Background: Anemia is common in persons 65 years and older with a prevalance of 10–11%. It is associated with increased mortality, hospitalization and decreased quality of life. Recent findings suggest that age-associated dysregulation of pro-inflammatory cytokines may negatively impact hematopoiesis, and erythropoietin (EPO) levels rise significantly with increasing age.

Objectives: To determine whether high EPO levels and pro-inflammatory cytokines increase the risk for anemia onset in older persons.

Methods: We investigated the effect of EPO levels and pro-inflammatory markers (Interleukin-6, Interleukin-1beta, CRP and Tumor Necrosis Factor-alpha) on the incidence of anemia in the InCHIANTI study, a longitudinal cohort of older adults. A sample of 1155 persons aged 65 years and older were randomly selected from the Chianti area, Italy. Baseline data were collected in 1998–2000 and participants were followed up 3 years later. Anemia was defined according to the WHO criteria as hemoglobin (Hb) <13 g/dL in men and <12 g/dL in women. A total of 679 participants were non-anemic at baseline and available for a longitudinal analysis. To assess the inflammatory state, an inflammation score was calculated by summing the total number of inflammatory markers in the upper tertile (range 0–4).

Results: The 3-year incidence of anemia was 9%. High levels of circulating EPO were significantly associated with an increased risk of developing anemia. Model 1 of Table 1 shows that participants in the highest tertile of EPO were 3 times more likely to develop anemia compared to those in the lowest tertile of EPO (p for trend=.002). Higher levels of inflammation were significantly associated with the onset of anemia. The incidence of anemia was 4 times higher in those with the highest inflammation score vs. those with the lowest score (Model 2, p for trend=.006). Model 3 shows that higher levels of EPO and inflammation remained significant predictors of anemia after adjusting for age, sex and Hb. As expected, those with low normal Hb were significantly more likely to develop anemia relative to those >=2 g/dl above the WHO cutoff (Model 3, p for trend<.0001). Adjusting for comorbidities did not substatively alter the results. Further, the combination of high levels of EPO and inflammation increased the risk for anemia by 9-fold compared to those with low levels of each [ORadj 9.1 (1.9, 43.3), p=0.006].

Conclusions: High levels of EPO and inflammation as well as low normal Hb levels and increasing age were associated with an increased risk of anemia in older adults. The synergistic effects between EPO and inflammation might reflect a compensatory effort to maintain Hb levels in the presence of high levels of inflammation.

Table 1:

Association of EPO, inflammation and Hb with 3−yr incidence of anemia.

Odds Ratio (95% CI)
Model 1Model 2Model 3
*p for trend<.25 
65–74 yrs. 1.0* 1.0* 1.0* 
75–84 yrs. 3.4 (1.7, 6.4) 3.3 (1.7, 6.4) 2.8 (1.4, 5.6) 
>=85yrs. 13.9 (6.2, 31.3) 14.4 (6.5, 32.0) 11.4 (4.9, 26.6) 
women (vs. men) .8 (.5, 1.4) .9 (.5, 1.6) .9, (.5, 1.6) 
Hb [g/dl above WHO]    
0–1   4.3 (1.8, 10.4) 
1−2   1.7 (.7, 4.5) 
>2   1.0* 
EPO tertiles    
low 1.0*  1.0* 
medium 1.4 (.6, 3.1)  1.4 (.6, 3.3) 
high 3.0 (1.4, 6.4)  2.4 (1.1, 5.4) 
Inflammation score    
0−1  1.0* 1.0* 
 1.8 (.9, 3.5) 1.7 (.8, 3.4) 
 2.0 (.9, 4.5) 1.9 (.8, 4.5) 
 4.6 (1.4, 15.3) 3.4 (1.0, 11.4) 
Odds Ratio (95% CI)
Model 1Model 2Model 3
*p for trend<.25 
65–74 yrs. 1.0* 1.0* 1.0* 
75–84 yrs. 3.4 (1.7, 6.4) 3.3 (1.7, 6.4) 2.8 (1.4, 5.6) 
>=85yrs. 13.9 (6.2, 31.3) 14.4 (6.5, 32.0) 11.4 (4.9, 26.6) 
women (vs. men) .8 (.5, 1.4) .9 (.5, 1.6) .9, (.5, 1.6) 
Hb [g/dl above WHO]    
0–1   4.3 (1.8, 10.4) 
1−2   1.7 (.7, 4.5) 
>2   1.0* 
EPO tertiles    
low 1.0*  1.0* 
medium 1.4 (.6, 3.1)  1.4 (.6, 3.3) 
high 3.0 (1.4, 6.4)  2.4 (1.1, 5.4) 
Inflammation score    
0−1  1.0* 1.0* 
 1.8 (.9, 3.5) 1.7 (.8, 3.4) 
 2.0 (.9, 4.5) 1.9 (.8, 4.5) 
 4.6 (1.4, 15.3) 3.4 (1.0, 11.4) 

Disclosure: No relevant conflicts of interest to declare.

Author notes

*

Corresponding author

Sign in via your Institution