Background: The best thing to do to reduce the harms caused by smoking is to quit, but some people may not want to do this or may feel that they are unable to stop smoking completely. Cutting down the number of cigarettes smoked daily or using different tobacco products, such as chewing tobacco or low-tar cigarettes, may reduce some of the harm caused by smoking. It may also help people to stop smoking completely in the long term. On the other hand, reducing smoking or using other tobacco products may not improve health and could reduce a person's motivation to quit smoking altogether. It is important that we review the evidence to find out whether these approaches could help smokers who do not want to or cannot quit to reduce the harm caused by their smoking. We were mainly interested in whether these approaches improved the health of smokers, but also looked at smoking reduction and quitting rates.

Study characteristics: We found 20 randomised controlled trials that tested ways to help people to cut down the number of cigarettes they smoked. Some of these just advised smokers to smoke less, but most also provided them with a product to help them cut down: nicotine replacement therapy (NRT), varenicline, bupropion, electronic cigarettes (ecigs), or snus (a form of smokeless, oral tobacco). We also found four randomized controlled trials that tested the effects of using cigarettes designed to reduce the damage caused by smoking: reduced tar, carbon or nicotine cigarettes. Most of the studies used NRT to help people to reduce their smoking. All of the studies included people who were not planning to quit smoking soon. The research is current to October 2015.

Key results: Eight studies (with 3081 smokers) found that using NRT roughly doubled the likelihood of halving the number of cigarettes smoked each day, compared to using a placebo. Using NRT in this way also nearly doubled the likelihood of quitting completely. One trial each tested bupropion, varenicline, ecigs and snus to help reduce the harms caused by smoking, and there was no evidence that any of these treatments helped smokers to reduce the number of cigarettes they were smoking each day. This may be because there has not yet been enough research into these methods. Only one of the trials testing cigarettes designed to reduce risk measured their effect on the number of people quitting smoking. It found that people were not more likely to quit smoking if they used reduced-nicotine cigarettes than if they smoked their usual cigarettes. We did not find any trials which reported the long-term health effects of the treatments, and so it remains uncertain how much health benefit there is from reducing the number of cigarettes smoked each day or smoking cigarettes designed to be less harmful.

Quality of evidence: The tobacco industry funded three of the included studies of cigarettes designed to reduce risk. None of the studies looked at whether there had been a long-term change in the health of the users. We rate the quality of the evidence looking at how many people quit smoking as 'low' or 'very low', generally because the findings are based on a small number of studies. We need more studies to investigate methods of reducing the harm caused by continued smoking. These need to measure the health of the users over a long period.