Scientists are to investigate why people with low cholesterol levels appear to be more likely to develop Parkinson's disease, following concerns that statins - given to control cholesterol - could cause an increase in the numbers of people with the illness.

About 2.3 million adults in the UK take statins to help control their cholesterol levels; the American scientists have found that those with lower levels of cholesterol are more likely to develop the degenerative neurological disorder of Parkinson's disease.

The link between statins and Parkinson's is not yet understood, and health charities last night urged caution. But the scientists behind the research warn that if they get confirmation of the finding, in their follow-up study of 16,000 people, there could be a surge in Parkinson's diagnoses in the next five years as the effects of the drug set in.

The initial study compared 124 people diagnosed with Parkinson's with a control group of 112. They found that the people with low levels of "bad" LDL cholesterol were in excess of three times more likely to be in the Parkinson's group than those with high cholesterol.

But they also found that those in the trial who took statins were less likely to develop Parkinson's disease, though the study's leader suggested this could be because the group with Parkinson's had had low cholesterol all their lives and that the effect of low cholesterol could be cumulative.

Statins are the world's biggest selling drug. The drug company Pfizer reported sales of $12.2bn (£6.2bn) for its statin, Lipitor, in 2005. The National Institute for Clinical Excellence last year recommended more people take them in the UK, raising the number of customers from 2.3 million to 5.2 million.

The head of the study, Xuemei Huang, at the University of North Carolina, said: "I'm definitely concerned [about the initial findings] which is why I'm conducting a prospective study of 16,000 people."

People should not stop taking statins, she said. The risk of heart disease in those who should be taking statins far outweighed the risk of developing Parkinson's.

The study, which is reported today in the journal Chemistry and Industry, has raised more questions than it has answered. It has not, for example, established whether low cholesterol is a cause or consequence of Parkinson's.

David Dexter, senior lecturer in neuropharmacology at Imperial College London, said: "Although the association is worrying, the study was carried out only in a small number of subjects and hence needs confirming in a larger population. Lower LDL-C levels may also be a consequence of Parkinson's and not a cause. Indeed, the study did not take into account the dietary intake of the two groups in the study, [which] may be important since some Parkinson's patients find it difficult to eat or even swallow food, thus reducing the intake of fats."

Dr Huang said the well-established link between Parkinson's and apoE2, a gene associated with lower LDL cholesterol, supported her theory that low LDL was the culprit in many cases of Parkinson's.

Kieran Breen, director of research at the Parkinson's Disease Society, said people should be wary of such a small study. "Further research into any link between low LDL cholesterol and cholesterol-lowering drugs with Parkinson's is needed. We hope that the proposed study will shed further light on this. The exact causes of Parkinson's are unknown. Research is ongoing. It is generally understood that Parkinson's [arises] from genetic and environmental factors."

Peter Weissberg, medical director of the British Heart Foundation, said: "We are concerned that any suggestion of a link between statins and Parkinson's disease would unnecessarily scare the millions of people benefiting from statins in the UK. There is no evidence to suggest that statins cause [the] disease. On the other hand, there is overwhelming evidence that statins save lives by preventing heart attacks and strokes."