The suppression of appetite by nicotine has been attributed to various and complex effects of nicotine on peptides that regulate food intake and body weight, such as hypothalamic neuropeptides (neuropeptide Y, orexins), adipokines (leptin, adiponectin) and other metabolic hormones (ghrelin, GLP-1) [38,52]. We focus on studies conducted in humans and mention only a few animal studies. The latter involve peptides, which have not received much attention in regard to smoking or smoking cessation in humans. We also present associations between smoking status, levels of peptides, appetite and body weight. A summary of the literature we identified is presented in Table 1.

Neuropeptide Y (NPY), a peptide that increases food intake, is classified as part of the pancreatic polypeptide family. This peptide shares 70% of its sequence homology with peptide YY (PYY) and it is located in the arcuate nucleus of the hypothalamus [33]. Some studies carried out in animals support a suppression of NPY expression, induced by nicotine administration [53,54,55]. Other laboratories reported an increase in NPY mRNA and its peptide after chronic nicotine treatment [56]. It was also suggested that chronic exposure to nicotine upregulates the NPY expression in forebrain areas involved in the regulation of feeding [56] with a simultaneous down-regulation of the NPY receptors [57]. This reduced hypothalamic NPY receptor density in response to nicotine could explain the decrease in food intake in smokers [58]. Studies that have measured plasma NPY levels in humans found these levels to be either elevated [59] in previously non-smoking adults after smoking or unchanged [60] in habitual smokers after acute smoking. Hussain at al. conducted the first study to show a relationship between cigarette smoking, NPY and body weight in humans. Smokers had decreased body weight and lower levels of NPY and leptin compared to never-smokers, whereas quitters had significantly increased body weight and higher levels of NPY and leptin compared to smokers. NPY levels were significantly correlated with body weight, BMI and waist circumference. NPY levels remained, in contrast to leptin, significantly lower in smokers and higher in quitters, even after adjusting for anthropometric parameters [61]. These results imply that the weight modulating effects of cigarette smoke directly involve effects on NPY expression independently of leptin [61]. Later, Stadler et al. reported that fasting NPY levels increased 3 months after smoking cessation [49].

Orexin-A and Orexin-B are neuropeptides derived from prepro-orexin/hypocretin, the production of which and subsequent cleavage into orexins A and B are almost exclusively located in the area of lateral hypothalamus [57]. Similar to NPY, orexins are positive regulators of food intake. Therefore, it could be expected that orexin levels would decrease upon nicotine administration. Paradoxically, a dose-dependent increase in prepro-orexin mRNA production upon chronic nicotine administration was reported [62]. In a prospective cohort study where 200 quitters and 85 non-smokers were compared over the course of 4 weeks, a decline in orexin levels was reported during the initial withdrawal period (defined as 24 h of abstinence) among smokers who relapsed in the first 4 weeks [63].

Leptin is an adipose-tissue-derived hormone that regulates both satiety and thermogenesis. Leptin suppresses food intake by decreasing appetite after eating and increasing metabolic rate [64]. It is released from adipocytes in direct proportion to fat mass [65]. There are conflicting results among different studies that examined the relation between smoking or smoking cessation and plasma leptin levels. It has been suggested that smoking, via nicotine mechanisms, might modify the sensitivity of hypothalamic leptin receptors, modulate leptin biosynthesis and consequently reduce body weight by augmenting the effects of leptin in the brain and by enhancing leptin binding or increasing the sensitivity of downstream transduction cascades [38,66]. Epidemiological studies [61,66] in different ethnic groups showed that plasma leptin, sOb-R leptin receptor and free leptin were significantly lower in smokers (moderate or heavy, cigarettes or sheesha) than non-smokers [65,67,68]. Other studies showed elevated plasma leptin concentrations in smokers [69,70] and suggested that this may be due to an increase in adipose tissue secretion of leptin or a decrease in leptin clearance [71]. In a recent study, female smokers not only showed significantly higher serum levels of leptin compared to female non-smokers, but also compared to male smokers [72]. However, most of the studies noted a significant increase in serum leptin levels after smoking cessation [69,73,74,75]. This increase was positively correlated with an increase in body weight, BMI and body fat mass [61,64,71,73,74]. Gonseth et al. reported a significant increase in serum leptin levels one year after smoking cessation, parallel with substantial weight in the same time period. This finding is in discordance with the theoretical effects of leptin. One explanation may be the development of leptin resistance in smokers [76]. More recently, Komiyama et al. found no significant changes in leptin levels in individuals with no increase in waist circumference after smoking cessation, but a significant increase in leptin levels in individuals with abdominal obesity. These observations may indicate that abdominal obesity caused leptin resistance despite smoking cessation [77]. However, another publication did not find an increase in leptin levels following smoking cessation, despite the weight gain [49]. One of the reasons for such a discrepancy might be the interaction between smoking and other factors such as diet, exercise, hormones and host inflammatory responses, which may also impair regulation of the actions of leptin [68].

