Table 3.
Association between MVM usage and self-reported overall health in sociodemographic subgroups
| Group | Self-rated overall health as excellent, very good or good, % (95% CI*), MVM Non-Users |
Self-rated overall health as excellent, very good or good, % (95% CI*), MVM Users | Adjusted effect of MVM usage on self-reported health, OR (95% CI*)† | FDR adjusted p value‡ |
| Age, years | ||||
| 18–44 | 92.3 (91.1 to 93.5) | 94.2 (93.6 to 94.8) | 1.3 (1.0 to 1.6) | 0.03 |
| 45–64 | 79.9 (77.8 to 82.1) | 85.3 (84.2 to 86.4) | 1.3 (1.1 to 1.6) | 0.009 |
| ≥65 | 77.2 (73.8 to 80.5) | 82.0 (80.6 to 83.4) | 1.2 (1.0 to 1.5) | 0.06 |
| Race | ||||
| White | 85.9 (84.7 to 87.2) | 89.1 (88.5 to 89.7) | 1.3 (1.1 to 1.7) | 0.009 |
| Non-white | 80.0 (77.2 to 82.7) | 84.2 (82.8 to 85.6) | 1.3 (1.1 to 1.5) | 0.007 |
| Sex | ||||
| Female | 84.0 (82.5 to 85.4) | 88.1 (87.4 to 88.9) | 1.2 (1.1 to 1.4) | 0.009 |
| Male | 85.9 (84.2 to 87.7) | 88.4 (87.5 to 89.3) | 1.3 (1.1 to 1.6) | 0.009 |
| Family income relative to federal poverty level, % (95% CI) | ||||
| <100 | 71.7 (68.0 to 75.4) | 75.6 (73.1 to 78.1) | 1.4 (1.1 to 1.8) | 0.007 |
| 100–199 | 76.4 (73.6 to 79.2) | 80.7 (79.0 to 82.4) | 1.4 (1.1 to 1.7) | 0.007 |
| 200–299 | 84.8 (82.1 to 87.5) | 87.3 (85.9 to 88.6) | 1.3 (1.0 to 1.7) | 0.04 |
| 300–399 | 89.6 (86.4 to 92.7) | 91.0 (89.6 to 92.4) | 1.3 (0.9 to 2.0) | 0.15 |
| ≥400 | 94.8 (93.5 to 96.1) | 95.2 (94.6 to 95.8) | 1.1 (0.8 to 1.6) | 0.23 |
| Education | ||||
| Did not graduate high school | 67.2 (63.1 to 71.3) | 71.9 (69.7 to 74.2) | 1.4 (1.1 to 1.9) | 0.01 |
| High school graduate | 84.1 (82.6 to 85.5) | 86.7 (85.9 to 87.4) | 1.2 (1.0 to 1.4) | 0.01 |
| College graduate or higher | 93.8 (92.4 to 95.1) | 95.3 (94.7 to 95.9) | 1.4 (1.0 to 1.9) | 0.03 |
*All confidence intervals were computed based on a Rao-Scott-scaled χ2 distribution for the loglikelihood from a binomial distribution using the Survey package in R.
†Estimates were produced after adjusting for age, sex, race, region, education level, income, employment status, health insurance status, presence of child in household, marital status, unmet medical care due to cost in the past year, and not seeing a health professional in office in the past 2 weeks.
‡FDR-adjusted p values were computed using the Benjamini-Hochberg procedure; p value was defined using a multivariate regression model controlling for age, sex, race, region, education level, income, employment status, health insurance status, presence of child in household, marital status, unmet medical care due to cost in the past year, and not seeing a health professional in office in the past 2 weeks.
FDR, false discovery rate; MVM, multivitamin and multimineral.