We used the National Health and Nutrition Examination Survey (NHANES) to describe anemia prevalence and trends for participants aged 65 and older. We included the 2001–2002 through 2015–2016 survey cycles, combining two survey cycles from 2001–2004 to 2013–2016 for the analysis. The National Center for Health Statistics (NCHS) conducts NHANES, a nationally representative multistage probability survey. Participants complete an in-home interview and a clinical and laboratory examination in a mobile examination center (MEC). Because of potential disclosure risk, age was censored at 80 and above in the public use NHANES data files starting in 2007, and we accessed data through the NCHS Research Data Center. The analytical sample consisted of 9,874 individuals.

Hemoglobin values are available as part of the NHANES complete blood count analysis performed in the MEC Clinical Laboratory Improvement Amendments–approved laboratory. Men with a hemoglobin level less than 13 g/dL and women with a hemoglobin level less than 12 g/dL were classified as anemic according to World Health Organization standards.³

We used SAS version 9.3 (SAS Institute, Inc., Cary, NC) and SAS-callable SUDAAN version 11.0.1 (Research Triangle Institute, Research Triangle Park, NC) for data analysis. We used survey design variables to account for complex sampling, t-statistics to compare proportions, and regression analysis trends. Statistical significance was determined using two-sided P<.05. We used the Korn and Graubard⁴ small percentage confidence interval (CI) method to calculate CIs for percentages.

The NCHS Research Ethics Review Board approved NHANES.

In 2013–16, we found an overall anemia prevalence of 14.1% (95% CI=11.8–16.7) for men and 10.2% (95% CI=8.5–12.1) for women aged 65 and older. In 2013–2016, anemia prevalence significantly increased with age (men: 65–74, 7.4%; 75–84, 21.5%; ≥ 85, 39.5%; Figure 1A; women: 65–74, 7.6%; 75–84, 11.0%; ≥ 85, 21.9%; Figure 1B). These trends with increasing age were noted in all survey cycles for men (Figure 1A) and all survey cycles except 2009–2012 for women (Figure 1B).

Over time, we observed a significant increasing linear trend in anemia prevalence from 2001–2004 (24.2%) to 2013–2016 (39.5%) in men aged 85 and older (Figure 1A). The prevalence of anemia in men aged 75 to 84 increased during this time period, but the increase was not significant. For men aged 65 to 74, there was no significant trend over time. In women, anemia prevalence increased from 2001–2004 to 2013–2016 for all age groups, but the linear trends were not significant (Figure 1B), although we identified a significant quadratic trend for women aged 65 to 74 (Figure 1B).

This report presents updated national estimates of anemia prevalence and shows increasing trends with age and over time. Previous literature from NHANES 1988–1994 showed that approximately 26% of men aged 85 and older had anemia², similar to our results in 2001–2004. However, our more recent analysis showed that anemia prevalence increased over time for men aged 85 and older and in 2013–2016 anemia prevalence was higher for both men and women aged 85 and older compared to younger age groups. Reasons for these patterns are probably multifactorial and could include changes in chronic health conditions contributing to anemia which could differ by age, sex, or over time. For example, the incidence of end-stage renal disease peaks at approximately age 80 to 85 and is higher in men⁵, and approximately 53% of persons with Stage 5 chronic kidney disease had anemia.⁶ Also, malnutrition in older persons is common⁷ and may be related to pathogenesis of chronic conditions such as inflammatory bowel disease⁸ or directly related to anemia².

Anemia estimates shown here may be an underestimate. Anemia in homebound older adults has been found to be 4 times as high as that reported from NHANES III data⁹, and NHANES does not include institutionalized persons, a population with higher rates of anemia¹⁰, but these estimates are representative of the general noninstitutionalized U.S. population.

Information on how anemia prevalence has changed over time within sex and age groups adds to the literature and supports the need for additional research. These most recent estimates of anemia in older adults from a nationally representative survey may help inform clinicians caring for this population.