Condition: Cholesterol & Lipids

The 2017 European Atherosclerosis Society consensus statement assessed whether LDL meets the formal tests for causation. Rare gene variants that raise LDL raise heart-disease risk in a dose-dependent way, and variants that lower LDL lower it correspondingly. Pooling more than 200 cohort studies, the Mendelian randomization studies and the randomized LDL-lowering trials, over two million participants with more than 20 million person-years and over 150,000 cardiovascular events, gives a consistent dose-dependent log-linear relationship between cumulative LDL exposure and risk. The naturally randomized genetic studies and the trials agree that any mechanism lowering LDL particle number reduces events in proportion to the absolute LDL reduction and its duration.

The Cholesterol Treatment Trialists' 2010 meta-analysis pooled individual data from 170,000 participants in 26 randomized trials, including trials of more versus less intensive statin therapy. Each 39 mg/dL (1.0 mmol/L) reduction in LDL cholesterol was associated with about a 22% proportional reduction in major vascular events (a composite of coronary death, non-fatal myocardial infarction, stroke and coronary revascularization), with more intensive regimens adding further benefit and no threshold below which lowering stopped helping. All-cause mortality fell largely through fewer coronary deaths.

IMPROVE-IT randomized 18,144 people after an acute coronary syndrome to simvastatin plus ezetimibe or simvastatin alone. The combination held LDL at 53.7 versus 69.5 mg/dL and lowered the primary composite from 34.7% to 32.7%, both 7-year Kaplan-Meier estimates, over a median 6 years of follow-up (hazard ratio 0.936). FOURIER randomized 27,564 people on statins to the PCSK9 inhibitor evolocumab or placebo, which cut LDL 59%, from a median 92 to 30 mg/dL, and reduced the primary composite from 11.3% to 9.8% (hazard ratio 0.85) over a median 2.2 years, with benefit holding even in those who started with the lowest LDL. Both are non-statin mechanisms, and both deliver the benefit predicted by the size of the LDL reduction.

The ACCELERATE trial randomized 12,092 people with high-risk vascular disease to the CETP inhibitor evacetrapib or placebo. Evacetrapib raised HDL cholesterol about 130 percent and lowered LDL cholesterol about 31 percent, yet the rate of cardiovascular events was not reduced and the trial was halted early for futility. This sits alongside a Mendelian randomization study in which a genetic variant that raises HDL cholesterol was not associated with a lower risk of myocardial infarction, while LDL-raising variants behaved as expected. Together the drug and the genetics point the same way: HDL cholesterol reads risk but does not appear to be a lever that lowers it when raised on its own.

The Cholesterol Treatment Trialists' 2012 meta-analysis of individual data from 27 randomized trials examined people at low baseline risk of vascular disease. Each 39 mg/dL (1.0 mmol/L) reduction in LDL cholesterol reduced major vascular events by about 21 percent per 39 mg/dL (1.0 mmol/L) reduction, and the proportional benefit held even in participants whose estimated 5-year risk was under 10 percent, and under 5 percent. Because the proportional effect is roughly constant across the risk range, the absolute benefit scales with baseline risk: sizeable in secondary prevention and higher-risk primary prevention, and modest in low-risk primary prevention.

The 2016 meta-analysis by Ho and colleagues pooled 58 randomized controlled trials (3,974 participants) of oat beta-glucan, a viscous soluble fiber. A median dose of 3.5 g/day lowered LDL cholesterol by 7 mg/dL (0.19 mmol/L, 95% CI -0.23 to -0.14), non-HDL cholesterol by 7.7 mg/dL (0.20 mmol/L), and apoB by 0.03 g/L versus control. Heterogeneity was high, reflecting differences in dose, food matrix and baseline cholesterol. The mechanism is physical: soluble fiber forms a gel that binds bile acids and reduces cholesterol reabsorption, so the liver pulls more LDL from the blood to make new bile.

The 2014 dose-response meta-analysis by Ras and colleagues pooled 124 studies (201 strata). Plant sterol and stanol intakes of 0.6 to 3.3 g/day gradually lowered LDL cholesterol by an average of 6 to 12%, with the effect continuing to rise up to about 3 g/day to an average of roughly 12%, and plant sterols and stanols behaving similarly. They compete with cholesterol for uptake in the intestine, so less is absorbed. There is no randomized trial showing they cut heart attacks; the case for them rests on lowering LDL, which the causal evidence links to risk.

The 2020 Cochrane review by Hooper and colleagues included 15 randomized controlled trials (about 59,000 participants) that reduced or modified dietary saturated fat for at least 24 months. In the 11 trials that reported cardiovascular events (about 53,300 participants), reducing saturated fat lowered them by 21% (risk ratio 0.79, 95% CI 0.66 to 0.93). Meta-regression showed greater reductions in saturated fat, reflected in greater falls in serum cholesterol, produced greater reductions in events, which explained most of the differences between trials. Replacing the calories with polyunsaturated fat or carbohydrate looked useful; replacement with monounsaturated fat was unclear. The number needed to treat in primary prevention was 56 over about four years.

Drouin-Chartier and colleagues (2020) analyzed three large US cohorts totaling over 5.5 million person-years and 14,806 incident cardiovascular events, then updated a meta-analysis of prospective cohorts (33 risk estimates, 1.72 million participants, 139,195 events). Consuming at least one egg per day carried a hazard ratio of 0.93 (95% CI 0.82 to 1.05) for cardiovascular disease in the pooled cohorts, and the meta-analysis found no association per additional egg per day (relative risk 0.98, 95% CI 0.93 to 1.03). For most people, saturated and trans fats raise LDL more than dietary cholesterol does, because the liver adjusts its own production. This is the settled correction on eggs specifically; people with diabetes and very high responders are less studied, and eggs eaten with bacon and buttered toast are a different exposure from eggs alone.

