Fatigue is a predictor for cardiovascular outcomes in patients undergoing hemodialysis

Abstract

Background and objectives: Despite potential significance of fatigue and its underlying components in the occurrence of cardiovascular diseases, epidemiologic data showing the link are virtually limited. This study was designed to examine whether fatigue symptoms or fatigue's underlying components are a predictor for cardiovascular diseases in high-risk subjects with ESRD.

Design, setting, participants, & measurements: 788 volunteer patients under hemodialysis therapy (506 male, 282 female) completed the survey between October and November 2005, with the follow-up period up to 26 months to monitor occurrence of fatal or nonfatal cardiovascular events. The questionnaire consisted of 64 questions, and promax rotation analysis of the principal component method conceptualized eight fatigue-related factors: fatigue itself, anxiety and depression, loss of attention and memory, pain, overwork, autonomic imbalance, sleep problems, and infection.

Results: 14.7% of the patients showed fatigue scores higher than twice the SD of the mean for healthy volunteers. These highly fatigued patients exhibited a significantly higher risk for cardiovascular events (hazard ratio: 2.17; P < 0.01), with the relationship independent of the well-known risk factors, including age, diabetes, cardiovascular disease history, and inflammation and malnutrition markers. Moreover, comparisons of the risk in key subgroups showed that the risk of high fatigue score for cardiovascular events was more prominent in well-nourished patients, including lower age, absence of past cardiovascular diseases, higher serum albumin, and high non-HDL cholesterol.

Conclusions: Fatigue can be an important predictor for cardiovascular events in patients with ESRD, with the relationship independent of the nutritional or inflammatory status.

Figure 1.

Fatigue score is a predictor for CVDs in patients undergoing hemodialysis treatment. (A) Univariate Cox proportional hazard analyses. For fatigue factors, the score exceeding twice SD of the mean of healthy subjects is represented as “high.” For continuous variables, data were expressed per 1 SD to make the variables scale and unit invariant. †, P < 0.01. CI, confidence interval. (B) Kaplan-Meier analyses for the association between high fatigue score and cardiovascular events.

Figure 2.

Comparisons of the risk of high fatigue score for cardiovascular events in key subgroups as analyzed by Cox proportional hazard analyses. Age, BMI, hemoglobin, albumin, CRP, non-HDL, and HDL were divided into high and low groups on the basis of the medians. *, P < 0.05; †, P < 0.01. CI, confidence interval.