Adiponectin is a hormonal active adipocytokine secreted by adipose tissue, with insulin-sensitizing, anti-inflammatory and anti-atherogenic properties [65]. Oxidative stress and inflammatory cytokines produced by smoking, lipolysis induced by nicotine and increased consumption of circulating adiponectin, which accumulates in the injured vascular walls of smokers, are some of the mechanisms proposed to contribute to lower levels of adiponectin in smokers compared to non-smokers [46,71,78,79]. Smoking cessation improves the adiponectin profile in different time periods following cessation [80]. It has been reported in many studies that plasma concentrations of adiponectin were significantly lower in active smokers compared to non-smokers [81,82] and quitters [71,80,83,84] among both healthy persons and patients with coronary heart disease (CHD) [85]. In addition, a dose-dependent association between smoking intensity and adiponectin levels in current smokers has been observed [82]. Bergmann and Siekmeier found that in non-obese middle-aged women even moderate cigarette smoking adversely influenced the serum concentration of adiponectin, whereas among obese women there were no significant differences in adiponectin and leptin concentrations between smokers and non-smokers [65]. Efstathiou et al. reported that post-cessation adiponectin levels were significantly increased after two months in a healthy Greek population [86]. Otsuka et al. reported that plasma adiponectin levels in Japanese patients were elevated 6 months after smoking cessation [87]. In a study where early effects of smoking cessation were followed, serum adiponectin levels tended to increase 1 week after the end of treatment, but after 9 weeks, they were significantly decreased in weight gainers. In weight maintainers, adiponectin levels increased slightly after smoking cessation, but changes were not significant [88]. More recently, Komiyama et al. reported that serum adiponectin levels did not decrease 1 year after smoking cessation, despite weight gain, and increased abdominal obesity. However, in individuals with less abdominal obesity and a smaller increase in waist circumference, total adiponectin levels increased 1 year after cessation, showing the apparent beneficial effect of smoking cessation on adiponectin levels [77].

Ghrelin is a known appetite-stimulating hormone synthesized and secreted in the stomach and hypothalamus [33]. Studies have showed that smoking acutely increased plasma levels of ghrelin, an unexpected finding in terms of explaining the known anorectic effect of smoking [89], as well as the observation that plasma ghrelin levels decreased after smoking cessation [74]. In another study, plasma concentrations of acetylated ghrelin, but not total ghrelin, were significantly higher in smokers than non-smokers [90]. More recently, in a study with a large sample size, total ghrelin serum levels were positively associated with active smoking [91]. Earlier, Kokkinos et al. found that fasting plasma ghrelin concentrations were not different between male smokers and non-smokers. Furthermore, smoking did not provoke any short-term change in ghrelin levels in smokers, but it induced a decline in non-smokers. Thus, it was suggested that if the anorectic effect of smoking is ghrelin-induced, this effect may be present only in people not habituated to smoke exposure. In habitual smokers, desensitization of receptors due to prolonged nicotine exposure could blunt ghrelin suppression by short-term smoking [92]. The lack of effect of acute smoking on serum ghrelin levels in smokers in a recent study was in line with this observation and, similarly, smoking interventions do not appear to affect levels of obestatin, which is a hormone encoded from the same gene as ghrelin [7]. These observations were not in line with previous ones, in which an acute increase in obestatin levels in smokers taking a smoke had been reported [93]. Pilhatsch et al. measured both appetite and ghrelin and found no effect of nicotine administration on ghrelin levels in healthy non-smokers, despite the decreased subjective appetite induced by nicotine. Taken together, these findings suggest that ghrelin may not play a major role in nicotine-related energy homeostasis [94]. In 26 healthy normal-weight never-smokers, Kroemer et al. found that nicotine administration decreased correlations with ghrelin levels in the mesocorticolimbic system, while subjects fasted. However, caloric load increased the modulatory effects of ghrelin on food-cue reactivity, particularly in the ventromedial prefrontal cortex and the amygdala. This effect was stronger during nicotine sessions [95]. There are also studies that showed alterations in ghrelin levels after nicotine restriction, but focused primarily on the predictive value of ghrelin measurements on smoking relapse and are outside the scope of this review [52].

Enteroendocrine-cell (EEC) secretory hormones include PYY, glucagon-like peptide-1 (GLP-1) and cholecystokinin (CCK). A few studies have demonstrated the effect of smoking cessation on these hormones in humans [33]. PYY is an anorexigenic 36 amino-acid peptide, which is secreted by the enteroendocrine cells of the distal gastrointestinal tract. GLP-1 is a peptide produced and released postprandially by the intestinal enteroendocrine cells that acts on the hypothalamus to decrease food intake. CCK is a peptide released by enteroendocrine cells within the small intestine in response to mixed meals, acting as a satiety signal [96]. Fasting GLP-1, visfatin and PYY levels were unchanged 3 months after smoking cessation in a study conducted by Stadler et al. [49], and levels of gastric inhibitory polypeptide (GIP), GLP-1, amylin, insulin, PYY and pancreatic polypeptide (PP), measured after a meal challenge, were also not affected after 3 months of abstinence, according to Pancova et al. However, changes in incretin levels earlier after smoking cessation may have occurred [73]. PYY suppresses food intake, and although no significant change was observed on its levels after smoking abstinence, according to the above-mentioned studies, body weight showed a significant increase. More recently, Yanakkoulia et al. tried to explain the observed acute decrease in energy intake associated with acute smoking by investigating a series of hormonal factors; however, they failed to observe any significant difference in the changes in blood concentrations of CCK, GLP-1, ghrelin and obestatin over time between smoking and control condition [7].