Chicago IL - To borrow from Winston Churchill, it's a "cholesterol paradox" wrapped in a mystery inside an enigma, counterintuitive at several levels but consistent with a good deal of epidemiologic and mechanistic evidence. A large observational study of patients with acute heart failure has suggested that short-term mortality goes up as admission serum total cholesterol (TC) levels go down, yet such patients appear to live longer if they are on statins [1].

Chronic heart failure has previously been associated with a so-called "reverse epidemiology," in which, for example, body-mass index, blood pressure, and serum cholesterol are inversely correlated with mortality [2,3], but it is unclear whether they are mediators or simply markers of risk. The current analysis, based on a European multicenter registry of patients hospitalized with HF, suggests the inverse TC-mortality relationship is independent and points to effects other than lipid modification as central to any clinical benefit HF patients may gain from the drugs, according to the investigators.

Twelve-week mortality for the study's >5000 patients admitted for acute HF was about 12.5%, but the adjusted mortality odds ratio was reduced by a significant 40% for those with serum TC >193 mg/dL (5 mmol/L) compared with lower levels and for those on vs not on lipid-lowering therapy, which was presumed to consist primarily of statins. The reductions were independent of other significant outcome predictors that included creatinine, LV systolic function, and treatment with beta blockers or ACE inhibitors, reported Dr Periaswamy Velavan (University of Hull, Kingston-upon-Hull, UK) recently at the American Heart Association 2006 Scientific Sessions.

Mean baseline serum total cholesterol levels, survivors vs nonsurvivors at 12 weeks

Population	Survivors, n=5123	Nonsurvivors, n=606	Combined
Lipid-lowering therapy, n=1506 (mg/dL)	202	198	202
No lipid-lowering therapy, n=4223 (mg/dL)	194	178	192*
Combined (mg/dL)	196	180†	194

The apparent paradox of an inverse TC-mortality relationship became clearer after analysis by TC quintiles. Those with the poorest prognosis, who had TC levels below 151 mg/dL, accounted for almost a third of the deaths. But the curve for TC vs mortality was U-shaped for the 26% of the cohort on lipid-lowering therapy. Among them, those with TC levels in the middle quintiles fared the best and those in the first and fifth quintiles did the worst. Lipid-lowering therapy was associated with a greatly reduced 12-week mortality regardless of admission TC levels.

Mortality by serum total cholesterol quintile, all heart-failure patients and those on lipid-lowering therapy

End point	<151 mg/dL	151-179 mg/dL	180-205 mg/dL	206-235 mg/dL	>235 mg/dL
Total population
Mortality (% within quintile)	16.6	10.8	8.9	9.0	7.4
Proportion of deaths (%)	32.7	19.6	17.4	16.6	13.7
Patients on lipid-lowering therapy
Mortality (% within quintile)	5.8	5.2	4.0	3.1	5.2
Proportion of deaths (%)	22.5	19.7	16.9	11.3	29.6

The findings argue against lipid lowering per se as the mechanism of statin benefit in the HF population, Velavan told heartwire. The drugs' other well-recognized benefits include anti-inflammatory effects and improvements in endothelial function, and they may also have antiarrhythmic and angiogenic properties, he observed. But it's unknown whether lowering initially "normal" or low TC levels with statins will improve outcomes in the setting of HF. The current limited analysis suggests that may be the case, but "only randomized trials can answer that question."

Velavan explained that low TC levels in HF may reflect increased metabolic stress due to endotoxins or oxidative or inflammatory processes. Patients with the lowest TC levels, he said, "are possibly the sickest patients, with the highest levels of inflammation. We believe that it is possibly the severe inflammatory response associated with the advanced heart-failure patient that causes the low cholesterol." In support of that view, he said, other conditions characterized by a severe inflammatory response and high mortality, such as trauma or septic shock, are known to be associated with low cholesterol levels.