Introduction: Anemia is associated with increased morbidity and mortality in the elderly, though the risk factors for and the consequences of hemoglobin (HGB) decline are poorly characterized.

Methods: We studied 5201 men and women ≥65 participating in the Cardiovascular Health Study. The cohort was followed biannually and had baseline and repeat hemograms 3 years later. HGB decline was defined as

  1. >1g/dL HGB drop, or

  2. incident anemia at 3 years by WHO criteria.

Results: 4006 participants survived to 3 years and had two HGB measures. The median HGB change was −0.2g/dL (IQR-0.8, 0.1). 961 (24%) participants had a >1g/dL HGB drop and 335 (8%) developed incident anemia. The left side of the table presents adjusted logistic regression models of baseline risk factors for HGB decline. Those with baseline cardiovascular disease (CVD), diabetes and kidney disease were more likely to develop >1g/dL HGB drop while only baseline kidney disease was associated with incident anemia. The table also shows the adjusted risk of HGB decline with concurrent development of co-morbid conditions. A >1g/dL drop in HGB was more likely in those who concurrently developed incident CVD, hypertension or inflammation. Incident anemia was more likely in participants with concurrent development of kidney disease or inflammation. Both incident anemia and a HGB drop >1g/dL were associated with subsequent 9-year mortality adjusting for age, race, gender, year 3 HGB, hypertension, CVD, diabetes, and renal disease; HRs (95% CI) 1.4 (1.2, 1.6) and 1.2 (1.1, 1.4) respectively.

Discussion: Among studied factors, baseline CVD, diabetes and kidney disease were risk factors for >1g/dL HGB drop while only baseline kidney disease was a risk factor for incident anemia. Incident CVD and hypertension were associated concurrently with >1g/dL HGB drop while kidney disease was associated with concurrent incident anemia. Inflammation development was the strongest risk factor accompanying HGB decline. HGB decline, especially a 1g/dL drop, was associated with subsequent mortality irrespective of HGB concentration. These data suggest that small HGB changes not captured by the WHO anemia criteria are associated with poor health outcomes and that inflammation is a major correlate of HGB decline in the elderly.

Table: Risk Factors for HGB Decline in Age-, Race-, Gender, and Baseline HGB-Adjusted Logistic Regression Models

Baseline Risk Factors for HGB DeclineRisk of HGB Decline with Concurrent Conditions
HGB Drop >1g/dLIncident AnemiaHGB Drop >1g/dLIncident Anemia
CVD 1.2 (1.1, 1.4) 1.0 (0.8, 1.3) 1.3 (1.1, 1.6) 1.0 (0.7, 1.3) 
Hypertension 1.1 (0.99, 1.3) 1.1 (0.8, 1.2) 1.4 (1.1, 1.7) 1.1 (0.8, 1.5) 
Diabetes 1.3 (1.1, 1.5) 1.1 (0.8, 1.4) 0.9 (0.6, 1.4) 0.8 (0.4, 1.7) 
Kidney Disease (GFR <60ml/min/1.73m21.2 (1.0, 1.3) 1.3 (1.1, 1.7) 1.1 (0.8, 1.4) 1.5 (1.0, 2.1) 
Inflammation CRP ≥10mg/dL or WBC≥15×109/mm3 1.0 (0.8, 1.3) 1.3 (0.99 1.8) 2.3 (1.8, 2.8) 2.3 (1.8, 3.0) 
Baseline Risk Factors for HGB DeclineRisk of HGB Decline with Concurrent Conditions
HGB Drop >1g/dLIncident AnemiaHGB Drop >1g/dLIncident Anemia
CVD 1.2 (1.1, 1.4) 1.0 (0.8, 1.3) 1.3 (1.1, 1.6) 1.0 (0.7, 1.3) 
Hypertension 1.1 (0.99, 1.3) 1.1 (0.8, 1.2) 1.4 (1.1, 1.7) 1.1 (0.8, 1.5) 
Diabetes 1.3 (1.1, 1.5) 1.1 (0.8, 1.4) 0.9 (0.6, 1.4) 0.8 (0.4, 1.7) 
Kidney Disease (GFR <60ml/min/1.73m21.2 (1.0, 1.3) 1.3 (1.1, 1.7) 1.1 (0.8, 1.4) 1.5 (1.0, 2.1) 
Inflammation CRP ≥10mg/dL or WBC≥15×109/mm3 1.0 (0.8, 1.3) 1.3 (0.99 1.8) 2.3 (1.8, 2.8) 2.3 (1.8, 3.0) 

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

Sign in via your Institution