REDUCE-IT randomized 8,179 statin-treated patients with elevated triglycerides (mostly for secondary prevention) to icosapent ethyl 2 g twice daily or placebo. The primary composite of cardiovascular death, myocardial infarction, stroke, revascularization or unstable angina fell from 22.0% to 17.2% (hazard ratio 0.75) over a median 4.9 years, with cardiovascular death also lower (4.3% vs 5.2%). Atrial fibrillation or flutter requiring hospitalization was more common (3.1% vs 2.1%) and serious bleeding trended higher (2.7% vs 2.1%). The benefit was larger than the modest triglyceride change alone would predict, and some of it has been debated because the mineral-oil placebo may have slightly raised risk in the comparison group. It is a specific high-dose prescription product, not a case for standard fish-oil capsules.

Kodama and colleagues (2007) pooled 25 randomized controlled trials of aerobic exercise training. Mean HDL cholesterol rose 2.53 mg/dL (0.065 mmol/L), a statistically significant but modest change. A minimum of about 900 kcal of weekly energy expenditure, or roughly 120 minutes of exercise per week, was needed to move it, and each extra 10 minutes per session was associated with about a 1.4 mg/dL further rise, while exercise frequency and intensity mattered less. Leaner people and those with higher total cholesterol responded more. Exercise also lowers triglycerides and improves blood pressure, glucose handling and weight, which is where most of its cardiovascular value sits.

Gerards and colleagues (2015) reviewed 20 randomized studies and found red yeast rice lowered LDL cholesterol by 39 mg/dL (1.02 mmol/L) versus placebo, an effect not different from statin therapy, because its active compound monacolin K is molecularly identical to lovastatin. Safety assessment in the trials was mostly of low quality. Cohen and colleagues (2017) analyzed 28 US brands: monacolin K was undetectable in two, and in the 26 that contained it the amount ranged more than 60-fold, from 0.09 to 5.48 mg per 1,200 mg, so the daily dose a label recommends could vary more than 120-fold between products. Taking it means taking a statin at an unknown dose, with the same possibility of muscle and liver effects and the same interactions, but without the labeling.

AIM-HIGH randomized 3,414 people with established cardiovascular disease and well-controlled LDL to extended-release niacin (1,500 to 2,000 mg/day) or placebo, all on simvastatin. Niacin raised median HDL from 35 to 42 mg/dL and lowered triglycerides from 164 to 122 mg/dL, yet the primary composite occurred in 16.4% on niacin versus 16.2% on placebo (hazard ratio 1.02), and the trial was stopped after a mean 3 years for futility. Later trials of other HDL-raising drugs told the same story. A low HDL flags higher risk, but moving the HDL number pharmacologically does not lower that risk, so HDL is best read as a marker rather than a target.

Marston and colleagues analyzed ApoB, LDL cholesterol and triglycerides in UK Biobank participants and in the FOURIER trial cohort. ApoB, which counts each atherogenic particle once, captured the risk of myocardial infarction; in models that included ApoB, LDL cholesterol and triglyceride concentrations contributed little additional predictive information. The practical consequence is discordance: some people with a reassuring LDL cholesterol carry a high ApoB, and some with a high LDL cholesterol carry fewer particles than the number suggests, and it is the particle count that tracks the events.

A Mendelian randomization study in the Copenhagen population found that genetically raised remnant cholesterol was associated with a correspondingly higher risk of ischemic heart disease, the pattern expected of a causal factor rather than a bystander. A 2021 consensus statement from the European Atherosclerosis Society reviewed the genetic, epidemiological and mechanistic evidence and concluded that triglyceride-rich lipoproteins and their remnants contribute causally to atherosclerotic cardiovascular disease, in part by provoking arterial inflammation that raised LDL alone does not.

A meta-analysis of 29 Western prospective studies, covering 262,525 participants and 10,158 incident coronary heart disease cases, compared the top and bottom thirds of fasting triglycerides. The odds ratio adjusted for baseline risk factors but not for other lipids was about 1.72, and it attenuated after additionally adjusting for HDL cholesterol while remaining a residual excess risk independent of the other lipids. The finding places triglycerides as a contributor to coronary risk in their own right, partly through the metabolic state behind them.

PREDIMED randomized about 7,400 Spanish adults at high cardiovascular risk to a Mediterranean diet supplemented with extra-virgin olive oil, the same diet supplemented with mixed nuts, or a control reduced-fat diet. Major cardiovascular events (myocardial infarction, stroke, cardiovascular death) occurred in 3.8% of the olive-oil group, 3.4% of the nut group and 4.4% of the control group, giving hazard ratios of 0.69 (olive oil) and 0.72 (nuts). The reported estimates come from a reanalysis that accounted for randomization irregularities at some sites, and the conclusion held. The pattern moves triglycerides, blood pressure, inflammation and glycemia together rather than acting mainly through LDL, which is why it earns its own line rather than sitting under diet-and-LDL.

Sniderman and colleagues (2011) meta-analyzed 12 epidemiological studies (233,455 subjects, 22,950 events). ApoB was the strongest single marker of cardiovascular risk (standardized relative risk ratio 1.43), non-HDL cholesterol was intermediate (1.34) and LDL cholesterol was weakest (1.25). Because most atherogenic particles carry one apoB molecule, apoB is a direct particle count, which matters when LDL cholesterol and particle number disagree, as they often do in people with high triglycerides, diabetes or metabolic syndrome. Non-HDL cholesterol (total minus HDL) needs no extra test and tracks apoB closely, so it is the practical first upgrade over LDL alone.