Table 2.
Negative binomial regression of activity-of-daily-living limitations at the Survey of Mid-Life Development in the United States (MIDUS) 3 controlling for limitations at MIDUS 1 (n = 949).
| Inflammation |
ADLs at MIDUS 3 |
|||||
|---|---|---|---|---|---|---|
| b | 95% CI | IRR | 95% CI | |||
| Age (10 years) | 0.06 | 0.00 | 0.13 | 1.29*** | 1.17 | 1.41 |
| Sex (1 = female) | 0.26*** | 0.13 | 0.39 | 1.26* | 1.03 | 1.55 |
| Race (1 = non-White) | 0.16 | −0.11 | 0.42 | 0.87 | 0.58 | 1.30 |
| Education | −0.15** | −0.23 | −0.06 | 0.83** | 0.73 | 0.94 |
| Multimorbidity (1 = yes) | 0.30** | 0.10 | 0.50 | 1.90*** | 1.51 | 2.40 |
| ADLs at MIDUS 1 | 0.08** | 0.03 | 0.13 | 1.20*** | 1.14 | 1.27 |
| Inflammation | – | – | 1.34*** | 1.18 | 1.53 | |
Beta coefficients are shown for predictors of inflammation and IRRs are shown for predictors of ADLs. Age, sex, race, education, and multimorbidity were assessed at MIDUS 1, inflammation at MIDUS 2, and ADLs at MIDUS 3.
*
p < 0.05; **p < 0.01; ***p < 0.001.
ADL, activity of daily living; CI, confidence interval; IRR, incidence rate ratio.