How smoking affects sleep: a polysomnographical analysis

Sleep Med. 2012 Dec;13(10):1286-92. doi: 10.1016/j.sleep.2012.06.026. Epub 2012 Sep 28.


Objective: Subjective quality of sleep is impaired in smokers compared with non-smokers, but there is only limited evidence from methodologically sound studies about differences in polysomnography (PSG) sleep characteristics. Therefore, this study used PSG to evaluate sleep in smokers and non-smokers while controlling for other parameters that affect sleep.

Methods: After an adaptation night, PSG sleep laboratory data were obtained from 44 smokers (29 men and 15 women, median age 29.6 years) and compared with PSG data from 44 healthy, sex- and age-matched never smokers. Exclusion criteria were alcohol or other substance abuse, psychiatric or endocrine diseases, and treatment with any kind of psychotropic medication. Nicotine and cotinine plasma levels were measured (in the smoking group) and subjective sleep quality assessed in both groups.

Results: The smokers had a Fagerström tolerance score of 6.4, consumed an average of 21.2 cigarettes per day and had been smoking for 13.1 years (median). Smokers had a shorter sleep period time, longer sleep latency, higher rapid eye movement sleep density, more sleep apneas and leg movements in sleep than non-smokers. There were no differences regarding parameters of spectral analysis of the sleep electroencephalogram as well as in the sleep efficiency measured by PSG. Nevertheless smokers rated their sleep efficiency lower on the Pittsburgh Sleep Quality Index compared with non-smoking individuals, but no differences were detected on the SF-A. Plasma cotinine level correlated negatively with slow wave sleep in the smoking group.

Conclusions: Smokers showed a number of insomnia-like sleep impairments. The findings suggest that it is important for sleep researchers to control smoking status in their analyses. Further research should focus on the causes and consequences of impaired sleep during tobacco cessation, as sleep disturbances are a known risk factor for early relapse after initial tobacco abstinence.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Case-Control Studies
  • Female
  • Humans
  • Male
  • Polysomnography
  • Sleep / drug effects*
  • Sleep / physiology
  • Sleep Stages / drug effects
  • Sleep Stages / physiology
  • Smoking / adverse effects*
  • Smoking / physiopathology