- Large primary care study from Naples finds that lower LDL cholesterol is linked with a higher chance of developing type 2 diabetes
- Statin treatment raises diabetes risk at every LDL level, but the association between low LDL and diabetes remains even after accounting for statins
- Results add to a complex picture of cholesterol and glucose biology rather than suggesting that high LDL cholesterol is safe
Statins are among the most widely prescribed medicines in the world.
They lower LDL cholesterol and clearly reduce heart attacks and strokes, yet they also carry a known side effect: a modest increase in the rate of new type 2 diabetes diagnoses, especially at higher doses.
Genetic studies have deepened the puzzle.
Variants that lower LDL cholesterol through pathways such as HMGCR and NPC1L1 are linked to a higher risk of type 2 diabetes.
People with familial hypercholesterolaemia, who have very high LDL cholesterol from birth and a high risk of coronary artery disease, appear less likely than average to develop type 2 diabetes.
Put simply, lower LDL cholesterol in genes often lines up with more diabetes, while extremely high LDL cholesterol in this inherited condition seems to line up with less.
The obvious next question is what happens in ordinary clinical practice when doctors measure LDL cholesterol in adults and follow them for years. Does lower LDL cholesterol in real life also track with more diabetes, and how much of that pattern is due to statins versus LDL cholesterol itself?
Study design and data source
To answer that question, researchers at the Department of Advanced Biomedical Sciences at Federico II University in Naples carried out a six year longitudinal analysis using an extensive primary care database.
A cooperative of 140 general practitioners in Naples share a single electronic medical record system that tracks:
- Diagnoses and problem lists
- Prescribed medicines
- Hospital and emergency admissions
- Laboratory measurements, including LDL cholesterol
- Vital signs and deaths
The database contains records for more than 200,000 adults whose age and location broadly reflect the city population. For this study, investigators applied strict inclusion and exclusion criteria and ended up with 13,674 adults aged between 19 and 90 years who did not have type 2 diabetes at baseline and had suitable data for analysis.
Slightly more than half, 7,140 people, were receiving statin therapy at the start of the observation period.
As expected, statin users were older on average, with a mean age of 70 years compared with 54 years in the 6,534 non users.
Follow up and new cases of type 2 diabetes
Participants were followed for a median of 71.6 months, a little under six years. Over that time:
- 1,819 people, around 13 percent of the cohort, developed type 2 diabetes
- 1,424 of these new diagnoses occurred in statin treated patients, about 20 percent of that group
- 395 occurred in non users, around 6 percent of that group
These raw percentages reflect both the effects of the medicine and the fact that those prescribed statins tended to be older and at higher cardiovascular risk.
The key analytic question was how baseline LDL cholesterol related to incident diabetes once age, statin use and other risk factors were taken into account.
How LDL cholesterol levels related to diabetes risk
The investigators treated LDL cholesterol as both a continuous measure and a categorical one. When they looked at it as a continuous variable, each increase of 10 milligrams per decilitre in LDL cholesterol was associated with a 10 percent lower hazard of developing diabetes during follow up, after adjustment for other factors. In other words, lower LDL cholesterol tended to go hand in hand with higher diabetes risk.
When they divided participants into four LDL cholesterol groups, the pattern held. The incidence of diabetes per 1,000 person years was:
- 27.6 cases in the low group, with LDL cholesterol below 84 milligrams per decilitre
- 17.4 cases in the medium group, 84 to less than 107 milligrams per decilitre
- 13.5 cases in the high group, 107 to less than 131 milligrams per decilitre
- 8.4 cases in the very high group, LDL cholesterol at or above 131 milligrams per decilitre
Across the cohort, then, the lowest LDL cholesterol was linked with the highest diabetes incidence, and the highest LDL cholesterol with the lowest observed incidence.
The role of statins
Statin treatment added a further layer.
Within every LDL cholesterol category, statin use was associated with a higher risk of developing diabetes compared with non use. Adjusted hazard ratios for statin treatment were roughly:
- 1.75 in the low LDL cholesterol group
- 1.63 in the medium group
- 1.54 in the high group
The largest relative increase was seen in people starting with very high LDL cholesterol levels, where statin use was linked with an adjusted hazard ratio of 2.41 for incident diabetes.
Taken together, these results suggest two overlapping effects. Statins themselves increase diabetes risk regardless of starting LDL cholesterol. At the same time, lower LDL cholesterol values are associated with more diabetes even when statin use is accounted for, with the lowest LDL cholesterol category carrying the highest risk.
How should this be interpreted
The findings fit with the genetic evidence that mechanisms lowering LDL cholesterol can raise diabetes risk, but they do not mean that high LDL cholesterol is healthy or that statin treatment should be avoided. High LDL cholesterol is a major cause of atherosclerotic cardiovascular disease, and the absolute cardiovascular benefits of statins are large, especially in people with existing heart disease or very high risk.
There are several possible explanations for the observed pattern:
- Biological links between cholesterol metabolism and insulin secretion, insulin sensitivity and glucose handling
- Genetic variants that influence both LDL cholesterol and diabetes risk in opposing directions
- Residual confounding, for example people with lower LDL cholesterol may differ systematically from those with higher values in ways that the study cannot fully capture
Observational studies can show associations but cannot prove that low LDL cholesterol directly causes diabetes. Nevertheless, the consistency between genetic findings and this large primary care cohort strengthens the idea that the relationship is not purely an artefact of statin prescribing.
Clinical and research implications
For clinicians, the study underlines points that are already emerging from guidelines and expert opinion:
- Statin therapy carries a real but modest increase in diabetes risk, which should be discussed with patients, especially those who are already at high risk of diabetes
- The cardiovascular benefits of lowering LDL cholesterol with statins generally outweigh this risk in people with raised cardiovascular risk
- Patients on statins, particularly at higher doses and with low LDL cholesterol, may merit closer monitoring for rising glucose or early diabetes
For researchers, the Naples data highlight the need to clarify mechanisms that link LDL cholesterol pathways with glucose metabolism and to identify strategies that preserve cardiovascular benefit while minimising diabetogenic effects. Future work may also explore whether certain patient groups or drug combinations can reduce this trade off.