Long-term Use of Nicotine Gum Is Associated With Hyperinsulinemia and Insulin Resistance
Background Insulin sensitivity and cardiovascular risk profile were examined in 20 healthy, nonobese, middle-aged men who were long-term users of nicotine-containing chewing gum and in 20 matched control subjects who did not use nicotine.
Methods and Results Long-term use of nicotine-containing chewing gum was associated with insulin resistance and hyperinsulinemia. The degree of insulin sensitivity correlated negatively to the extent of nicotine use measured as plasma cotinine levels.
Conclusions These findings suggest that nicotine is the major constituent in cigarette smoke that leads to insulin resistance, metabolic abnormalities associated with the insulin resistance syndrome, and increased cardiovascular morbidity. Thus, the use of nicotine replacement therapy during smoking cessation should be transient and limited.
Nicotine-replacement therapy is an effective aid in smoking cessation programs.1 However, the metabolic effects of the long-term use of nicotine-replacement therapy are not well documented. In contrast, the effects of cigarette smoking have been studied extensively.
Smoking is associated with lower HDL cholesterol2 as well as an increased propensity for NIDDM.3 These abnormalities are well-known risk factors for cardiovascular disease and are reminiscent of those seen in conjunction with insulin resistance and IRS.4
A recent study5 has also shown that smoking decreases insulin sensitivity in the short term, and cross-sectional studies have shown that smokers are insulin resistant6 and exhibit glucose intolerance,7 postprandial lipid intolerance,8 and other markers of IRS.6 9 Furthermore, healthy, middle-aged smoking men have an increased prevalence of atherogenic small, dense LDL particles compared with nonsmoking control subjects of similar age and BMI (B. Eliasson, MD, et al, unpublished observations, 1996). The extent of the metabolic abnormalities related to IRS in male smokers is a function of smoking habits,9 and the beneficial effects of smoking cessation on HDL cholesterol10 and insulin sensitivity are seen after 8 weeks (B. Eliasson, MD, et al, unpublished observations, 1996).
The aim of this investigation was to examine the metabolic effects of long-term nicotine gum chewing to evaluate the potential role of nicotine in eliciting insulin resistance and IRS.
The methods used in this study have been described in detail previously.9 The protocol was approved by the Ethics Committee of Go¨teborg University.
Twenty middle-aged, healthy, and nonobese long-term NGCs were recruited via a newspaper advertisement. They all had been long-term NGCs for >11 months (mean±SD, 50±57 months). All but 3 were previous smokers. Three other subjects were previous users of wet snuff. All subjects had a BMI <27 kg/m2. The nonsmoking control subjects, who were carefully selected to have similar ranges of age, BMI, and BF, had not regularly smoked or consumed nicotine in any form for >20 years.
Subject characteristics are presented in Table 1⇓. Blood pressure, body weight, BMI, and WHR were recorded. Lean body mass was determined by the measurement of naturally occurring 40K in a whole-body counter as previously described.9
Samples for the determination of metabolic variables were taken after an overnight fast. Lipoprotein and hepatic lipase activities before and after a heparin bolus (100 U/kg IV) were determined. A euglycemic hyperinsulinemic clamp was performed for 120 minutes as previously reported.9 The degree of insulin sensitivity (M/I) was determined as the amount of glucose infused per kilogram of lean body mass per minute (GIR), corrected for steady state plasma insulin levels. The steady state blood glucose level during clamping was 4.9±0.1 mmol/L (mean±SD) in both groups.
Plasma nicotine and cotinine levels were determined for confirmation of nicotine abstinence the night before blood sampling and the euglycemic clamp. All subjects had nicotine levels <8 ng/mL. The laboratory methods used in the analyses of plasma samples have been described previously.9
Statistically, the results were tested for normality, and unpaired t tests or Mann-Whitney U tests were used accordingly. Linear regression analysis was used to analyze relations between data. Multiple regression analysis was performed to evaluate the independent relations between variables. StatView 4.5 (Abacus Concepts Inc) was used for all statistical calculations.
Anthropometric and metabolic data of NGCs and control subjects are presented in Tables 1 and 2⇑⇓. There were no differences in age, body weight, BF, or BMI, although NGCs had significantly higher WHRs. They also consumed more alcohol than the nonsmokers.
