–2.66) |
| Minority Race§ | 1.49** | (1.11–1.99) | 1.40 | (1.00–1.95) |
| Use of VA health care | 2.74** | (2.08–3.62) | 1.59** | (1.16–2.18) |
| Service-connected injury | 4.16** | (3.16–5.48) | 2.81** | (2.05–3.83) |
| Combat exposure | 1.10** | (1.06–1.13) | 1.01 | (0.97–1.05) |
| Overweight | 1.34* | (1.03–1.74) | 1.23 | (0.92–1.66) |
| Probable PTSD | 6.72** | (4.63–9.74) | 2.93** | (1.81–4.76) |
| Probable depression | 3.88** | (2.80–5.38) | 1.53 | (1.00–2.36) |
| Current smoking | 1.45* | (1.06–2.03) | 2.71* | (1.15–6.37) |
| Current smoking × gender‡ | – | – | 0.39* | (0.15–0.99) |
| Unadjusted | Adjusted† | |||
|---|---|---|---|---|
| Variable | OR | (95% CI) | OR | (95% CI) |
| Age (>30 y) | 0.97 | (0.72–1.31) | 1.06 | (0.75–1.50) |
| Male gender‡ | 1.29 | (0.90–1.84) | 1.68* | (1.06–2.66) |
| Minority Race§ | 1.49** | (1.11–1.99) | 1.40 | (1.00–1.95) |
| Use of VA health care | 2.74** | (2.08–3.62) | 1.59** | (1.16–2.18) |
| Service-connected injury | 4.16** | (3.16–5.48) | 2.81** | (2.05–3.83) |
| Combat exposure | 1.10** | (1.06–1.13) | 1.01 | (0.97–1.05) |
| Overweight | 1.34* | (1.03–1.74) | 1.23 | (0.92–1.66) |
| Probable PTSD | 6.72** | (4.63–9.74) | 2.93** | (1.81–4.76) |
| Probable depression | 3.88** | (2.80–5.38) | 1.53 | (1.00–2.36) |
| Current smoking | 1.45* | (1.06–2.03) | 2.71* | (1.15–6.37) |
| Current smoking × gender‡ | – | – | 0.39* | (0.15–0.99) |
CI = confidence interval; OR = odds ratio; PTSD = post-traumatic stress disorder; VA= Veterans Affairs.
P < 0.05.
P < 0.01.
Adjusted for age, gender, race, use of VA health care, service-connected injury, combat exposure, self-reported weight, PTSD, depression, and smoking status.
Females comprised the reference group.
Race was modeled as a dichotomous variable (white vs other).
Modeled probability of reporting moderate to severe musculoskeletal pain.
In a similar model in which we substituted lifetime smoking for current smoking, the Gender X Smoking interaction remained significant (P = 0.049). Male veterans who never smoked were more likely than female veterans who never smoked to report moderate to severe pain (OR = 2.00, 95% CI = 1.12–3.56); however, there were no gender differences in pain reporting among veterans with any smoking history (OR = 0.90, 95% CI = 0.51–1.57). Female veterans were more likely to report moderate to severe pain if they were lifetime smokers (OR = 2.30, 95% CI = 1.11–4.78). Male veterans were not significantly more likely to report moderate to severe pain if they were lifetime smokers (OR = 1.04, 95% CI = 0.75–1.43).
Discussion
This study examined cross-sectional associations between cigarette smoking, gender, and musculoskeletal pain in a large random sample of Afghanistan and Iraq era veterans in the Mid-Atlantic region of the United States and included veterans who were receiving care from the VHA health care system and those who were not. Consistent with prior research examining pain and smoking [12], both current and lifetime smoking were associated with increased reports of moderate to severe musculoskeletal pain. This effect, however, was moderated by gender, where smoking was more strongly related to increased pain among women. Findings from the present study built upon past research in this area by providing evidence that gender moderates the association between smoking status and musculoskeletal pain.
Similar to previous findings among Afghanistan and Iraq era veterans, almost a third of the current sample (30%) reported moderate to severe pain [24]. Our results indicated that for women, but not men, smoking was linked to moderate to severe pain. This result is in contrast to a recently reported finding that failed to observe gender moderation of the effect of smoking on pain severity among Afghanistan and Iraq era VHA users [24]. However, the discrepancy between study results may be accounted for by sampling; the Volkman and colleagues [24] study focused on VHA health care users only whereas the present study examined nonusers of VHA health care as well.
