Table 2.
Prevalence and Adjusted Odds Ratios of Metformin Use Among US Adults with Self-reported Prediabetes, NHANES 2005–2014, N = 1607
| Prevalence of metformin use (SE) | Adjusted* OR (95% CI) | |
|---|---|---|
| Age group | ||
| 20–40 years | 4.5 (1.4) | 1 (ref) |
| 40–65 years | 4.9 (0.9) | 1.28 (0.54, 3.06) |
| 65+ years | 9.0 (1.9) | 2.52 (0.99, 6.41) |
| Gender | ||
| Male | 5.7 (1.1) | 1 (ref) |
| Female | 6.1 (1.0) | 1.15 (0.67, 1.98) |
| Race | ||
| Non-Hispanic White | 5.8 (1.0) | 1 (ref) |
| Non-Hispanic Black | 6.5 (1.6) | 1.29 (0.65, 2.53) |
| Mexican American | 6.4 (1.9) | 1.36 (0.63, 2.95) |
| BMI category, kg/m2 | ||
| < 25 kg/m2 | 3.0 (0.9) | 1 (ref) |
| 25–29 kg/m2 | 4.7 (1.0) | 1.53 (0.59, 3.98) |
| 30–34 kg/m2 | 5.6 (1.5) | 2.07 (0.92, 4.64) |
| ≥ 35 kg/m2 | 9.8 (2.0) | 4.00 (1.76, 9.08) |
| Waist circumference category | ||
| Low risk | 3.9 (0.8) | 1 (ref) |
| High risk | 6.9 (1.1) | 1.89 (1.02, 3.51) |
| Education | ||
| Less than high school | 5.7 (1.5) | 1 (ref) |
| High school | 6.6 (1.7) | 1.14 (0.52, 2.48) |
| More than high school | 5.7 (1.0) | 1.08 (0.54, 2.16) |
| Hypertension | ||
| No | 4.2 (1.3) | 1 (ref) |
| Yes | 7.2 (1.1) | 1.39 (0.63, 3.06) |
| High cholesterol | ||
| No | 5.0 (1.3) | 1 (ref) |
| Yes | 8.2 (2.2) | 1.49 (0.71, 3.09) |
| ASCVD 10-year risk | ||
| < 7.5% | 5.3 (1.2) | 1 (ref) |
| ≥ 7.5% | 9.0 (2.3) | 1.31 (0.75, 2.30) |
| History of cardiovascular disease | ||
| No | 5.5 (0.7) | 1 (ref) |
| Yes | 8.4 (2.4) | 1.26 (0.70, 2.27) |
| Smoking status | ||
| Never | 4.9 (1.0) | 1 (ref) |
| Former | 7.3 (1.4) | 1.34 (0.78, 2.32) |
| Current | 6.1 (2.2)† | 1.48 (0.58, 3.76) |
| Tried losing weight in the past year | ||
| No | 3.5 (0.6) | 1 (ref) |
| Yes | 6.5 (1.3) | 2.13 (1.17, 3.90) |
| Physical activity‡ | ||
| < 600 METS/week | 6.8 (1.3) | 1 (ref) |
| ≥ 600 METS/week | 6.5 (1.1) | 1.03 (0.61, 1.76) |
| Family history of diabetes | ||
| No | 6.7 (1.4) | 1 (ref) |
| Yes | 5.2 (0.8) | 0.75 (0.43, 1.32) |
Abbreviations: ASCVD, atherosclerotic cardiovascular disease; BMI, body mass index; CI, confidence interval; METS, metabolic equivalents; OR, odds ratio: SE, standard error
Statistically significant OR estimates are italicized
*Multiple logistic regression for age groups was adjusted for gender and race; multiple logistic regression for gender was adjusted for age and race; multiple logistic regression for race/ethnicity was adjusted for age and gender; all other multiple logistic regressions were adjusted for age, gender, and race
†Estimate considered unreliable per NCHS analytic guidelines as the standard error estimate exceeds 30% of the proportion estimate
‡Beginning in 2007, NCHS used the World Health Organization Global Physical Activity Questionnaire