Tobacco Smoking, Eating Behaviors, and Body Weight: A Review

Abstract

Purpose of review:

This narrative review provides an overview of the relationships among tobacco smoking, eating behaviors, and body weight. The aims are to (1) examine the concurrent and longitudinal associations between tobacco smoking and body weight, (2) describe potential mechanisms underlying the relationships between smoking and body weight, with a focus on mechanisms related to eating behaviors and appetite, and (3) discuss management of concomitant tobacco smoking and obesity.

Recent findings:

Adolescents who smoke tobacco tend to have body mass indexes (BMI) the same as or higher than nonsmokers. However, adult tobacco smokers tend to have lower BMIs and unhealthier diets relative to nonsmokers. Smoking cessation is associated with a mean body weight gain of 4.67 kg after 12 months of abstinence, though there is substantial variability. An emerging literature suggests that metabolic factors known to regulate food intake (e.g., ghrelin, leptin) may also play an important role in smoking-related behaviors. While the neural mechanisms underlying tobacco smoking-induced weight gain remain unclear, brain imaging studies indicate that smoking and eating cues overlap in several brain regions associated with learning, memory, motivation and reward. Behavioral and pharmacological treatments have shown short-term effects in limiting post-cessation weight gain; however, their longer-term efficacy is limited.

Summary:

Further studies are needed to identify the exact mechanisms underlying smoking, eating behaviors, and body weight. Moreover, effective treatment options are needed to prevent long-term weight gain during smoking abstinence.

Introduction

Tobacco smoking and obesity are leading causes of morbidity and mortality worldwide [1, 2]. Globally, more than 1.1 billion people 15 years of age and older are tobacco users [3••]. Tobacco smoking is harmful to nearly every organ in the body, and approximately 12% of all deaths among those 30 years of age and older can be attributed to tobacco smoking [4]. With regard to obesity, 650 million adults around the world are obese, defined as a body mass index (BMI) ≥30 kg/m² [5]. Similar to tobacco smoking, obesity increases the risk of negative health outcomes including cardiovascular disease, type 2 diabetes, and cancer [1, 2]. Further, the co-occurrence of obesity and tobacco smoking increases mortality and morbidity above either condition alone [6–8]. The life expectancy of individuals who are obese and tobacco smokers is 13 years less than normal weight non-smokers [9]. Thus, tobacco smoking, obesity, and their comorbidity are significant public health priorities.

This narrative review provides an overview of the relationships among tobacco smoking, eating behaviors, and body weight. The aims are to (1) examine the concurrent and longitudinal associations between cigarette smoking and weight, (2) describe potential mechanisms underlying the relationships between smoking and weight, with a focus on those related to eating behaviors and appetite, and (3) discuss management of co-occurring cigarette smoking and obesity. The review focused on the most current literature (from 2013 to 2019).

Epidemiology of Tobacco Smoking and Body Weight

Smoking and Body Weight

Adolescents (11-17 years of age), particularly girls, frequently report initiating and continuing to smoke cigarettes for weight control [10]. Compared to girls in the 5^th grade who had normal body weight, girls who were overweight or obese were more likely to initiate smoking through age 18 [11]. This relationship was not found for boys [11]. The average BMI for smokers under 18 years of age is typically the same as or higher than nonsmokers. A review of weight status and cigarette smoking in adolescents found that seven of 12 studies that examined boys found a positive association between body weight and cigarette smoking, whereas only five of 16 studies that examined girls found a positive association [12]. In a more recent study, data from 2,733 seventh, ninth, and 11^th graders from the Texas Adolescent Tobacco and Marketing Surveillance system showed that compared to boys of normal weight, boys who were obese had a higher odds of past 30-day cigarette smoking (AOR=4.52, 95% CI=1.31, 15.51) [13]. No association was found between obese and normal weight girls. In a longitudinal study of 6,683 adolescents from the National Longitudinal Survey of Adolescent Health, in female adolescents, each two-unit increase in BMI z-score was associated with a 4-5% increase in the probability of smoking in adulthood [14•]. Adolescent females with obesity (BMI z-score of 2) had a 23.6% probability of becoming a smoker during adulthood, while those who were of normal weight (BMI z-score of 0) had an 18.3% probability. Among female cigarette smokers, adolescent BMI was also associated with smoking frequency in adulthood [14]. In contrast to female adolescent tobacco smokers, no significant associations between smoking behaviors and body weight were found in male adolescent smokers.

