HDL cholesterol

What is being measured

HDL cholesterol is not a substance, but a measure. What is measured is the amount of cholesterol contained in a specific class of transport particles known as high-density lipoproteins. These particles take up excess cholesterol from tissues and blood vessel walls and transport it to the Liver back; the structure of lipoproteins is described under Cholesterol and lipoproteins.

The figure on the test results therefore says something about the ‘cargo’ and nothing about the work. It does not state how many particles are in circulation, how effectively they take up cholesterol, or how much of it actually reaches the liver. Everything else hinges precisely on this gap, because the return transport is the reason for the test’s good reputation, and yet it is not measured.

The reference range, and what it doesn’t tell you

Each test result includes a reference range, and this varies depending on the laboratory and the measurement method. You should therefore refer to the range stated on your own test result, rather than a figure found on a forum. Typical ranges are 40 to 90 mg/dl for men and 50 to 100 mg/dl for women; values below 25 mg/dl fall significantly outside this range.

These limits are conventions in clinical laboratory medicine. They indicate where most people fall within the range, and none of the studies listed below has examined them as target values.

However, there is a threshold in the literature, and it is significantly lower than the usual target figures would suggest: a value below 20 mg/dl is considered a severe HDL deficiency. In an analysis of 258,252 people, this affected 0.33 per cent of men and 0.099 per cent of women.

The difference between a reference range and a target value is precisely the point here. A reference range indicates where most people fall. A target value assumes that changing the figure will alter the course of events, and that is exactly what has been tested for this value.

Why a high value does not guarantee protection

In observational studies, low HDL levels are associated with a higher incidence of cardiovascular events. It was therefore concluded that higher HDL levels offer protection. This conclusion has been tested in three independent ways, and it does not hold up to scrutiny in any of them.

The genes

If a high level offers protection, people who, due to a gene variant, have higher HDL levels throughout their lives should suffer heart attacks less frequently. A study investigating this question utilised a variant in the gene for endothelial lipase, which raises HDL cholesterol by 0.14 millimoles per litre. The risk of a heart attack was unchanged: odds ratio 0.99 with a confidence interval of 0.88 to 1.11. The same calculation, applied to LDL, however, confirmed exactly what the observational studies show.

That is the most important part of the findings. It is not the case that the test is incapable of detecting anything. It detects the link with LDL, but not the link with HDL.

The medicines

The second approach is the direct one: administer a substance that raises HDL levels, and see what happens.

Torcetrapib raised HDL cholesterol by 72.1 per cent in 15,067 high-risk patients within one year, whilst simultaneously reducing LDL by 24.9 per cent. The study was terminated early because more people died: a hazard ratio of 1.58 for all-cause mortality and 1.25 for cardiovascular events. Blood pressure rose by 5.4 mmHg, and it remained unclear whether this was the sole explanation.

Niacin was tested twice. In the first study, 3,414 patients with low HDL levels who were already on statin therapy were given niacin as an additional treatment. Blood lipid levels improved, but the clinical endpoint did not: 16.4 per cent versus 16.2 per cent, hazard ratio 1.02. The study was terminated due to lack of efficacy. The second study, with 25,673 participants, was significantly larger. Niacin reduced LDL by 10 mg/dl and increased HDL by 6 mg/dl, without reducing the incidence of serious vascular events (13.2 per cent versus 13.7 per cent). Serious adverse events and blood glucose disturbances increased.

The upper end

The third approach is to look at the people with the highest levels. In two population cohorts comprising a total of 116,508 participants and 745,452 person-years, the relationship between HDL cholesterol and overall mortality was U-shaped in both sexes. In the highest category, the adjusted hazard ratio was 2.06 for men with HDL levels of 3.0 millimoles per litre or higher, and 1.68 for women with levels of 3.5 millimoles per litre or higher.

This is an observation, not an attribution of causes; the study itself requires confirmation from other studies. What it rules out is the notion of a straight line where more is always better. The range above 60 mg/dl is therefore not the ‘hall of fame’ that it has long been regarded as.

What a low value means in that case

He remains an indicator, and a good one at that. Only he does not point to himself, but to whatever has weighed him down, and that varies from case to case.

