Endocrine Effects of Tobacco Smoking

Konstantinos Tziomalos; Faidon Charsoulis


Clin Endocrinol. 2004;61(6) 

In This Article

Hypothalamic–Pituitary Axis

Over many years a large number of studies have demonstrated that exposure to cigarette smoke produces marked neuroendocrine changes. The effects of nicotine on the hypothalamic–pituitary axis (HPA) were first studied by Balfour (1989), who showed that nicotine was a potent activator of the HPA. The basic mechanism of action of nicotine on these systems appears to be contingent on its ability to mimic the effects of acetylcholine at selected central nicotinic acetylcholinergic receptors (Rosecransland & Karin, 1998). There are at least three types of nicotine binding sites in the hypothalamus (Fuxe et al., 1989). Nicotine stimulates neuronal firing directly by acting at the binding sites or indirectly by causing the release of acetylcholine or monoaminergic neurotransmitters (such as dopamine, noradrenaline and serotonin) (Pickworth & Fant, 1998). Nicotine can thus affect HPA function through a variety of paths (Rosecransland & Karin, 1998).

Cigarette smoking causes an acute increase in circulating levels of ACTH; intense smoking is necessary to induce these changes (Seyler et al., 1984; Pomerleau, 1992). It appears that it is the administration of nicotine and not other components of tobacco smoke that increases ACTH (Pickworth & Fant, 1998). However, neither the site of action of nicotine nor its precise mechanism of action on ACTH secretion has been elucidated. Experimental data from rats show that nicotine does not act directly at the hypothalamic paraventricular nucleus, the site of the corticotrophin-releasing factor neurones crucial to the regulation of ACTH. However, brainstem catecholaminergic regions projecting to the paraventricular nucleus showed a regionally selective and dose-dependent sensitivity to nicotine, particularly the noradrenergic/adrenergic nucleus tractus solitarius (Matta et al., 1998). Catecholamine release in response to nicotine, which could then affect pituitary secretion directly, has been demonstrated both in vitro (Hall & Turner, 1972; Westfall, 1974) and in vivo in animals (Armitage et al., 1966; Hillhouse et al., 1975; Giorguieff-Chesselet et al., 1979; Andersson et al., 1981, 1983; Fuxe et al., 1983; Tsagarakis et al., 1988). Therefore, in animals, it is brainstem catecholaminergic neurones that play a role in the central effect of nicotine on ACTH secretion. Additionally, it is possible that neurotransmitters other than noradrenaline may be involved in the acute ACTH response to systemic nicotine (Matta et al., 1998).

In contrast to the acute effect of smoking on ACTH levels, the latter are not altered in chronic smokers (del Arbol et al., 2000). This is probably due to desensitization of the central nicotinic cholinergic receptors involved (Fuxe et al., 1989).

Several of the symptoms of tobacco abstinence are characteristic of the stress response; the latter is associated with increases in plasma levels of ACTH (Selye, 1976). If tobacco withdrawal is a stressful event, nicotine abstinence would be expected to lead to increases in the plasma levels of ACTH. However, ACTH levels did not significantly increase during nicotine abstinence over ad libitum smoking levels (Pickworth et al., 1996).

Acute nicotine administration stimulates prolactin release (Wilkins et al., 1982; Rasmussen, 1995). However, serum prolactin levels are significantly lower in both male and female chronic smokers who smoke more than 10 cigarettes per day (Andersen et al., 1984). This apparent discrepancy could be explained by a similar mechanism to that with ACTH (Fuxe et al., 1989).

A wide body of literature has shown that smoking increases plasma levels of vasopressin (Fuxe et al., 1989; Pomerleau & Rosecrans, 1989; Pomerleau, 1992; al'Absi et al., 2003). It has been postulated that the acute release of vasopressin may be responsible for the acute pressor response after smoking (Pickworth & Fant, 1998). Symptomatic hyponatraemia associated with this effect of smoking in vasopressin levels has been reported, particularly in long-term psychiatric patients (Blum, 1984).

Considerable indirect evidence from studies in humans supports the notion that some responses to smoking are mediated by forebrain beta-endorphinergic opioid mechanisms (Karras & Kane, 1980; Gorelick et al., 1989). Furthermore, experimental data from rats suggest that while acute nicotine administration stimulates release of beta-endorphin from forebrain neurones (Davenport et al., 1990; Boyadjieva & Sarkar, 1997), chronic nicotine inhibits pro-opiomelanocortin gene expression and thus, probably, biosynthesis of beta-endorphin and other opiomelanocortins (Rasmussen, 1998). It can therefore be reasonably hypothesized that diminished forebrain beta-endorphin biosynthesis, in response to long-term nicotine exposure by chronic cigarette smoking, could favour continued acute self-administration of nicotine in order to induce acute release of available beta-endorphin, minimizing the opioid withdrawal that would otherwise occur due to tonically decreased beta-endorphin synthesis (Rasmussen, 1998).

Smokers are generally lighter than nonsmokers (Khosla & Lowe, 1971). The forebrain beta-endorphinergic system appears to have a primarily permissive role in regulating eating (Morley, 1989), so it would be reasonable to hypothesize that chronic nicotine-induced suppression of forebrain pro-opiomelanocortin neuronal activity may in part mediate the associated reduction in body weight (Rasmussen, 1998).

The circadian system and cigarette smoking also appear to interact (O'Hara et al., 1988). Smoking varies considerably over the course of a day for most smokers (Hasenfratz et al., 1992; Jarvik et al., 1993), and some data suggest that peak activity may be altered in smokers (Jacober et al., 1994). In addition, vacillating feelings of stress and arousal over the course of a day appear to influence smokers' behaviour, possibly indicative of circadian modulation (Parrott, 1993).