Starting at young adulthood, smokers tend to have lower body weight than nonsmokers. The National Longitudinal Study of Adolescent to Adult Health followed a nationally representative cohort of 13,361 adolescents (mean age of 15.5 years) into young adulthood (mean age of 28.4 years) [15•]. In the sample, 44% were nonsmokers, 23% were early establishers who started smoking before age 14, 21% were late establishers who started smoking after age 14, and 12% were former smokers. After approximately 12 years of follow-up, as compared to age-matched non-smokers, early establishers had a BMI −1.27 points lower (95% confidence interval (CI)=−1.56, −0.98), late establishers a BMI −0.84 points lower (95% CI=−1.16, −0.52), and former smokers a BMI −0.63 (−0.93, −0.34) points lower [15].

Cross-sectional data indicate that adults who are current tobacco smokers tend to have lower BMIs than never or former smokers [16, 17]. Moreover, current tobacco smokers are less likely to be obese than never smokers (OR=1.09, 95% CI 1.08, 1.11) [18]. However, among those who smoke cigarettes, 26.3% also have obesity [19]. Thus, while on average, adult smokers tend to have lower body weights compared to non-smokers, a substantial number of individuals who smoke are obese.

Smoking Cessation and Weight Gain

Weight gain is a frequently cited barrier to smoking cessation. Weight gain after cessation is associated with smoking relapse, and 52% of women and 32% of men who have had a previous quit attempt report that weight gain is one of the primary reasons for smoking relapse [20]. Smoking cessation is associated with a mean increase of 2.85 kg at 3 months and 4.67 kg in body weight after 12 months of abstinence [21]. Most post-cessation weight gain occurs during the first 3 months of a quit attempt [21]. Pooled results from a meta-analysis of 35 prospective cohort studies demonstrated that abstinent smokers gained 4.1 kg compared to active smokers who gained 1.5 kg over an average of 5.2 years of follow-up [22••]. There is significant variability in weight gain during smoking cessation ranging from 16% of untreated abstinent smokers losing weight to 13% of abstinent smokers gaining more than 10 kg of body weight [21]. Predictors of higher weight gain include being obese, black, under 55 years of age, and a heavy smoker (≥25 cigarettes smoked per day) [19, 23, 24]. A recent study of 12,204 adults from the National Health and Nutrition Examination Survey examined smoking cessation-induced weight gain (i.e., difference in body weight gain between abstinent smokers and active smokers) over 10 years [19••]. A total of 65.2% participants were never smokers, 25.3% were active smokers, and 9.5% were abstinent smokers. At 10 years, active smokers gained 3.5 kg versus 8.4 kg for abstinent smokers (a smoking cessation-induced body weight gain of 4.9 kg). There was a dose response relationship between cigarettes smoked per day and body weight gain. Smoking cessation-induced body weight gain for those who smoked 1-14, 15-24, and ≥25 cigarettes per day was 2.0, 6.0, and 10.3 kg, respectively.

Mechanisms Mediating the Effects of Tobacco Smoking and Smoking Cessation on Body Weight

Cigarettes contain more than 4,000 different chemicals including at least 70 compounds known to cause cancer [25]. A typical cigarette contains approximately 10-15 mg of nicotine, a psychoactive molecule that reaches the brain about 10-20 seconds after inhalation [26]. Approximately 10% of nicotine contained in each cigarette is absorbed into the systemic circulation [27]. In addition to mediating the rewarding and reinforcing effects of tobacco, nicotine also plays a critical role in the effects of cigarette smoking on weight gain due to its ability to decrease appetite and increase metabolic rate [28, 29]. The mechanisms underlying smoking’s effects on energy balance are not fully understood [30]. Below we summarize evidence of the underlying relationships between tobacco smoking and body weight as related to food intake and appetite.