NGCs had greater plasminogen activator inhibitor-1 activity and higher fasting insulin and C-peptide levels (Table 2⇑). NGCs exhibited a lower M/I measured during the euglycemic clamp. During the clamp, NGCs suppressed FFA and C-peptide levels less than control subjects, and the steady state levels of insulin were also higher in NGCs (Fig 1A and 1B⇓⇓). There was a tendency toward higher total cholesterol, LDL cholesterol, and triglyceride levels in NGCs, although these differences were not significant.
Plasma cotinine levels, a measure of the extent of long-term nicotine consumption, were significantly and negatively correlated with M/I (Fig 2⇓). Correlations between M/I, percent BF, and metabolic variables are shown in Table 3⇓. As expected, M/I was negatively correlated to both fasting insulin and C-peptide levels as well as clamp steady state levels of insulin and C-peptide. M/I was also negatively correlated to steady state FFA levels during the clamp and to percent BF in the total study population, whereas a positive correlation was seen for HDL cholesterol and postheparin lipoprotein lipase levels in control subjects (Table 3⇓).
Because BF mass, alcohol, and nicotine consumption can affect M/I, these factors were analyzed in a multiple linear regression analysis in which M/I was used as the dependent variable. The regression model was statistically significant (r2=.29, P=.011), and the three independent variables were all statistically significant (alcohol: r=.38, P=.025; nicotine use or nonuse: r=−.33, P=.051; percent BF: r=−.31, P=.046). It should be noted that nicotine use and percent BF were negatively related to M/I whereas alcohol use was positively correlated.
This study shows that in healthy, middle-aged men, long-term exposure to smoke-free nicotine by the use of nicotine-containing chewing gum is associated with hyperinsulinemia and insulin resistance. Because insulin resistance and IRS are important risk factors, this finding is consistent with increased cardiovascular morbidity and mortality in long-term users of smokeless tobacco.11 Thus, these data support the view that nicotine is the major constituent in cigarette smoke that causes the metabolic aberrations linked to cardiovascular morbidity, ie, insulin resistance and various manifestations of IRS, including an increased risk for NIDDM.3 6 7 8 9
The M/I in NGCs was negatively correlated to nicotine intake, as reflected by plasma cotinine levels. This is in agreement with our recent data in smokers showing significant relations between smoking habits, degree of insulin resistance, and various manifestations of IRS.9
The euglycemic hyperinsulinemic clamp technique is the gold standard for measuring peripheral insulin sensitivity in vivo.12 13 To directly compare insulin sensitivity between different groups of individuals, it is necessary for the same plasma insulin levels to be reached. If this does not occur, because of different rates of insulin clearance or suppression of endogenous insulin production, GIR may be corrected for the different plasma insulin levels.12 This is illustrated in the present study, in which GIR, calculated as the mean amount of glucose infused between 90 and 120 minutes during the clamp divided by lean body mass, was similar for NGCs and the control group. During the clamp, however, NGCs had 34% higher insulin levels than the control group (79 and 59 mU/L, respectively). When we corrected for this, NGCs were significantly more insulin resistant than the control subjects, and the degree of insulin resistance was positively correlated to cotinine levels.
Taken together, our data show that NGCs had higher endogenous insulin secretion than the control group, both in the fasting state and during the euglycemic clamp. This is illustrated by their higher C-peptide levels, the best indicator of insulin secretion. The reason for the impaired suppression of endogenous insulin release during the euglycemic clamp in NGCs is unclear. However, the difference compared with the control group was of sufficient magnitude to completely account for the higher insulin levels seen in NGCs during the clamp. Interestingly, we have seen the same perturbation in smokers (B. Eliasson, MD, et al, unpublished observations, 1996). C-peptide suppression by insulin infusion previously has been demonstrated to be impaired in obese subjects,14 but the pathophysiological mechanisms are still unknown.
Fasting hyperinsulinemia is associated with atherosclerosis and IRS,4 although its role as an independent risk factor for cardiovascular disease has been questioned.15 Recent studies, however, have shown that fasting hyperinsulinemia is an independent risk factor for ischemic heart disease in men.16 It is likely that hyperinsulinemia is a compensatory response to insulin resistance.
