Background
In patients with chronic heart failure (CHF) increasing levels of total serum cholesterol are associated with improved survival – while statin usage is not. The impact of statin treatment on the “reverse epidemiology” of cholesterol is unclear.

Methods
2992 consecutive patients with non-ischemic CHF due to left ventricular systolic dysfunction from the Norwegian CHF Registry and the CHF Registries of the Universities of Hull, UK, and Heidelberg, Germany, were studied. 1736 patients were individually double-matched on both cholesterol levels and the individual propensity scores for statin treatment. All-cause mortality was analyzed as a function of baseline cholesterol and statin use in both the general and the matched sample.

Results
1209 patients (40.4%) received a statin. During a follow-up of 13,740 patient-years, 360 statin users (29.8%) and 573 (32.1%) statin non-users died. When grouped according to total cholesterol levels as low (≤3.6 mmol/L), moderate (3.7–4.9 mmol/L), high (4.8–6.2 mmol/L), and very high (>6.2 mmol/L), we found improved survival with very high as compared with low cholesterol levels. This association was present in statin users and non-users in both the general and matched sample (p < 0.05 for each group comparison). The negative association of total cholesterol and mortality persisted when cholesterol was treated as a continuous variable (HR 0.83, 95%CI 0.77–0.90, p < 0.001 for matched patients), but it was less pronounced in statin users than in non-users (F-test p < 0.001).

Conclusions
Statins attenuate but do not eliminate the reverse epidemiological association between increasing total serum cholesterol and improved survival in patients with non-ischemic CHF.