Table 2.
Multivariate stepwise linear regression analyses of variables impacting health-related quality of life in SHIELD respondents*
| Beta coefficient (SE) | ||
|---|---|---|
| Significant variables | Physical Component Score† | Mental Component Score† |
| Diabetes risk/diagnosis | ||
| 0–2 risk factors | (reference) | (reference) |
| 3–5 risk factors | −4.82 (0.25) | −1.71 (0.24) |
| Type 2 diabetes | −6.53 (0.27) | −1.72 (0.25) |
| Income ($) | ||
| < 22,500 | −7.81 (0.29) | −5.42 (0.28) |
| 22,500–39,999 | −4.45 (0.30) | −2.94 (0.28) |
| 40,000–59,999 | −2.76 (0.30) | −1.56 (0.29) |
| 60,000–89,999 | −1.40 (0.30) | −1.18 (0.29) |
| ≥ 90,000 | (reference) | (reference) |
| Age (years) | ||
| 18–24 | 1.80 (0.56) | 1.59 (0.53) |
| 25–34 | 1.00 (0.38) | −1.26 (0.36) |
| 35–44 | (reference) | (reference) |
| 45–54 | −2.82 (0.30) | 1.42 (0.29) |
| 55–64 | −4.62 (0.32) | 3.62 (0.30) |
| 65–74 | −5.27 (0.34) | 6.30 (0.32) |
| ≥ 75 | −8.78 (0.36) | 5.40 (0.35) |
| BMI (kg/m2) | ||
| Underweight (< 18.5) | Dropped | −1.63 (0.81) |
| Normal weight (18.5–24.9) | (reference) | (reference) |
| Overweight (25.0–29.9) | Dropped | Dropped |
| Obese (> 30) | −4.00 (0.21) | −0.75 (0.20) |
| Female | −2.31 (0.19) | −1.90 (0.18) |
| Male | (reference) | (reference) |
| Race | ||
| White | (reference) | (reference) |
| Black | 0.95 (0.34) | 0.76 (0.33) |
| Other | −1.34 (0.56) | −1.10 (0.54) |
| Geographic region | ||
| East South Central | −1.41 (0.37) | −0.71 (0.36) |
| New England | 1.01 (0.42) | Dropped |
| West North Central | Dropped | 0.93 (0.34) |
| East North Central | Dropped | 0.52 (0.23) |
| Pacific | (reference) | (reference) |
| Household size | ||
| 1 | (reference) | (reference) |
| 3 | Dropped | −0.89 (0.24) |
| ≥ 5 | −0.71 (0.32) | −1.31 (0.31) |
*
Scores indicate change from reference group: gender = male, race = white, household income = ≥ $90,000, age = 35–44, BMI = normal weight, group = low risk, geographic region = Pacific, household size = 1.
†
p < 0.05 vs. reference group for all values. Dropped = level of variable dropped from model during stepwise regression, no beta coefficient computed. SHIELD, Study to Help Improve Early evaluation and management of risk factors Leading to Diabetes; BMI, body mass index.