LONDON, July 10 (Reuters) - In research that turns on its head previous thinking about links between schizophrenia and smoking, scientists say they have found that cigarettes may be a causal factor in the development of psychosis.

After analyzing almost 15,000 tobacco users and 273,000 non users and their relative rates of psychosis - where patients can experience delusions, paranoia and hear voices in their heads - the researchers said cigarette smoking appears to increase risk.

"While it's always hard to determine the direction of causality, our findings indicate that smoking should be taken seriously as a possible risk factor for developing psychosis," James MacCabe, a psychosis expert who co-led the research at King's College London's Institute of Psychiatry, told reporters.

He added, however, that tobacco was only one of many factors, including certain genetic, diet, lifestyle and other influences, raising a person's risk of developing schizophrenia.

Schizophrenia, a severe psychiatric disorder that affects around one in 100 people, typically begins in early adulthood. Its most common symptoms are disruptions in thinking and perception, and patients often have psychotic experiences.

Although the link between smoking cigarettes and schizophrenia has been noted before, until now many doctors have followed a self-medication hypothesis whereby patients smoke to counteract the stressful symptoms of schizophrenia or the side-effects of antipsychotic medication.

For this study, McCabe's team analyzed rates of smoking in people presenting with their first episode of psychosis and found that 57 per cent of these individuals were smokers.

People with a first episode of psychosis were three times more likely to be smokers than those in the control groups.

Robin Murray, a professor of psychiatric research at King's who worked with McCabe, said activity in the brain's dopamine system might be one explanation of a possible causal link between smoking and psychosis.

"Excess dopamine is the best biological explanation we have for psychotic illnesses," he told the briefing. "It's possible that nicotine exposure, by increasing the release of dopamine, causes psychosis to develop."

Previous studies, some by Murray, have also linked cannabis use to psychosis. But there is much debate about whether this is causal or whether there may be shared genes which predispose people to both cannabis use and schizophrenia.

McCabe said the new results on smoking suggest "it might even be possible that the real villain is tobacco, not cannabis" -- since cannabis users often combine the drug with tobacco.

(Reporting by Kate Kelland; Editing by Mark Heinrich)