Eating Behaviors

In preclinical studies, nicotine exposure suppressed food intake, decreased meal size, and resulted in longer inter-meal intervals [31, 32]. Moreover, nicotine has been shown to have anorexic effects and alter feeding patterns in animal models [33]. While it is commonly recognized that cigarette smoking is an appetite suppressant, the effects of smoking on appetite in humans have not always been observed. Compared to nonsmokers, smokers report equal or higher intakes of dietary energy, total and saturated fat, and energy dense foods [34•–36]. Relative to nonsmokers or abstinent smokers, current smokers also report more frequent cravings for high-fat foods and fast-food fats [36]. Smokers who are obese have reduced perceptions of fat and sweetness compared to normal weight smokers and nonsmokers [37]. Smoking cessation is associated with increased caloric intake. During acute abstinence, smokers consume 152 more calories during an ad libitum snack meal compared to the total calories consumed during a snack meal while actively smoking [38]. After 6 to 8 weeks of smoking cessation, abstinent smokers consume about 122-227 kcals more [39, 40]. Increases in food intake may be related to several mechanisms. Smoking may have been helping to control overeating and binge eating [41]. Individuals who quit smoking may also substitute eating for the ‘hand-to-mouth’ behavior of smoking. Smoking may have also been used to cope with stress and negative emotions, and to increase reward [42]. When individuals quit smoking, they may turn to food for comfort.

Metabolic Hormones

Preclinical and clinical studies have begun to elucidate the relationships between cigarette smoking and homeostatic and hedonic control of food intake [43, 44, 30]. Studies have indicated a role for various metabolic hormones in tobacco craving, smoking relapse, and the rewarding properties of tobacco [43, 44, 30]. For example, ghrelin is a peptide secreted from the stomach and peripheral tissues. Ghrelin regulates meal initiation and increased circulating ghrelin is associated with increased hunger and food cravings [45–47]. Ghrelin receptors are expressed throughout the brain including nuclei known to play critical roles in the rewarding and reinforcing effects of food and drugs of abuse (e.g., ventral tegmental area) [43, 48, 49]. In addition to the effects of ghrelin on hunger, it has been postulated that ghrelin also augments drug-seeking behaviors, potentially by activating the mesolimbic dopamine system and promoting craving-induced drug seeking [50, 43]. However, the relationships between circulating levels of ghrelin and tobacco smoking-mediated behaviors are mixed. In preclinical studies, rats exposed to cigarette smoke have significantly higher plasma acyl ghrelin compared to controls [51]. Some studies have found that plasma ghrelin levels do not differ between tobacco smokers and nonsmokers [52–54]. Other studies have found that plasma concentrations of acetylated ghrelin, the active form of ghrelin that can bind to its cognate receptor growth hormone secretagogue receptor 1a [55], were significantly higher in smokers than nonsmokers [56, 57]. There are acute effects of smoking on ghrelin, and smoking cessation leads to decreased ghrelin levels [52]. Participants who smoked within 60 minutes prior to a blood sample had higher levels of ghrelin than those who had not smoked in the last 180 minutes [57], and ghrelin may continue to decline with longer-term smoking cessation [58]. Higher baseline ghrelin also predicts increased risk of smoking relapse at 4 weeks [59].

Results for other metabolic hormones have also been mixed. Glucagon-like peptide-1 (GLP-1) is secreted by the L cells of the small intestine and neurons in the nucleus tractus solitarius of the caudal brainstem.[60] This hormone helps to regulate food intake by inhibiting appetite centers in the brain, increasing satiety, and slowing gastric emptying [61]. GLP-1 pathways have roles in both homeostatic and hedonic food regulation [62]. GLP-1 receptor agonists have been shown to reduce nicotine-mediated behaviors in rodents such as nicotine-induced locomotion and nicotine consumption [63, 64••]. GLP-1 receptor knockout mice self-administer more nicotine than wild-type controls [64]. Collectively, these studies suggest that systemic administration of a GLP-1 receptor agonist may decrease the reinforcing and rewarding efficacy of nicotine.

