The Association between Serum Copper and Anemia in the Adult NHANES II Population.

Anemia is prevalent in older adults. Although the majority of cases are due to nutritional deficiencies, chronic inflammation, or renal disease, anemia remains unexplained in about one third of cases (Guralnik et al., Blood, 2004, p2263). Copper deficiency has been reported as a rare cause of anemia and neutropenia. We hypothesized that copper deficiency could account for some cases of unexplained anemia in adults. To test this hypothesis, we examined the relationship between serum copper and unexplained anemia in 11,240 adults (15 years and older) who participated in the Second National Health and Nutrition Examination Survey (NHANES II). Anemia was defined as hemoglobin level below 13.5 g/dL in males and below 11.5 g/dL in females. In this study population, 638 or 5.7% of adults were anemic. Of these, anemia was unexplained by vitamin B12, folate, or iron deficiencies, chronic inflammation, or renal disease in 421 or 3.7% of the total subjects studied. Thus, unexplained anemia constituted approximately 65% of all anemia cases in this sample of US adults.

Median serum copper level was significantly higher in all anemia cases (126 μg/dL, p <0.0001), in explained anemia (133 μg/dL, p<0.0001), and in unexplained anemia (122 μg/dL, p<0.02) as compared to non-anemic adults (119 μg/dL, ref). Spline regression showed a U-shaped relationship between serum copper levels and unexplained anemia. The prevalence of unexplained anemia increased significantly at both lower and higher copper levels as compared to the median (119 μg/dL): odds ratios for unexplained anemia were 1.2 (95% CI: 1.1–1.3) in the association with a decrease of serum copper from median to 10th percentile (92 μg/dL) and 1.8 (95% CI: 1.6–2.2) in the association with an increase of serum copper to 90th percentile (161 μg/dL).

This study likely underestimates the prevalence of anemia of chronic illness, as only a single marker of inflammation (serum iron) was available to capture these subjects. Furthermore, copper levels are known to be elevated in inflammation. Thus, unrecognized cases of anemia of chronic illness might explain the observed association between high serum copper levels and unexplained anemia. On the other hand, subjects with low serum copper levels and unexplained anemia might have nutritional deficiency.

In conclusion, this analysis of the NHANES II data confirms (1) that unexplained anemia is common among the US adults, and (2) that both lower and higher than median serum copper levels are positively associated with unexplained anemia. These findings support the need for further study of copper status in adults with unexplained anemia.