A retrospective study of 36,460 Chinese coronary artery disease patients found that those admitted to hospital with low apolipoprotein B (ApoB below 65 mg/dL), the very target that lipid guidelines tell physicians to chase, were roughly 30 percent more likely to die over the following five years than patients with higher ApoB. The authors then show the paradox is an illusion. Patients with low ApoB were older, sicker, and above all malnourished: nearly 90 percent carried some degree of malnutrition versus 49 percent in the high-ApoB group. Once nutritional status and bilirubin were entered into the model, the direction of the association flipped and higher ApoB reverted to being associated with worse survival, in a straight-line dose-response fashion. The paper is a case study in reverse causation and confounding by illness rather than a challenge to lipid-lowering therapy.
For thirty years the lipid hypothesis has told a simple story: the more atherogenic particles circulating in your blood, the more plaque you accumulate and the sooner you die. ApoB is arguably the cleanest measure of that particle burden, because exactly one ApoB molecule sits on every LDL, VLDL and IDL particle. So when a large hospital cohort reports that the patients with the lowest ApoB died first, it demands an explanation.
The team at Guangdong Provincial People’s Hospital pulled records on 36,460 patients with angiographically confirmed coronary artery disease treated between 2007 and 2018, and followed them for a median of five years. Just over one in eight died. Splitting the cohort at the European guideline goal of 65 mg/dL, the low-ApoB group had a mortality rate of 15.5 percent against 11.9 percent in the high-ApoB group.
The interesting work is what happens next. The authors add confounders in layers. Adjusting for age, sex and the usual comorbidities shrinks the excess risk to nothing. Adding two more variables, nutritional status and total bilirubin, reverses it entirely: now the low-ApoB group has the survival advantage, and the relationship between ApoB and death becomes linear rather than the J-shaped curve seen in the raw data.
The big idea is that a low ApoB reading in an acutely unwell hospitalised patient is frequently not a sign of good lipid control. It is a downstream readout of something else going wrong. Cholesterol-carrying lipoproteins are manufactured in the liver, and they fall when a patient is catabolic, inflamed, malnourished, or has liver dysfunction. Bilirubin, an endogenous antioxidant released under stress, rises in the same circumstances. Low ApoB, in that setting, is a biomarker of frailty wearing the costume of a treatment success.
This is the same trap that produced the obesity paradox, the cholesterol paradox in heart failure, and the blood pressure paradox in the very old. In each case, a protective-looking association in sick populations evaporates or reverses once the underlying illness is accounted for. What this study adds is a specific, measurable mediator: malnutrition, quantified by a routine laboratory score, that explains almost all of the anomaly.
The practical message for cardiology is unglamorous but useful. Read a low ApoB in a hospitalised patient in context. If it arrives alongside low albumin, low lymphocytes and rising bilirubin, the patient does not have excellent lipids. The patient is sick.
Actionable Insights
The single most useful takeaway is a reasoning skill rather than a supplement. When you see a headline claiming that low cholesterol, low LDL or low ApoB predicts death, check whether the population was already ill. In this cohort the raw numbers looked alarming: 15.5 percent of low-ApoB patients died versus 11.9 percent of high-ApoB patients, a relative risk of 1.30 and an absolute gap of 3.5 percentage points over five years. That translates to one extra death for every 28 people in the low-ApoB group. But the low-ApoB group was not healthier. Roughly 90 percent of them were malnourished compared with 49 percent of the high-ApoB group, which corresponds to an odds ratio near 9. That is an enormous imbalance, and it is doing the work.
Once malnutrition was accounted for, the direction reversed and higher ApoB again predicted higher mortality. The magnitude, however, is modest: a 10 percent relative reduction, which against a 12.5 percent baseline death rate is about 1.2 percentage points over five years, or roughly 80 people needing lower ApoB to avoid one death.
Second takeaway: nutritional markers are longevity markers. Albumin, lymphocyte count and total cholesterol combined into a simple score separated survivors from non-survivors more powerfully than the lipid exposure did. If you track biomarkers, track these three.
Context and Source
- Open Access Paper: Paradoxical Association Between Baseline Apolipoprotein B and Prognosis in Coronary Artery Disease: A 36,460 Chinese Cohort Study.
- Authors: Li H, Wang B, Mai Z, Yu S, Zhou Z, Lu H, Lai W, Li Q, Yang Y, Deng J, Tan N, Chen J, Liu J, Liu Y, Chen S.
- Institution: Guangdong Cardiovascular Institute, Guangdong Provincial People’s Hospital, Guangdong Academy of Medical Sciences, Guangzhou, China. Additional affiliations at South China University of Technology and Southern Medical University.
- Country: China (single centre).
-
Journal: Frontiers in Cardiovascular Medicine, Published 25 January 2022.
( Impact evaluation: The impact score of this journal is 3.0 (most recent Journal Impact Factor; it was 3.6 in the 2022 JCR release, contemporaneous with publication; Scopus-derived CiteScore is reported around 8.2 and SJR places it in Q1 for cardiology). Evaluated against a typical high-end range of 0 to 60+ for top general science and elite specialty journals, therefore this is a Low impact journal in absolute terms, and a mid-tier journal within its own cardiology subfield.