Nicotine also leads to functional impairment and loss of β-cell function[65]. Smokers, compared to nonsmokers, have more insulin resistance, increased fasting plasma glucose, and decreased insulin sensitivity[66]. After smoking cessation, insulin sensitivity improves [67]. Leptin is an adipokine that suppresses food intake [68]. Serum leptin levels are lower in tobacco smokers compared to nonsmokers [69], but not after adjustment for BMI [70]. Leptin levels increase after smoking cessation [71, 72]. Plasma leptin is positively correlated with nicotine craving intensity during early nicotine abstinence [73, 74]. Further studies are necessary to clarify these inconsistencies and to investigate other metabolic and weight-regulating hormones that may be related to smoking behaviors, such as neuropeptide Y.

Neurobiological Pathways

The rewarding effects of food and nicotine are mediated, in part, by common neurobiological pathways [75, 76•]. This shared circuity suggests that nicotine exposure can influence food intake. Indeed, nicotine produces neuroadaptations that result in altered energy homeostasis and change the hedonic value of food [77]. Thus, nicotine may influence body weight by altering the reinforcing properties of food [78].

Neuroimaging studies have reported that brain regions involved in learning, memory, motivation, and reward such as the insula, amygdala, striatum, and orbital frontal cortex are activated by both food and smoking cues [79]. Smokers, relative to nonsmokers, had a greater neural response in the hypothalamus to a milk shake stimuli, suggesting that nicotine influences hypothalamic circuits [80]. Compared to adolescents who are never smokers, adolescents who were light smokers showed decreased activation to pleasurable foods in the insula, putamen, inferior frontal cortex, and Rolandic operculum [81]. Likewise, a study in adults found that smokers, relative to non-smokers, during exposure to favorite food cues, had diminished neural activation in the caudate, putamen, insula, thalamus, brainstem, and cerebellum [82]. During exposure to favorite-food imagery, smokers compared to nonsmokers showed lower connectivity in the supramarginal gyrus, but greater connectivity between the supramarginal gyrus and the corticostriatal-limbic system [83]. These results suggest that smokers may be desensitized to natural food-related rewards in motivation-reward regions of the brain.

Several neurotransmitters and neurophysiological mechanisms contribute to nicotine relapse and may also underlie the relationship between tobacco smoking and weight changes. For example, nicotine’s effects on the mesolimbic reward system and on dopamine transmission are known to mediate nicotine reinforcement [29]. Nicotine activates pro-opiomelanocortin (POMC) neurons, which help to control appetite and weight, through nicotinic acetylcholine receptors. Expression of POMC and subsequent activation of hypothalamic melanocortin 4 receptors were found to be critical to the anorectic effects of nicotine [84].

Smoking Treatments, Obesity, and Post-Cessation Weight Gain

Smoking cessation is associated with a number of benefits including decreased risks of mortality and smoking-related diseases [85]. Smoking cessation is typically recommended regardless of baseline weight. Guidelines also recommend that the primary emphasis is placed on smoking cessation and advise against simultaneously initiating weight loss and smoking cessation [86]. However, a major concern for many smokers is post-cessation weight gain [87]. Interventions such as lifestyle changes and pharmacotherapy may be effective during smoking cessation while also helping individuals achieve a healthier weight.

Lifestyle Interventions

Effects of smoking on lifestyle interventions.

Smoking appears to neither impede nor be beneficial to weight loss efforts based on lifestyle interventions. In the Look AHEAD study, participants in an intensive lifestyle intervention (ILI) who were current smokers had no differences in weight loss outcomes compared to never smokers and abstinent smokers [88•]. Participation in the ILI was not associated with compensatory smoking or a likelihood of quitting or relapsing. Similarly, in trials of other behavioral or dietary intervention strategies (low-carbohydrate, low-fat, Mediterranean), tobacco smokers and nonsmokers did not differ in weight loss [89, 90].

Effects of behavioral interventions to prevent post-cessation weight gain.