An analysis of 258,252 individuals shows just how significant a factor this is. A severe HDL deficiency of less than 20 mg/dl was rare, occurring in 370 men (0.33 per cent) and 144 women (0.099 per cent). In 40.1 per cent of cases, there was a secondary cause. Of the individuals whose genomes were sequenced, 48.8 per cent carried a mutation or variant in one of four genes, and the prevalence of atherosclerotic cardiovascular diseases differed between them by a tenfold: 40.0 per cent for APOA1, 37.0 per cent for ABCA1, 11.1 per cent for LPL and 4.0 per cent for LCAT, compared with 6.4 per cent among those without a mutation.

The levels are therefore similarly low, and the reported prevalence ranges from below average to six times that level. In LCAT deficiency, the enzyme responsible for esterifying cholesterol within HDL particles is absent; HDL levels are very low, and in this assessment are less frequently affected than in people without any mutation. In APOA1, the particle’s main protein is missing, and the situation is reversed. The rare congenital forms are found among Cholesterol classified; they are known as familial hypoalphalipoproteinaemia, Tangier disease and LCAT deficiency.

The subgroups are small: 201 out of 308 suitable individuals were sequenced, spread across four genes. The individual percentages are therefore imprecise. It is not possible to determine exactly how wide the range between 4 and 40 per cent is from the reported figures alone; however, this does not affect the direction of the findings. Individuals undergoing testosterone therapy were excluded from this analysis.

What needs to be measured

The factor that answers this question is efflux capacity: the amount of cholesterol that a person’s serum can remove from cells. In a study of 2,924 individuals followed for a median of 9.4 years, the risk of a first cardiovascular event was 67 per cent lower in the highest quartile of efflux capacity than in the lowest (hazard ratio 0.33). The HDL cholesterol level of the same individuals was not associated with the events in the fully adjusted model.

This test is not carried out by GPs. It is a laboratory procedure used in research, and it is unclear whether it can be influenced. It is mentioned here because it explains why the figure on the report and the question behind it do not match.

Which shifts the value

Androgens lower HDL, whilst oestrogens raise it. This inverse relationship has long been described and also applies to LDL and lipoprotein(a). On the androgen side, a pathway has been described that involves hepatic lipase, the enzyme that breaks down HDL particles in the liver; the figure relating to this is given below and is taken from a single small study. The consequences for the heart and blood vessels are listed under Cardiovascular risks associated with AAS.

Swallowed versus injected

Eleven strength athletes is a small number, but the study design makes the comparison robust: it was a crossover trial, meaning the same men received both treatments in succession. The Apolipoprotein A-I fell by 40 per cent under stanozolol and by 8 per cent under testosterone. The authors conclude that the adverse lipoprotein effects of orally ingested 17-alpha-alkylated steroids differ from those of injected testosterone.

What the experiment does not separate is the underlying cause: Stanozolol is both 17-alpha-alkylated and was swallowed, and both factors coincide here.

The trial involved a daily dose of 6 milligrams of stanozolol. It does not state how this compares to the quantities actually used. However, a review of the available studies identifies precisely this limitation: most laboratory studies have not tested the doses typically abused, and as a result, the adverse effects are more likely to be underestimated than overestimated.

A similar finding has emerged from a study of healthy male strength athletes. In 19 of them, who self-administered several substances simultaneously over a period of eight or fourteen weeks, HDL cholesterol fell from 1.08 to 0.43 millimoles per litre, apolipoprotein A-I from 1.41 to 0.71 grams per litre, and apolipoprotein B rose from 0.96 to 1.32. In 16 control subjects control subjects, there was no change. In the second, blinded and placebo-controlled part of the same study, nandrolone decanoate alone, injected at a dose of 200 mg per week, did not alter any of these values.

The apolipoprotein B component is the more significant factor here, as each LDL particle carries exactly one of them: the value counts particles rather than cargo. And as far as LDL is concerned, causality has been established. Over 200 meta-analyses from cohort studies, Mendelian randomisation and randomised trials involving a combined total of over two million participants reveal a dose-dependent relationship between LDL load and disease risk, and a reduction in the number of LDL particles reduces the risk proportionally to that reduction. The increase in ApoB among the substances has therefore not been tested, but it points in the same direction as the measure that has been tested.

