Quitting smoking and diabetes risk

Tuesday January 5 2010

“People who give up smoking are prone to developing diabetes because they gain weight,” The Times reported. It said a study has found that quitters are twice as likely as smokers, and 70% more likely than non-smokers, to have type 2 diabetes.

This study found that smokers and recent quitters had a greater risk of diabetes compared to those who had never smoked, but that three years after quitting this risk had reduced. The suggestion that this is because quitters are more likely to gain weight is logical, but it cannot be proven by this cohort study.

The results of this study do not mean that smoking is protective to health. Smokers and former smokers were at greater risk of diabetes than those who had never smoked, and the benefits of giving up far outweigh any temporary increase in risk. Instead, these findings emphasise the importance of an active lifestyle and a healthy balanced diet, and demonstrate the importance of providing quitters with education and support to achieve this.

Where did the story come from?

This research was carried out by Hsin-Chieh Yeh and colleagues from Johns Hopkins University, Baltimore; the Federal University of Rio Grande do Sul, Brazil; and the University of North Carolina, Chapel Hill. The study was funded by the National Heart, Lung, and Blood Institute and the National Institute of Diabetes and Digestive and Kidney Disease. It was published in the Annals of Internal Medicine .

What kind of research was this?

This cohort study enrolled a large group of middle-aged people who were free of diabetes, and followed them up over nine years to assess whether quitting smoking affected the risk of diabetes.

Where a randomised controlled trial (RCT) would be unethical, a cohort study is the best alternative for examining whether a particular exposure, in this case quitting smoking, increases the risk of a certain disease developing over time. The research needs to ensure that people are free from the disease at the beginning of the study, and take into account other confounding factors that could influence any observed associations.

What did the research involve?

The data for this study were obtained from a previous study on atherosclerosis called the Atherosclerosis Risk in Communities (ARIC) study, which recruited middle-aged people from several sites in the US. The ARIC recruits visited a clinic between 1987 and 1989, and then had three follow-up visits scheduled at about three-year intervals from 1990 to 1998. From this point onwards until 2004, they were only contacted by phone. Smoking status and number of cigarettes smoked was assessed at every follow-up. The development of diabetes up to the last clinic visit in 1998 was determined by fasting blood glucose levels, and from 1998 up to 2004 by self-report of a doctor’s diagnosis of diabetes or use of diabetes medications.

For this particular study, the 17-year follow-up information from the ARIC study was used for the 9,398 middle-aged adults who were free from diabetes when ARIC began and in the first three years of follow-up, and who had information on smoking status at each point during follow-up. For all participants, physical examination, various other medical data and information on other lifestyle factors were collected during follow-up, and various analyses carried out.

People were grouped according to how much they smoked at the beginning of the study. This was calculated as pack-years smoked (average number of cigarettes per day multiplied by the years of smoking divided by 20). People who were lifelong non-smokers formed the control group. For each category the incidence of diabetes during follow-up was calculated.

To assess the effect of quitting smoking on diabetes risk, the researchers looked at the effect of a change in smoking status from the start of the study to the first three-year follow-up, and risk of diabetes at the three- and nine-year follow-ups. They also looked at the change in smoking status and the effects on various metabolic variables, such as weight, waist and hip circumference, blood pressure and cholesterol. Other statistical analyses were then carried out, including an assessment of how various measures at the beginning of the study could affect weight-gain risk, how various other factors affected diabetes risk, and analyses using only self-reported data.

That multiple statistical tests were carried out is a slight drawback to the study. There are also likely to have been some unavoidable inaccuracies in self-reported measures such as duration of smoking, number of cigarettes smoked and time since quitting.

What were the basic results?

The study found that smoking increased the risk of diabetes, and that there was a dose-response relationship, meaning that the more packs smoked the greater the diabetes risk. Quitting smoking was also associated with increased risk compared to never smoking. New quitters at three-year follow-up (380 of them) were 1.73 times more likely than those who had never smoked to develop diabetes. However, when the analysis was adjusted for weight change, white blood cell count at the beginning of the study and all other known risk factors for diabetes (including gender, BMI, waist circumference, physical activity, triglyceride level, cholesterol, blood pressure), the incidence was 1.24 times greater in quitters than those who had never smoked, but this was no longer significant.

The highest risk of diabetes for quitters occurred in the first three years, but gradually reduced to zero at 12 years. Former smokers who had smoked more than three years ago did not have a significantly increased risk of diabetes.

How did the researchers interpret the results?

The researchers concluded that cigarette smoking increases the risk of developing type 2 diabetes, but that quitting smoking also increases the risk in the short term. They advise that smokers who have other risk factors for diabetes receive smoking cessation care coupled with strategies for diabetes prevention and early detection.


Smoking is associated with an increased risk of diabetes, and the current study confirms this. However, the effect of quitting smoking on diabetes risk has been unclear until now. This study found that quitting is associated with an increased risk of developing diabetes in the short term, but that this risk decreases over time. When the researchers adjusted their analyses for weight change since quitting this affected the risk.

This was a well-conducted study that carried out extensive follow-up on a large number of participants. However, there are some points to consider:

  • As the authors acknowledge, though they adjusted for various established diabetes risk factors, there is still the possibility of residual confounding from unmeasured factors.
  • Several of the self-reported measures, most notably smoking status, frequency of smoking and time since quitting, are likely to involve some degree of inaccuracy.
  • Multiple statistical tests were carried out, and this is a slight drawback to this research, as it increases the risk that the findings are due to chance only. However, this likelihood is reduced by the fact that the study specified its research hypothesis prior to the start of the study.
  • The theory is that, while quitting smoking may reduce inflammation in the body and so reduce diabetes risk, the weight gain that quitters often experience could adversely influence this risk. Although this pattern may be suggested by these results, no firm conclusions can be made. Adjusting for weight change did reduce the strength of the association between quitting and diabetes risk but the risk remained significant, indicating that there are other factors involved. In addition, the reasons for the person's weight gain have not been examined.

The recommendation of the researchers seems sensible. Smokers who quit should seek advice on avoiding weight gain, diabetes prevention and how to spot the early signs of the disease.

Analysis by Bazian
Edited by NHS Choices