Table 4. Separate, independent and interaction effects of education and wealth on obesity in the low-income countries, DHS data.
| Benin | India | Nigeria | ||||
| N = 14883 | N = 113063 | N = 28901 | ||||
| OR (95%CI) | OR (95%CI) | OR (95%CI) | ||||
| Unadjusted | Adjusted1 | Unadjusted | Adjusted1 | Unadjusted | Adjusted1 | |
| Model using education | ||||||
| Education (level) | ||||||
| None/primary | 1 | 1 | 1 | 1 | 1 | 1 |
| Secondary/higher | 2.16 (1.84–2.53) | 2.01 (1.67–2.44) | 2.06 (1.92–2.21) | 2.23 (2.07-2.41) | 1.84 (1.68–2.04) | 2.39 (2.11-2.70) |
| Model using wealth | ||||||
| Wealth quintile (linear) | 2.09 (1.94–2.23) | 1.99 (1.84–2.16) | 2.77 (2.65–2.89) | 2.41 (2.30–2.53) | 1.76 (1.68–1.83) | 1.80 (1.70-1.90) |
| Model both education and wealth | ||||||
| Education (level) | ||||||
| None/primary | 1 | 1 | 1 | |||
| Secondary/higher | 1.26 (1.04–2.12) | 1.18 (1.09–1.28) | 1.32 (1.16–1.52) | |||
| Wealth quintile (linear) | 1.95 (1.79–2.12) | 2.33 (2.23–2.46) | 1.71 (1.62-1.81) | |||
| Model including education and wealth and a wealth-by-education interaction 2 | ||||||
| Wealth effect within education level 3 | ||||||
| None/primary | 1.98 (1.81–2.15) | 2.25 (2.11–2.39) | 1.72 (1.61–1.84) | |||
| Secondary/higher | 1.67 (1.30–2.16) | 2.48 (2.29–2.70) | 1.70 (1.55–1.86) | |||
| Interaction estimate4 | 0.85 (0.65– 1.10) | 1.10 (1.00–1.21) | 0.99 (0.89–1.10) | |||
| P-value for the interaction estimate5 | 0.2 | 0.06 | 0.8 | |||
1
Estimates adjusted for age group, urban/rural residence, and parity.
2
Estimates for education not shown in this section of the table.
3
OR for obesity associated with an increase in one wealth quintile within each education level calculated using the interaction estimate obtained from the model including an interaction between education and wealth.
4
Wealth effect in the secondary/higher education divided by the wealth effect in the no/primary education group e.g. for Benin, 0.85 = 1.67/1.98).
5
Test of whether the wealth effects differ by education level.