The study's lead author, Dr Rebecca McKetin, said that ''there have always been questions about causality from those who say methamphetamine users aren't 'turned mad' by the drug but have a pre-existing psychotic condition. What's unique about this study is that it excludes those users and still finds such a strong link between use and psychotic symptoms in a large cohort over a period of years''.

Dr McKetin, formerly of the National Drug and Alcohol Research Centre in Sydney and now at the Australian National University, said she was surprised by the strength of the results but that they will ''come as no surprise to police officers in Kings Cross who report anecdotally about users who will 'go mad' one day and not the next''.

The authors studied 278 methamphetamine users from Sydney and Brisbane over a four-year period between 2006 and 2010. A correlation between methamphetamine use and psychosis had already been found by many studies, but it was uncertain how many of those ice-users had an existing psychotic condition, characterised by hallucinations and delusions of being persecuted. What made the present study original and significant was that it followed the users over an extended period, and it excluded anyone with existing psychotic tendencies.

In the four years the drug-users were observed, the results were compelling. The incidence of psychosis went up sharply with the amount of methamphetamine being used. When participants were abstinent, the likelihood of psychotic episodes dropped to 7 per cent; those who were using the drug one to 15 times a month had a 27 per cent likelihood of suffering an episode, and those using more than 16 times a month had a 48 per cent likelihood. Among all users, the odds of suffering a psychotic episode were 5.3 times greater when they were using than when they were abstinent.

These were all people who had not been sufferers before. Of those episodes, 71 per cent involved suspiciousness and paranoia, 51 per cent were hallucinations, and 35 per cent included periods of delusional thinking. There was a strong connection between psychoses and polydrug use (most of the subjects also smoked cannabis and tobacco and drank alcohol frequently), but once these factors were adjusted for, there remained a link between psychosis and methamphetamine use.

Given the strength of the connection between heavy use and psychosis, the authors wrote, ''there is a good argument for providing methamphetamine treatment as a first-line intervention to reduce rates of psychosis among this population''. At present, users suffering a psychotic episode are most likely to come into contact with hospital emergency departments and police. As well as hospitalisation, methamphetamine use has been involved in numerous crimes of violence since the drug was first widely used in Australia 15 years ago.

After a period in which legislation limiting precursor drugs such as pseudoephedrine cut the supply of ice, Dr McKetin said ''there are telltale signs that methamphetamine is making a comeback in Australia''. She said the key to treatment was to put users on a long-term treatment plan akin to that which is given to sufferers from paranoid schizophrenia, ''instead of treating them for the episode and then turning them back onto the street''.