Will it recover?

The answer depends on what we are talking about. After six weeks of stanozolol, the levels of lecithin-cholesterol acyltransferase, lipoprotein(a) and apolipoprotein D had returned to near baseline levels five weeks after discontinuation. In the case of the athletes who used several substances at the same time, however, the changes were still evident six weeks after they had stopped taking them.

The same study also notes that recovery after fourteen weeks of use was slower than after eight.

Both statements appear side by side and do not contradict one another; they describe different types of stress. The blanket statement that HDL levels recover within weeks only applies to the first case. Anyone wishing to demonstrate a trend should measure it, rather than simply assuming it.

Other factors affecting the value

• Oral androgens: In this comparison, the sharpest fall, a third in six weeks. The potency of another substance depends on the substance and the dose, and is therefore not specified.
• Injectable androgens: Testosterone enanthate reduced the level significantly less in the same comparison, whilst nandrolone decanoate on its own had no effect on most lipid levels. That is not a rule that applies to all injectable substances.
• Endurance training, Giving up smoking, reducing belly fat: are regarded as measures that increase the value . This is not substantiated in the studies cited here, and they are recommended for other reasons in any case. This is what distinguishes them from a measure that merely increases the figure.

What this means in strength training

The value remains useful in the test results, but for a different reason to the one attributed to it: under the tested oral stanozolol, it shifted by a third over six weeks, It therefore indicates that there is a change in the lipid profile. It is not an indicator for the liver; the Liver values beside it.

• A baseline figure before you start is essential. Without it, it will be impossible to tell later on whether a reading of 32 mg/dl is a drop or simply your normal level.
• This value, together with LDL and Triglycerides read, and, where available, with apolipoprotein B. A fall in HDL alongside a rise in ApoB is a different situation to a fall in HDL on its own.
• The key lies in the substance, not in the figures. A dose or duration threshold below which the lipid shift does not occur has not been established; it has only been shown that recovery after fourteen weeks was slower than after eight. On the other hand, drugs designed specifically to raise HDL levels have not shown any clinical benefit in large-scale studies.
• There is little that helps counteract the androgenic effect. None of the studies cited here has investigated whether any of the standard measures counteracts the effect of ongoing oral administration. Whether Cardio or Omega-3 Whether they are sensible does not depend on this figure.
• A check-up after weaning is part of the process, and not after six weeks, when several substances were being administered at the same time. At that point, the levels were still elevated.

None of the studies cited demonstrates a lower limit below which a targeted increase in this figure improves the clinical outcome. Thresholds such as 35 mg/dl stem from the same line of reasoning as the upper target value and are not mentioned in any of them.

Classification

The whole difference lies between a good indicator and a worthwhile target, and in the case of HDL, the two diverge. In observational studies, low levels are associated with a higher incidence of adverse events. However, no benefit from raising HDL levels is evident either in cases of genetically higher HDL or with drugs that raise it, and the findings at the upper end further contradict the assumption that more is always better. For LDL, the situation is the reverse: here, numerous cohort studies, Mendelian randomisation studies and randomised trials reveal a consistent dose-dependent relationship.

In practical terms, this means that a drop in HDL levels whilst taking an oral medication is not a figure that should be corrected, but rather an indication that blood lipid levels have shifted in an unfavourable direction. It does not indicate the overall extent of the risk; for that, the other values are needed. What the data do reveal is this: it does not make sense to treat this value in isolation as a treatment target, and if it falls, the cause and the rest of the lipid profile must be taken into account. Incidentally, it has not yet been demonstrated that hormone therapy reduces cardiovascular risk in the case of testosterone.

The most interesting question remains unanswered. In a cohort, efflux capacity predicts what the HDL level does not. It is not known whether it can be improved or whether that would make any difference. Until then, the figure on the report remains what it is: a sensitive indicator whose significance depends on what triggered it.

This article was created using AI systems. Find out more