Consistent with current study findings, civilian sample studies have reported gender differences in the association between smoking and pain. In a sample of older adults, smoking was related to pain intensity for both men and women, but current smoking was only related to pain prevalence among women [48]. However, that study conducted separate analyses by gender and unfortunately did not include a test of gender moderation. In an earlier study, smoking status was associated with increased pain intensity for women but not men though no formal testing of gender modification was conducted [49].
Regarding potential explanations for gender moderation of the smoking-pain relationship, sex differences in smoking-related analgesia, sex hormones, pain perception, and socialization may account for this association. Though there is evidence of smoking-related analgesia in both women and men [3], there is mixed evidence of sex-related analgesic effects. Gender moderates the effect of smoking status on pain sensitivity, such that smoking status is associated with decreased pain sensitivity to ischemic pain only for female smokers and cold-pressor pain for male smokers [4]. Decreased pain sensitivity may reflect longstanding patterns of negative reinforcement, where smoking behavior is rewarded by a decrease in pain over time [3]. Given that sex differences in response to classical and operant conditioning can be partially attributed to sex hormones [50], the extent to which the analgesic effects of nicotine reinforce the link between smoking and pain may differ by gender. Additionally, the relationship between gender, smoking, and pain may also relate to psychiatric comorbidity as both smoking and pain are more strongly related to psychological symptoms in women, relative to men [51,52]. Though multidisciplinary psychological approaches (e.g., cognitive behavioral therapy) [50] to pain management already include coping with stress, results from the current study suggest that it might also be fruitful to target smoking cessation as a pain management and health promotion technique within the context of behavioral/multidisciplinary pain management. In a recent, large study of women, the strength of the relationship of persistent back pain to stress was similar to the strength of the relationship between back pain and smoking [53]. For women with chronic pain, smoking abstinence and stress reduction both provide pain-relevant health promotion goals that are amenable to multidisciplinary intervention efforts.
There are limitations associated with this study. Results are based on a cross-sectional survey, which does not permit any causal conclusions. The response rate (29%) is relatively low, which may affect generalizability to all Afghanistan/Iraq era veterans. However, this response rate is consistent with other population-based mail surveys of Afghanistan and Iraq era veterans (e.g., 21.9–33%) [11,51,52]. Concerns about the potential representativeness of the sample are tempered by the consistency of demographic variables and rates of probable PTSD and pain with other large studies of Afghanistan/Iraq era veterans [24,29]. Musculoskeletal pain severity was assessed with a single item, and future research would benefit from detailed, multimethod assessment of pain and pain-related functional impairment.
Despite the limitations, the current study adds to the growing knowledge that smoking is significantly associated with increased pain, particularly among women. There is a clear need to reduce smoking among veterans both male and female. Although specialty clinic–based tobacco cessation programs have been shown to be efficacious in reducing smoking [55,56], such programs are infrequently attended [57–59]. Leveraging electronic medical records to identify smokers and implementing proactive methods to offer smoking cessation interventions that are consistent with clinical practice guidelines may be an ideal strategy to reduce smoking in VA and other large health care organizations. Novel interventions that increase the reach of intensive smoking cessation treatment are needed to maximize quit rates. Pain management approaches that incorporate smoking cessation might be particularly worthy of research and clinical attention in female veterans. Additionally, education about pain and its relationship with cigarette smoking should be integrated into smoking cessation interventions as a way to further encourage abstinence among veterans. Clinicians working with veterans who have chronic pain might prioritize smoking cessation as a modifiable short-term behavioral goal for their patients.
Funding sources: This work was supported in part by the VA Mid-Atlantic Mental Illness Research, Education, and Clinical Center (MIRECC) and by the Department of Veterans Affairs (VA) Office of Research and Development (ORD) Health Services Research and Development Service (HSR&D; I01HX001109) and the National Institutes of Health (R01CA196304). Dr. Dedert was supported by a Career Development Award (IK2CX000718) from the Clinical Science Research and Development Service (CSR&D) of the VA ORD. Dr. Beckham was supported by a VA Research Career Scientist Award (11SRCS-009) from the CSR&D of the VA ORD.
Disclosure and conflicts of interest: Funding sources had no role in the design, execution, analysis, or interpretation of the data, or the decision to submit results for publication. The content is solely the responsibility of the authors and does not necessarily represent the official views of the VA or the US government, or any of the institutions with which the authors are affiliated. The authors have no competing financial interests to report.
