Table 3. Sensitivity Analyses Examining the Robustness of Effect Estimates Reported in Table 2 to Unmeasured Confounding in a Sample of 13,887 Adolescents Participating in the COMPASS Study, Canada.
Variable | Adjusted model, β (SE) [95% CI] | E-valuea for point estimate | E valuea for 95% CI |
---|---|---|---|
Depressive symptoms | |||
SSB composite indexb | 0.04 (0.01) [0.01 to 0.06] | 1.08 | 1.06 |
Anxiety symptoms | |||
SSB composite indexb | 0.02 (0.01) [0 to 0.05] | 1.06 | 1.01 |
Psychological well-being | |||
Vegetables and fruit | 0.06 (0.02) [0.03 to 0.10] | 1.11 | 1.06 |
SSB composite indexb | −0.03 (0.01) [−0.05 to −0.01] | 1.08 | 1.04 |
Abbreviations: CESD-R-10, Center for Epidemiologic Studies Depression Scale Revised-10; COMPASS, Cannabis, Obesity, Mental health, Physical activity, Alcohol, Smoking, and Sedentary behavior; GAD-7, Generalized Anxiety Disorder-7; SSB, sugar-sweetened beverage.
To calculate E-values, the following SDs for the outcomes of interest were used: 6.01 for the CESD-R-10 score, 5.55 for the GAD-7 score, and 5.49 for the Flourishing Scale score at 1-year follow-up. E-values are reported on the risk ratio scale.
Responses for these 3 categories of SSBs were summed to derive a composite SSB index, ranging from 0 (did not consume any SSBs in the 3 categories) to 21 (consumed SSBs from all 3 categories every day during a usual week) (13).