Prevalence of vitamin D deficiency in anemia subtypes and in nonanemic persons ≥ 60 years: NHANES III, phase 2, 1991-1994
| . | Nonanemic . | Anemia subtypes . | P, comparison among anemia subtypes* . | |||
|---|---|---|---|---|---|---|
| Anemia of chronic inflammation . | Anemia of nutrient deficiency . | Anemia of chronic kidney disease . | Unexplained anemia . | |||
| No (%) | 2302 | 42 (11.9) | 125 (33.0) | 84 (33.4) | 57 (21.7) | NA |
| Age, y, mean | 70.6 | 70.2 | 74.3 | 73.9 | 71.9 | < .001 |
| Female sex, % | 58.3 | 48.8 | 43.6 | 53.7 | 46.1 | .99 |
| Race/ethnicity, % | < .001 | |||||
| Non-Hispanic white | 84.4 | 55.4 | 80.7 | 69.4 | 67.3 | |
| Non-Hispanic black | 6.8 | 43.9 | 16.7 | 12.9 | 21.7 | |
| Mexican American | 2.4 | 0.7 | 2.6 | 2.2 | 0.8 | |
| Other | 6.4 | 0.0 | 0.0 | 15.4 | 10.3 | |
| Hemoglobin level, g/dL, mean | 14.2 | 11.7 | 11.6 | 11.6 | 11.8 | < .001 |
| Vitamin D deficiency, % | 33.3 | 56.0 | 47.7 | 37.1 | 33.0 | .003 |
| Vitamin D level, ng/mL, mean | 24.5 | 21.5 | 21.8 | 23.6 | 24.2 | .002 |
| . | Nonanemic . | Anemia subtypes . | P, comparison among anemia subtypes* . | |||
|---|---|---|---|---|---|---|
| Anemia of chronic inflammation . | Anemia of nutrient deficiency . | Anemia of chronic kidney disease . | Unexplained anemia . | |||
| No (%) | 2302 | 42 (11.9) | 125 (33.0) | 84 (33.4) | 57 (21.7) | NA |
| Age, y, mean | 70.6 | 70.2 | 74.3 | 73.9 | 71.9 | < .001 |
| Female sex, % | 58.3 | 48.8 | 43.6 | 53.7 | 46.1 | .99 |
| Race/ethnicity, % | < .001 | |||||
| Non-Hispanic white | 84.4 | 55.4 | 80.7 | 69.4 | 67.3 | |
| Non-Hispanic black | 6.8 | 43.9 | 16.7 | 12.9 | 21.7 | |
| Mexican American | 2.4 | 0.7 | 2.6 | 2.2 | 0.8 | |
| Other | 6.4 | 0.0 | 0.0 | 15.4 | 10.3 | |
| Hemoglobin level, g/dL, mean | 14.2 | 11.7 | 11.6 | 11.6 | 11.8 | < .001 |
| Vitamin D deficiency, % | 33.3 | 56.0 | 47.7 | 37.1 | 33.0 | .003 |
| Vitamin D level, ng/mL, mean | 24.5 | 21.5 | 21.8 | 23.6 | 24.2 | .002 |
Data are presented as weighted estimates. Variance could not be estimated because of small group sizes.
Comparisons are made with multinomial logistic regression of unweighted samples because of small group sizes.