Lifestyle interventions combined with smoking cessation treatment have demonstrated some initial successes in reducing weight gain without impeding smoking abstinence. However, the long-term effects of these treatments are minimal [91]. A meta-analysis examined 10 randomized controlled trials that compared combined behavioral smoking treatment and behavioral weight control to behavioral smoking treatment alone [92]. Compared to smoking treatment alone, participants who received both smoking treatment and weight treatment showed increased abstinence (OR=1.29, 95% CI=1.01, 1.64) and reduced weight gain (g=−0.30, 95% CI=−0.63, −0.04) during short-term abstinence (<3 months) [91]. However, differences in abstinence and weight between combined interventions for smoking plus weight management and smoking treatment alone were not significant during long-term abstinence (>6 months). Study results suggest that a sequential treatment approach (i.e., weight management treatment after smoking treatment) is more effective for reducing weight gain following a quit attempt compared to simultaneous treatment or cessation treatment alone [93]. In addition, personalized weight management support appears to reduce post-cessation weight gain at 12 months (MD=−2.58 kg, 95% CI −5.11 to −0.55) without reducing abstinence [91]. Exercise interventions did not significantly reduce post-cessation weight gain at the end of treatment, but a significant reduction was found at 12 months (MD=−2.07 kg, 95% CI=−3.78, −0.36).[91]

Existing and Potential Pharmacotherapies

Effects of smoking cessation pharmacotherapies on body weight.

Clinical practice guidelines indicate that smoking cessation interventions should include tobacco dependence counseling, in addition to at least one FDA-approved medication (if feasible and not medically contraindicated) [94]. The first-line tobacco cessation pharmacotherapies include nicotine replacement therapies, a nicotinic acetylcholine receptor partial agonist (varenicline) and a norepinephrine & dopamine reuptake inhibitor/nicotinic acetylcholine receptor antagonist (bupropion) [95]. Compared to placebo, bupropion limited post-cessation weight gain at the end of treatment (treatment period ranged from 7-14 weeks; mean difference (MD)=−1.12 kg, 95% CI=−1.47, −0.77). However, results were not sustained at >6 months [91]. Varenicline, relative to placebo, significantly reduced post-cessation weight gain at the end of treatment (treatment period ranged from 6-12 weeks; MD=−0.41 kg, 95% CI=−0.63 to −0.19), but the effect was not maintained at 6 or 12 months [91]. Compared to varenicline, bupropion resulted in significantly less weight gain at the end of treatment (−0.51 kg, 95% CI=−0.93, −0.09). Nicotine replacement therapy, regardless of route of administration, resulted in less post-cessation weight gain at the end of treatment than placebo (median treatment length of 12 weeks; MD=−0.69 kg, 95% CI=−0.88, −0.51), but there was no evidence of an effect at 12 months [91]. A systematic review of 12 studies evaluated polydrug therapy (i.e., nicotine replacement therapy and either rimonabant, naltrexone, nicotine inhaler, bupropion, nortriptyline, selegiline, pralidoxime and ipratropium, or bupropion plus varenicline or naltrexone plus bupropion). In the short-term follow-up, 7 of the 12 studies had less post-cessation weight gain than those in the group of individual drugs or placebo [96]. The other studies showed no difference. The effect of combination therapy in the long-term was not well documented.

Obesity pharmacotherapies.

Currently, five medications are FDA-approved for chronic weight management when used as an adjunct to a reduced-calorie diet and physical activity. These include orlistat (a pancreatic and gastric lipase inhibitor), lorcaserin (a 5HT2c receptor agonist), phentermine-topiramate (a norepinephrine-releasing agent/gamma-aminobutyric acid receptor modulation agent), naltrexone-bupropion (an opioid antagonist/dopamine and norepinephrine reuptake inhibitor), and liraglutide (a GLP-1 agonist). These medications are approved for use in individuals with a BMI≥30 kg/m² or BMI≥27 kg/m² with an obesity-related co-morbidity such as hypertension or type 2 diabetes. Lorcaserin and naltrexone-bupropion have been tested in studies to limit post-cessation weight gain as will be discussed below.

