Alcohol, Tobacco Use in Youth With and Without Chronic Pain
Alcohol, Tobacco Use in Youth With and Without Chronic Pain
Participants included 95 adolescents with mixed chronic pain conditions between the ages of 12 and 18 years (M = 15.60 years, SD = 2.6), and 91 adolescents without chronic pain recruited from the community (M = 15.10 years, SD = 2.0). Sample characteristics are shown in Table I. Participants in both groups were primarily female (71%) and White (80%). Adolescents with and without chronic pain did not differ on age or sex; however, participants without chronic pain were more likely to have parents with a higher income, χ (5, N = 173) = 23.83, p < .0001, and the ethnic composition of the two groups differed, χ (5, N = 181) = 11.63, p < .05, with more White participants in the chronic pain group (83%) compared with the group without chronic pain (76%). Youth with and without chronic pain reported similar levels of loneliness, t(170) = −0.50, p = .62. As expected, youth with chronic pain reported significantly higher depressive symptoms, t(172) = 2.60, p = .01; pain intensity, t(180) = 10.83, p < .0001; and activity limitations, t(158) = 13.52, p < .0001, than youth without chronic pain (Table I).
As hypothesized, adolescents with chronic pain were less likely to drink alcohol as compared with peers without chronic pain (7.4% vs. 22%, respectively), χ (1, N = 175) = 7.94, p < .01 (Table I). However, contrary to our hypothesis, adolescents with chronic pain were as likely to use tobacco compared with their peers without chronic pain (9% vs. 8%, respectively), χ (1, N = 170) = 0.03, p > .05.
Overall, rates of current alcohol and tobacco use were low across groups. Of the 30 youth (16.2%) in the full sample who endorsed current alcohol or tobacco use, 20 youth (10.8%) used either alcohol or tobacco and 10 youth (5.4%) used both. The number of youth who used one versus both substances did not significantly differ between groups, χ (2, N = 170) = 4.71, p = .10.
Logistic regression was used to test whether psychosocial and pain-related factors were associated with current substance use in the full sample while controlling for demographic factors. Group status (pain vs. healthy), demographic covariates (income, race, sex, age), loneliness, depressive symptoms, pain intensity, and activity limitations were entered as predictors in a single step in the logistic regression model. The dependent variable was the dichotomous substance use variable (0 = no tobacco or alcohol use, 1 = positive tobacco and/or alcohol use). As shown in Table II, age, loneliness, depressive symptoms, and activity limitations were associated with substance use, generally in the predicted directions. Older age was associated with a twofold increase in risk for substance use, odds ratio (OR) = 2.24, 95% CI = 1.51–3.34. A one-unit increase in depressive symptoms was associated with a 1.12-fold increase in risk for substance use, 95% CI = 1.05–1.20. A one-unit increase in activity limitations was associated with a decrease in risk for substance use by a factor of 0.91, 95% CI = 0.85–0.99. However, in the unexpected direction, a one-unit increase in loneliness was associated with a 0.94 decrease in risk for substance use, 95% CI = 0.89–0.99. Nagelkerke's R indicated that the final model accounted for 48% of the variability in substance use across groups.
Results from this analysis partially supported our hypothesis. After controlling for group status and demographic factors, higher depressive symptoms and lower activity limitations were associated with a higher likelihood for current substance use. However, contrary to our hypothesis, higher loneliness was associated with a decreased risk for substance use and pain intensity was not associated with substance use.
To explore whether a different pattern of associated factors emerged within each group, separate logistic regression models were conducted for youth with and without chronic pain (Table III). Demographic covariates (income, race, sex, age), loneliness, depressive symptoms, pain intensity, and activity limitations were entered as predictors in a single step in the logistic regression model. The dependent variable was the dichotomous substance use variable (0 = no tobacco or alcohol use, 1 = positive tobacco and/or alcohol use).
Youth Without Chronic Pain. Among youth without chronic pain, age, depressive symptoms, and loneliness were associated with substance use (Table III). Older age was associated with a threefold increase in risk for substance use, OR = 3.19, 95% CI = 1.47–6.91. A one-unit increase in depressive symptoms increased the risk for substance use by a factor of 1.20, 95% CI = 1.02–1.40. A one-unit increase in loneliness decreased risk for substance use by a factor of 0.88, 95% CI = 0.79–0.99. Pain intensity and activity limitations were not associated with substance use among youth without chronic pain. Nagelkerke's R indicated that the final model accounted for 70% of the variance in substance use among youth without chronic pain.
Youth With Chronic Pain. An identical exploratory logistic regression was conducted to examine factors associated with substance use among youth with chronic pain (Table III). Similar to findings for youth without chronic pain, older age was the only demographic factor associated with increased risk for substance use among youth with chronic pain, OR = 2.13, 95% CI = 1.11–4.10. A different pattern of psychological and pain-related risk factors for substance use emerged for youth with chronic pain relative to findings for youth without chronic pain. No association was found between depression, loneliness or pain intensity with substance use among youth with chronic pain. However, a one-unit increase in activity limitations was associated with a decreased risk for substance use among youth with chronic pain, OR = 0.86, 95% CI = 0.76–0.98. Nagelkerke's R indicated that the final model accounted for 35% of the variance in substance use among youth with chronic pain.
