Table 3.
Standardized effects of influence and income when entered jointly in linear GEE models of biomarkers and in binary logistic GEE models of disease diagnoses, with age and BMI as controls
| Effect of influence | Effect of income | |||||
|---|---|---|---|---|---|---|
| Health measure | n | Number of observations | Β | Β | ||
| Cortisol | 103 | 171 | −0.27** | 0.19* | ||
| Sed. rate | 114 | 251 | −0.09 | −0.07 | ||
| CRP | 58 | 58 | −0.01 | −0.09 | ||
| Lymphocytes | 112 | 241 | 0.04 | −0.04 | ||
| Systolic BP | 166 | 461 | −0.06 | 0.04 | ||
| Diastolic BP | 166 | 461 | −0.09 | 0.17 | ||
| Health measure | n | Number of observations | Effect of influence | Effect of income | ||
|---|---|---|---|---|---|---|
| OR | 95% CI | OR | 95% CI | |||
| Cardiovascular | 166 | 456 | 1.14 | 0.74–1.75 | 1.65 | 0.81–3.34 |
| Intestinal parasites | 108 | 231 | 0.84 | 0.57–1.24 | 0.90 | 0.61–1.32 |
| Respiratory infection | 166 | 456 | 0.73* | 0.54–0.99 | 1.10 | 0.82–1.47 |
| Skin infection | 166 | 456 | 1.02 | 0.69–1.52 | 0.81 | 0.57–1.15 |
Notes: The lymphocytes model also controls for current infection status. Odds ratios (OR) are for increments of 1 SD of influence or income. The sample sizes vary because different random samples of urine, serum and feces were collected and analyzed. *P < 0.05; **P < 0.01.