Background
Cigarette smoking is estimated to cause over five millions deaths per year worldwide making it the leading preventable major cause of disease and premature death with a high burden of costs for health care systems [
1,
2]. Although smoking cessation is obviously associated with individual health benefits, several barriers addressing physiologic and psychosocial aspects of nicotine addiction often prevent willingness to quit and successful smoking cessation. One of the most important causes of failure in smoking cessation is the fear of excessive weight gain after successful quitting [
3,
4]. In comparison with their non-smoking counterparts a lower body weight, a deceleration of age-related weight increase and an 80 percent probability of weight gain occurring within 6 months after quitting is commonly described for smokers [
5‐
8].
Several factors have been examined to identify determinants of postcessational weight gain. These include, although partly controversially discussed, younger age [
9], female gender [
10], ethnicity, low socioeconomic status [
11], baseline weight, determinants of tobacco dependence, a sedentary lifestyle [
12] and poor eating habits [
13]. In contrast, there is some evidence that both nicotine replacement therapy (NRT) and bupropion alone or prescribed in combination with NRT can at least partly prevent weight gain following smoking cessation [
14‐
16]. Success rates of most smoking cessation interventions remain modest, but may be increased in future through prematurely identifying determinants of weight gain. The main objective of the present study was to explore postcessational weight changes in a first step and thereafter searching for predictors of marked weight gain in an employed population of health care and industrial workers. For this purpose, longitudinal, prospective data for analysis were taken from our comprehensive smoking cessation program [
17].
Discussion
To our knowledge, the present study is one of the first of its kind to examine long-term weight gain in order to identify predicting factors in a cohort of healthy mid-aged workers participating in a comprehensive worksite related health intervention.
As a main finding of the present study, we observed no significant weight changes following smoking cessation in more than a half of long-term quitters. In the subgroup of long-term with marked weight gain we revealed a mean weight increase of 5 kg after quitting with a high variability between individuals. We were also able to identify some determinants which are potentially useful to prematurely identify smokers at greatest risk.
The relationship between smoking cessation and consecutive weight gain is considered to be complex in nature and most likely multifactorially conditioned. In line with previous research, smoking cessation was clearly associated with substantial weight gain in many smokers of our population. Thereby it should be mentioned that the frequency, the amount and pattern of weight gain differed widely between investigated populations in the past [
27,
28]. In general, assessment of postcessational weight changes is made complex by several facts. The majority of healthy adults in developed countries tend to gain weight with age as it has be shown in several large population studies [
10,
29]. More recent data have also confirmed this observation in a smoking population [
30]. Furthermore, the individual body weight might also be relevant influenced by different cultural and nutritionals factors [
31,
32]. In addition, studies in homozygotic twins have demonstrated that the postcessational weight gain might be in large a part of compensation for underweight in the smoking twin compared to the non-smoking twin [
33]. Therefore, a genetic component of individual body weight might also affect postcessational weight gain.
Finally, contrary reports about weight gain following smoking cessation could also be explained –at least to some extent- by several methodological reasons including the use of only self- reported height and weight, failure to biochemically confirm smoking status, missing control- groups, using more often point- versus continuous abstinence and limited follow-up periods [
34].
Relevant weight gain might both negatively affect the outcome of a quit attempt and at least party attenuate the health benefits of smoking cessation due to an increased probability for metabolic and cardio-vascular disorders in high-risk subjects [
35,
36].
We found a mean weight gain of 5 kg occurring in the first year after quitting, which is in line with various earlier reports describing an increase between 4 to 6 kilograms occurring within the first year in continuously abstinent outpatient smokers [
4,
37‐
40]. As one might expect, we observed greater weight gain over time in permanent abstainers as compared to non-sustained quitters and ongoing smokers. Significant gain of body weight occurred mostly within the first year post quitting and showed thereafter a tendency to stabilize until 24 months. Although evidence whether weight increase beyond one year continues is conflicting, our observation stands in agreement with some newer investigations [
41‐
43].
Some randomized clinical studies postulated that both NRT and bupropion might prevent or at least attenuate weight gain [
39,
44,
45]. However, other investigations could not confirm this finding [
40,
46]. In the present study, we found no differences in weight changes post- quitting in regards to several modalities of nicotine replacement with or without bupropion as compared to counselling advice only. Nevertheless, it should be mentioned, that our finding was made in a non- randomized, “real-life” setting with unmatched treatment groups. The non-effect of NRT on weight might be therefore due to a self-selection of NRT by heavier smokers with an overall higher risk for weight gain. Furthermore, we did not provide pharmacological support with varenicline, which might be more effective in attenuating weight gain than NRT based on latest findings [
47].