Results
Descriptive Statistics
Participants included 95 adolescents with mixed chronic pain conditions between the ages of 12 and 18 years (M = 15.60 years, SD = 2.6), and 91 adolescents without chronic pain recruited from the community (M = 15.10 years, SD = 2.0). Sample characteristics are shown in Table I. Participants in both groups were primarily female (71%) and White (80%). Adolescents with and without chronic pain did not differ on age or sex; however, participants without chronic pain were more likely to have parents with a higher income, χ (5, N = 173) = 23.83, p < .0001, and the ethnic composition of the two groups differed, χ (5, N = 181) = 11.63, p < .05, with more White participants in the chronic pain group (83%) compared with the group without chronic pain (76%). Youth with and without chronic pain reported similar levels of loneliness, t(170) = −0.50, p = .62. As expected, youth with chronic pain reported significantly higher depressive symptoms, t(172) = 2.60, p = .01; pain intensity, t(180) = 10.83, p < .0001; and activity limitations, t(158) = 13.52, p < .0001, than youth without chronic pain (Table I).
Rates of Current Alcohol and Tobacco use in Youth With and Without Chronic Pain
As hypothesized, adolescents with chronic pain were less likely to drink alcohol as compared with peers without chronic pain (7.4% vs. 22%, respectively), χ (1, N = 175) = 7.94, p < .01 (Table I). However, contrary to our hypothesis, adolescents with chronic pain were as likely to use tobacco compared with their peers without chronic pain (9% vs. 8%, respectively), χ (1, N = 170) = 0.03, p > .05.
Overall, rates of current alcohol and tobacco use were low across groups. Of the 30 youth (16.2%) in the full sample who endorsed current alcohol or tobacco use, 20 youth (10.8%) used either alcohol or tobacco and 10 youth (5.4%) used both. The number of youth who used one versus both substances did not significantly differ between groups, χ (2, N = 170) = 4.71, p = .10.
Predictors of Substance use Across Groups
Logistic regression was used to test whether psychosocial and pain-related factors were associated with current substance use in the full sample while controlling for demographic factors. Group status (pain vs. healthy), demographic covariates (income, race, sex, age), loneliness, depressive symptoms, pain intensity, and activity limitations were entered as predictors in a single step in the logistic regression model. The dependent variable was the dichotomous substance use variable (0 = no tobacco or alcohol use, 1 = positive tobacco and/or alcohol use). As shown in Table II, age, loneliness, depressive symptoms, and activity limitations were associated with substance use, generally in the predicted directions. Older age was associated with a twofold increase in risk for substance use, odds ratio (OR) = 2.24, 95% CI = 1.51–3.34. A one-unit increase in depressive symptoms was associated with a 1.12-fold increase in risk for substance use, 95% CI = 1.05–1.20. A one-unit increase in activity limitations was associated with a decrease in risk for substance use by a factor of 0.91, 95% CI = 0.85–0.99. However, in the unexpected direction, a one-unit increase in loneliness was associated with a 0.94 decrease in risk for substance use, 95% CI = 0.89–0.99. Nagelkerke's R indicated that the final model accounted for 48% of the variability in substance use across groups.
Results from this analysis partially supported our hypothesis. After controlling for group status and demographic factors, higher depressive symptoms and lower activity limitations were associated with a higher likelihood for current substance use. However, contrary to our hypothesis, higher loneliness was associated with a decreased risk for substance use and pain intensity was not associated with substance use.
Exploratory Analysis: Predictors of Substance use Within Groups
To explore whether a different pattern of associated factors emerged within each group, separate logistic regression models were conducted for youth with and without chronic pain (Table III). Demographic covariates (income, race, sex, age), loneliness, depressive symptoms, pain intensity, and activity limitations were entered as predictors in a single step in the logistic regression model. The dependent variable was the dichotomous substance use variable (0 = no tobacco or alcohol use, 1 = positive tobacco and/or alcohol use).
Youth Without Chronic Pain. Among youth without chronic pain, age, depressive symptoms, and loneliness were associated with substance use (Table III). Older age was associated with a threefold increase in risk for substance use, OR = 3.19, 95% CI = 1.47–6.91. A one-unit increase in depressive symptoms increased the risk for substance use by a factor of 1.20, 95% CI = 1.02–1.40. A one-unit increase in loneliness decreased risk for substance use by a factor of 0.88, 95% CI = 0.79–0.99. Pain intensity and activity limitations were not associated with substance use among youth without chronic pain. Nagelkerke's R indicated that the final model accounted for 70% of the variance in substance use among youth without chronic pain.
Youth With Chronic Pain. An identical exploratory logistic regression was conducted to examine factors associated with substance use among youth with chronic pain (Table III). Similar to findings for youth without chronic pain, older age was the only demographic factor associated with increased risk for substance use among youth with chronic pain, OR = 2.13, 95% CI = 1.11–4.10. A different pattern of psychological and pain-related risk factors for substance use emerged for youth with chronic pain relative to findings for youth without chronic pain. No association was found between depression, loneliness or pain intensity with substance use among youth with chronic pain. However, a one-unit increase in activity limitations was associated with a decreased risk for substance use among youth with chronic pain, OR = 0.86, 95% CI = 0.76–0.98. Nagelkerke's R indicated that the final model accounted for 35% of the variance in substance use among youth with chronic pain.
Source...