We have shown recently that smokers have higher fibrinogen and triglyceride levels, lower HDL cholesterol, an increased proportion of atherogenic small, dense LDL particles, and higher postheparin hepatic lipase than nonsmokers (Reference 9 and B. Eliasson, MD, et al, unpublished observations, 1996). Differences in fibrinogen and lipoprotein levels were not found in NGCs, although a nonsignificant trend was discernible in this small group of individuals.
Smokers, like NGCs, also have greater plasminogen activator inhibitor-1 activity and higher fasting C-peptide levels as well as elevated FFA and C-peptide levels during the euglycemic clamp compared with a group of well-matched, nonsmoking control subjects (B. Eliasson, MD, et al, unpublished observations, 1996). In addition, smokers exhibit lipid intolerance with a delayed postprandial lipid elimination after a mixed test meal in spite of fasting normotriglyceridemia.8 Whether NGCs also exhibit this abnormality is unclear but may well be the case, because postprandial lipid intolerance seems to be closely linked to insulin resistance and IRS.17
It should be emphasized that long-term users of nicotine gum may also have different life styles that may contribute to the metabolic aberrations. These aspects were not specifically evaluated in this study. Alcohol consumption, however, was positively correlated with insulin sensitivity in the NGCs, which supports previous data showing a beneficial effect of moderate intake of alcohol on insulin sensitivity.18 In contrast, a negative relation was found between M/I and cotinine levels, which suggests a direct pathophysiological effect of nicotine consumption habits. This is also in agreement with experimental studies5 7 showing short-term negative effects of smoking on insulin sensitivity and glucose tolerance.
The present study shows that long-term use of nicotine-containing chewing gum in nonsmoking, middle-aged men is associated with insulin resistance, hyperinsulinemia, and other manifestations of IRS. The negative relation between insulin sensitivity and cotinine levels suggests that the use of nicotine-replacement therapy during smoking cessation should be transient and of limited extent.
Selected Abbreviations and Acronyms
|BMI||=||body mass index|
|FFA||=||free fatty acids|
|GIR||=||glucose infusion rate|
|M/I||=||insulin sensitivity index|
|NGC||=||nicotine gum chewer|
|NIDDM||=||non–insulin-dependent diabetes mellitus|
|WHR||=||waist-hip circumference ratio|
This study was supported by Pharmacia Sweden AB, Helsingborg, Sweden; the Swedish Medical Research Council (project B-3506); and the Arne and Inga Britt Lundberg Foundation. The excellent technical assistance of Margareta Lande´n is gratefully acknowledged.
- Received May 14, 1996.
- Revision received June 27, 1996.
- Accepted July 2, 1996.
- Copyright © 1996 by American Heart Association
Craig WY, Palomaki GE, Haddow JE. Cigarette smoking and serum lipid and lipoprotein concentrations: an analysis of published data. BMJ.. 1989;298:784-788.
Rimm EB, Chan J, Stampfer MJ, Colditz GA, Willett WC. Prospective study of cigarette smoking, alcohol use, and the risk of diabetes in men. BMJ.. 1995;310:555-559.
Reaven GM. Role of insulin resistance in human disease. Diabetes.. 1988;37:1595-1607.
Frati AC, Iniestra F, Raul Ariza C. Acute effect of cigarette smoking on glucose tolerance and other cardiovascular risk factors. Diabetes Care.. 1996;19:112-118.
Eliasson B, Attvall S, Taskinen MR, Smith U. The insulin resistance syndrome in smokers is related to smoking habits. Arterioscler Thromb.. 1994;14:1946-1950.
Stubbe I, Eskilsson J, Nilsson-Ehle P. High-density lipoprotein concentrations increase after stopping smoking. Br Med J.. 1982;284:1511-1513.
DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol.. 1979;237:E214-E223.
Stern MP. Diabetes and cardiovascular disease: the ‘common soil’ hypothesis. Diabetes.. 1995;44:369-374.
Jeppesen J, Hollenbeck CB, Zhou M-Y, Coulston AM, Jones C, Chen Y-DI, Reaven GM. Relation between insulin resistance, hyperinsulinemia, postheparin plasma lipoprotein lipase activity, and postprandial lipemia. Arterioscler Thromb Vasc Biol.. 1995;15:320-324.
Facchini F, Chen Y-DI, Reaven GM. Light-to-moderate alcohol intake is associated with enhanced insulin sensitivity. Diabetes Care.. 1994;17:115-119.