Postcessational weight gain is considered both to negatively affect the outcome of a quit attempt and at least party attenuate the health benefits of smoking cessation due to an increased probability for metabolic and cardio-vascular disorders in high- risk subjects [
35,
36]. Data from the present study showed that smokers with excessive postcessational weight gain (>10 kg) had a significant higher rate of smoking relapse. Therefore, meaningful predictor variables would be helpful to identify smokers at risk for postcessational weight gain. Predictors identified by existing reports have a number of limitations including restricted generality (data were often collected in clinical populations), which probably differ in weight changes as compared to outpatient, “healthy” smokers [
48‐
51]. In addition, results of population-based investigations often relied on retrospective reports without any objective assessment of weight and smoking status. Additionally, long-term success rates of cessation interventions remained often low. In the present study weight increases revealed a broad range of variation with moderate weight gain (<5 kg) in 58 % and excessive weight gain (>10 kg) in only 10 % of former smokers respectively [
11,
37]. Therefore many tested predictor variables for a 5 % increase of baseline weight might be therefore inaccurate in the individual case.
A lower educational level was the strongest risk factors for weight gain after quitting in the present study. Although data describing the influence of socioeconomic variables on body weight gain are scarce, adopted health habits including limited resources for changing nutritional intake and a more sedentary lifestyle in less educated smokers may contribute to relevant weight gain following smoking cessation [
11]. Addressing motivational factors, employees who reported to smoke for reducing boredom in their life were significant more vulnerable for weight increase. This may represent an increased caloric intake as vicarious satisfaction. In contrast, employees who reported to utilize smoking as weight control practice had surprisingly neither an increased risk to gain weight after smoking cessation nor lower abstinence rates, although this subgroup of smokers was associated with higher nicotine dependence and lower resources of self- efficacy to control body weight [
52]. This observation can’t be fully explained, because data about important confounding variables like restraint eating or increased physical activities during quit attempts were not collected in the present study. Furthermore, the association between individual weight concerns and weight gain after cessation remains controversially discussed in literature [
53,
54].
The subgroup of subjects suffering from diabetes had a greater change for weight gain. This seems to be particularly relevant due to the potentially elevated risk for uncontrolled metabolic situation, morbidity and mortality following smoking cessation [
55,
56].
The lower risk of weight gain in subjects with coronary heart disease seems surprising at first glance. We assume that employable subjects with coronary heart disease might display a higher alertness for healthier lifestyles. We didn’t found any related articles in order to scientifically substantiate this observation. Additionally, the sample size participants with coronary heart disease (n = 28) was quite low, so the strength of this observation needs further proof due to a low statistical power.
In accordance with previous investigations, smokers with a lower amount of cigarettes smoked per day at baseline seem to be less vulnerable for weight gain suggesting that lower levels of nicotine dependence may be protective against body weight increase [
40,
57]. Importantly, it should be mentioned, that our population consisted mainly of low and medium dependent smokers (FTND mean: 4.42 pts, SD ±2.30) and therefore other determinants of nicotine dependence, namely FTND, carbon monoxide and intensity of craving might have failed to independently predict weight gain. The observation, that individual characteristics like age and gender showed no association with postcessational weight gain are in line with some, but not all earlier reports [
11,
12,
58,
59]. Our findings are also in accordance with some earlier investigations showing that smokers with normal baseline BMI (20 kg/m
2) are less vulnerable for marked gain than those with higher baseline weight (BMI 30 kg/m
2) [
58,
60].
The present study has some strengths and limitations. The strengths consist of its community based approach and a high rate of successful long-term quitters (over 35 % at 24 months). In line with a more recent published study we also analysed weight changes for continuous abstainers, part-time quitters and ongoing smokers at different points of time through a longitudinal model [
40]. The later aspect ensured that weight changes of all participants were taken into account and not only weight changes from those who remained permanently abstinent or completed the whole follow-up.
One of our most important limitations is the absence of a control group of lifelong non- smokers that would certainly have added to the strength of the study. Nevertheless, in contrast to many similar studies a relevant number of 156 unsuccessful smokers was followed up for 24 months. These participants represented a kind of internal control group enabling us to better estimate the impact of permanent smoking cessation on long-term body weight.
In summary marked weight gain following smoking cessation occurred in less than a half of smokers following smoking cessation. The majority of smokers intending to quit smoking has concerns about postcessational weight gain and the experience of a minority of quitters who gain large amounts of weight may discourage them from quitting. Therefore, it seems important that health care providers alleviate weight concerns and encourage smokers for a quit attempt. Numerous demographic, socioeconomic, smoking related, cultural and even genetic factors were associated with postcessational weight gain. Therefore, it seems improbable that a single or even a subset of variables as demonstrated by the moderate performance of our multidimensional model can predict individual weight gain following smoking cessation. Nevertheless, early recognition of those smokers at risk for excessive weight gain could enhance the success rate of smoking cessation interventions. Further longitudinal well- designed studies are needed to further characterize the determinants of smokers who are particularly vulnerable for weight gain.
Competing interests
The authors declare that they have no competing interests.
Author’s contribution
All authors had full access to all of the data in the study and took responsibility for the integrity of the data and the accuracy of the data analysis. AS contributed to data analysis, interpretation of results and drafted the manuscript. BS, MK and AM contributed substantially to recruitment and data acquisition, KF gave critical revision for important intellectual content, MT contributed to study conception, interpretation of data and gave critical revision for the manuscript, DS was the primary investigator of the study and contributed to study conception, design, data analysis, interpretation and critical revision of the manuscript for important intellectual content. All authors read and approved the final manuscript.