Lorcaserin has demonstrated some efficacy at reducing post-cessation weight gain in short-term studies. An open-label, single arm, pilot study assessed the preliminary safety and efficacy of combining varenicline 1 mg twice daily and lorcaserin 10 mg twice daily among 20 individuals who were active tobacco smokers and overweight or obese. Smoking abstinence was 50% at week 12 and 30% at week 26. Among those with smoking abstinence, weight gain was +1.1±3.9 kg at week 12 [97•]. In a larger 12-week, randomized, double-blind placebo-controlled trial, 603 smokers with a BMI of 18.5 to 35 kg/m² who averaged at least 10 cigarettes/day were randomized to lorcaserin 10 mg once daily, lorcaserin 10 mg twice daily, or placebo. All participants received standardized smoking cessation counseling weekly. Abstinence rates for month 3 were 5.6%, 8.7% and 15.3% for placebo, lorcaserin daily, and lorcaserin twice a day, respectively [98•]. Changes in weight from baseline to week 12 were −0.01, −0.35, and −0.98 kg for the placebo, lorcaserin daily, and lorcaserin twice a day groups, respectively.

Bupropion combined with naltrexone has also been examined for post-smoking cessation weight gain. In an-open label study of naltrexone (32 mg/day) and bupropion (360 mg/day) plus behavioral counseling, 30 overweight or obese tobacco smokers had continuous abstinence rates of 40.7%. The average weight gain at week 24 was +0.4% of initial body weight at week 24, which was not significantly different from baseline [99]. A preliminary, open-label study compared naltrexone 25 mg/day combined with bupropion 300 mg/day to bupropion 300 mg/day alone. All participants received a psychosocial intervention. Continuously abstinent participants in the naltrexone bupropion group gained less weight (M=1.67 pounds) than those in the bupropion alone group (M=3.17 pounds), though the effect was not statistically significant [100]. Abstinence rates did not differ between groups. In a randomized, double-blind, parallel group design study, 121 treatment-seeking cigarette smokers were randomized to bupropion (300 mg/day) and naltrexone (50 mg/day) or bupropion (300 mg/day) and placebo. At 7 weeks, 54.1% of participants assigned to bupropion and naltrexone achieved abstinence, which was significantly greater than the 33.3% on bupropion placebo [101•]. However, this effect was not sustained at 6-months. There was no effect of medication, time, or their interaction on weight.

Bariatric Surgery

Bariatric surgery provides the most effective and durable treatment for obesity [102]. It also has benefits of improving obesity-related comorbidities such as type 2 diabetes, hyperlipidemia, and hypertension. Cigarette smoking is a predictor of poor post-surgical outcomes including serious complications, infection, prolonged intubation, and pneumonia [103, 104]. Hence, some bariatric surgery practices require smoking cessation prior to bariatric surgery. In a sample of 84 patients who had laparoscopic sleeve gastrectomy, at a median follow-up of 8.0 years, 46% of previously smoking patients successfully quit [105]. The 55.9% excess weight loss of the actively smoking participants did not differ significantly (p=0.16) from the 48.7% excess body weight loss among abstinent smokers.

Conclusion and Future Directions

Weight management in tobacco smokers is challenging, and further research is needed to examine mechanisms responsible for smoking cessation-related weight gain. In particular, studies are needed that examine the hormonal and neural mechanisms underlying these relationships. These studies may identify molecular substrates that could serve as targets for novel pharmacotherapies to treat nicotine addiction and post-cessation body weight gain. New medications are needed that are effective for promoting smoking abstinence and weight control. For example, testing other obesity medications for post-cessation weight gain (e.g., GLP-1 receptor agonists such as liraglutide) warrants further research as well as other integrated behavioral strategies that target smoking cessation and weight management.

Smokers often have poor dietary habits and are vulnerable to excess weight gain after smoking cessation. The underlying mechanisms linking smoking, eating behaviors, and weight regulation are not well understood. However, growing data support the potential of hormonal and neural mechanisms. Future studies are needed that examine these mechanisms as well as test novel treatments for co-occurring smoking and weight issues.