TABLE 3.
Odds ratios of stunting at 20–24 mo, based on features of growth before 20 mo1
| Model I | Model II | Model III | |
|---|---|---|---|
| Predictors | |||
| Baseline odds | 0.076 (0.065, 0.089) | 0.070 (0.059, 0.082) | 0.066 (0.055, 0.079) |
| Sex (boy) | 0.95 (0.82, 1.10) | 0.93 (0.80, 1.08) | 0.93 (0.80, 1.08) |
| Ever stunted | 15 (13, 18) | 14 (12, 17) | 13 (11, 15) |
| Ever wasted | 1.4 (1.2, 1.7) | 1.3 (1.0, 1.6) | |
| Intercept2 | 0.89 (0.79, 1.00) | ||
| Slope2 | 0.00095 (0.00013, 0.00707) | ||
| CCV2 | 2.4 (1.2, 4.5) | ||
| Diagnostics | |||
| Error rate, % | 24 | 24 | 24 |
| False-negative rate, % | 15 | 15 | 18 |
| False-positive rate, % | 27 | 27 | 26 |
1
Values are ORs (95% CIs) or percentages, as indicated. ORs (95% CIs) based on logistic regression, n = 5160. CCV, coefficient of cyclic variation; WLZ, weight-for-length z score.
2
Random effects estimates extracted from WLZ trajectories. The random effects represent individual variation around the mean trajectory of WLZ (with mean zero), e.g., the CCV of the random effects for the Fourier terms quantifies the magnitude of the seasonal variation in WLZ experienced by an individual above/below the seasonal variation described by the average trajectory.