Want to lose weight, stop smoking and cut down on your alcohol consumption? An easy way to do all three is a tempting prospect in the post-Christmas gloom. This new healthy lifestyle could be as easy as just popping a pill, if the hype surrounding the new "wonder drug" Acomplia is to be believed.

A flurry of articles last year have lauded this drug, which is being developed by the newly merged French company Sanofi Aventis. The drug has been shown to help people quit smoking - without the usual expanding waistline - and induce weight loss. Trials to see if the drug cuts patients' alcohol intake are under way and some experts have suggested it could be a potential treatment for schizophrenia and serious drug addiction.

Doctors are crying out for help to slim their overweight patients. Obesity drastically raises the risk of other ailments such as diabetes and heart disease. Pharmaceutical companies are scrambling to develop new drugs to tackle the problem, and Sanofi's drug, which is also known as rimonabant, is one of the nearest to being marketed.

It will not be easy to become a sylph-like picture of health, even when using the drug. All the patients in the Acomplia weight-loss trials were on a diet that cut their consumption by 600 calories a day. While the reduction in weight was significantly greater for people who took the drug compared with those taking a placebo, it did not have this effect on all patients. Some gained weight and a third did not lose more than 5% of their body weight.

The main problem with the drug, however, is that the effects stop after a year. The average weight loss is about 7kg (15.4 lbs) - with the average weight being 102kg - in the first year of taking the drug, compared with a 2kg loss in those taking placeboes. No more weight is lost in the second year. When patients stopped taking the drug in the trials, they returned to their original weight.

This suggests that the drug would have to be taken over a long period of time just to retain the small weight loss achieved in the first year.

"Either it's a kick-start to a lifestyle change, which doesn't seem to have worked, or you have to take the drug for a long time just to get the benefits of the original weight loss," said Stewart Adkins, a pharmaceutical analyst at the investment bank Lehman Brothers. "I'm not sure there is enough data to know what the risk reward is if that is so."

The risks will be closely studied if the drug has to be taken over a long time. Acomplia acts on the brain, which is especially likely to raise concerns. The proportions of patient numbers are small but those taking it in trials showed more depression, anxiety, nausea and irritability than those on a placebo, although these effects seemed to fade in the second year.

However, doctors might prefer this to the side-effects of other obesity drugs on the market. One, Xenical, has the unfortunate effect of causing involuntary defecation in some patients.

Because Acomplia affects the brain, it is the side-effects that are difficult to measure that worry some experts. In rodents, the drug changes the behaviour of the animals. Researchers have suggested that the receptor the drug acts on is involved in the reward system.

"Smoking is behavioural; it is fine if it stops smoking but this drug has to be given for 40 years [for obesity]," said Professor Steve Bloom, the head of investigative medicine at Imperial College London. "Smoking is a complex behaviour, if the drug stops it then it must affect other behaviours."

The drug acts on a relatively recently discovered system in the body, the cannabinoid system, so called because cannabis produces its effects there. Acomplia does the opposite of cannabis: it turns off the receptors while cannabis excites them, though its mechanism is not fully understood. The body also produces molecules that bind to the receptors and, as the drug seems to compete with them, it could disrupt their effects.

"The other thing that endogenous cannabinoids [the body's own molecules] do that are pretty important: they help people to forget things," said Professor David Kendall, head of the University of Nottingham's medical school. "Forgetting about making an idiot of yourself in the pub or forgetting your grandma dying is a good thing. If you are getting an antagonist [to those molecules], will that stop? They are rather subtle things but clinical trials don't tend to look for them."

Mice or rats that are given the drug display a variety of responses in academic tests: increased memory and movement, a greater sense of pain, vomiting and nausea, an increase in the propulsion of food through the gut, to name a few.

The drug also boosts the levels of neurotransmitters in parts of rats' brains, an action also produced by some psychiatric drugs. The doses in these animal experiments may not be equivalent to the doses in the recent human trials, so the effects might not be the same. However, there is a rumour in academic circles that 20mg, the highest dose in the recent trials, is not the most effective at reducing weight in humans, but it was chosen because higher doses produced bad side-effects.

"[Cannabinoid] receptors are in the brain, spinal cord, Fallopian tubes and are part of the stress system," said Prof Bloom. "The part that regulates appetite is quite small but we zap the whole thing when we give Acomplia. This means the chances of side-effects are increased."

Sanofi-Aventis would argue that the clinical trials showed few side-effects and they have studied the drug in humans for two years. But the long-term effects cannot be studied in time to be given to regulators this year, when the company plans to file for approval to market the drug. The arthritis drug Vioxx was on sale for years before it was pulled last September after long-term data showed an increase in heart attacks and strokes in patients.

Regulators may be anxious not to repeat the episode, despite the need to find new ways to trim the nation's waistlines.