Table 4.
Fully Adjusted Random-Effects Logistic Regression Model Examining the Association of Change in Testosterone Levels With the Odds of High Depressive Symptoms From Visits 1 Through 8a
| Covariate/Estimated Parameter | OR (95% CI) | P Value |
|---|---|---|
| Change in logtestosterone level from baseline | 1.23 (1.04–1.45) | .01 |
| Baseline logtestosterone | 1.27 (1.03–1.57) | .03 |
| Status, premenopause as reference | .002 | |
| Early perimenopause | 1.51 (1.17–1.94) | |
| Late perimenopause | 1.82 (1.31–2.54) | |
| Postmenopause | 1.87 (1.33–2.63) | |
| Race/ethnicity, white as reference | .001 | |
| African American | 1.11 (0.85–1.45) | |
| Chinese | 1.20 (0.72–1.98) | |
| Hispanic | 2.69 (1.32–5.48) | |
| Japanese | 2.21 (1.36–3.59) | |
| Baseline age | 0.96 (0.92–0.996) | .03 |
| Aging | 0.93 (0.90–0.96) | <.001 |
| Baseline education, ≥college as reference | <.001 | |
| ≤High school | 1.99 (1.52–2.61) | |
| >High school/some college | 1.42 (1.12–1.79) | |
| Current smoker | 1.41 (1.09–1.81) | .008 |
| Medication for nerves or depression 2 times/wk in past month | 2.70 (2.15–3.40) | <.001 |
| BMI | 1.01 (0.999–1.03) | .06 |
| Any vasomotor symptoms | 1.61 (1.39–1.87) | <.001 |
| Social support | 0.81 (0.79–0.83) | <.001 |
| Upsetting life events, 0 as reference | <.001 | |
| 1 | 2.62 (2.20–3.11) | |
| ≥2 | 5.99 (5.07–7.08) | |
| Study site | .001 |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CES-D, Center for Epidemiological Studies Depression Scale; CI, confidence interval; OR, odds ratio.
a
High depressive symptoms are indicated by a CES-D score of 16 or higher. All covariates are time varying unless otherwise noted.