Skip navigation

Acupuncture to stop smoking


This is by way of a late millennial reminder for those still smoking cigarettes. Some may be tempted to try acupuncture to help stop. But does it work? A first-class systematic review says not [1, 2]. Actually there are two reviews, one [1] recently published in the Cochrane Library and dates June 1999. The other [2] is also recently published, but its completion date is in 1997. The paper took two years from acceptance to publication, an interesting comment on the information age! We use information from the more contemporary review.


Typically broad for this group, it included nine databases. Trials had to be randomised comparisons of acupuncture with sham acupuncture, another intervention, or no treatment. Various techniques were examined and information obtained on that and other critical issues to acupuncturists. The only outcome they were interested was complete abstinence from smoking early after treatment and at six and 12 months.


There were 20 controlled studies for analysis. At early times, there was no difference between acupuncture and sham acupuncture (Figure 1, Table 1) with 2069 patients. There was no difference in abstinence rates at six months with 719 patients (Figure 2) or 12 months (Table 1). Where there was useful data on comparison with no treatment, there was no difference in abstinence rates at six months.

Figure: Effect of acupuncture on smoking cessation. Early outcome


Figure: Effect of acupuncture on smoking cessation. Six month outcome


Table: Acupuncture for smoking cessation compared with sham acupuncture
Time Number of trials Number of patients Relative benefit (95%CI) NNT (95%CI)
Early 14 2069 1.1 (0.99 to 1.3) 27 (13 to -710)
Six months 7 719 1.26 (0.93 to 1.7) 21 (10 to -179
Twelve months 3 1196 1.01 (0.77 to 1.3) 493 (25 to -27)

There was few trials comparing acupuncture with no treatment controls, and these gave inconsistent results. There was no difference between acupuncture at the ear and at other sites.


This is smashing stuff in more ways than one. Interesting is that the points on the graph fall right on the line of equality. The authors make some comments on the blinding of acupuncture studies (difficult), on the overall quality of randomisation (generally poor), and on the fact that few studies tested for unequivocal smoking cessation. These factors would tend towards bias for acupuncture, which makes the profound negative even worse. The paper tells us that acupuncture for smoking is a poor buy.

We also have to beware thinking of placebo response here as people stopping smoking because of some feature of the treatment other than acupuncture. The reality is that the smoking cessation rates vary widely because the studies are small. They are not much different (early time point 0-54%, weighted average 27%; six months 0-34%, weighted average 14%) from the sorts of cessation rate with nicotine replacement ( Bandolier 54 ). Here the average placebo rate for nicotine replacement was about 12% with at least six months of follow up.

People entering trials of smoking cessation want to stop smoking. Some of them succeed. With acupuncture, no more succeed. With nicotine replacement about 1 more will succeed for every 13 people who use it.


  1. AR White, H Rampes, E Ernst. Acupuncture for smoking cessation (Cochrane review). In Cochrane Library issue 1, 2000.
  2. AR White, KL Resch, E Ernst. A meta-analysis of acupuncture techniques for smoking cessation. Tobacco Control 1999 8: 393-7.
previous or next story in this issue