Effect of long-term changes in diet and exercise on plasma leptin concentrations

¹ From the Institute for Nutrition Research, University of Oslo; the Norwegian University of Sport and Physical Education, Oslo; and the Departments of Preventive Cardiology and Clinical Chemistry and the Life Insurance Companies, Institute for Medical Statistics, Ullevål University Hospital, Oslo.

² Supported by the Research Council of Norway (112770/320), the Throne-Holst Foundation, and the Freia Foundation.

³ Reprints not available. Address correspondence to JE Reseland, Institute for Nutrition Research, University of Oslo, PO Box 1046 Blindern, N-0316 Oslo, Norway. E-mail: j.e.reseland{at}basalmed.uio.no.

ABSTRACT  
Background: Although it is known that plasma leptin concentrations correlate with the amount of adipose tissue in the body, little information is available on the long-term effects on leptin concentrations of changes in diet and exercise.

Objective: We wanted to examine whether changes in dietary energy sources and exercise-mediated energy expenditure, alone or in combination, affect plasma leptin concentrations.

Design: In a randomized, 2 x 2 factorial trial, 186 men with metabolic syndrome were divided into 4 groups: diet, exercise, a combination of diet and exercise, and control. Data on dietary intake, physical fitness, and demographics were collected and plasma leptin concentrations were measured before and after a 1-y intervention period.

Results: Plasma leptin concentrations, body mass index, and fat mass decreased in association with long-term reductions in food intake as well as increased physical activity. By adjusting for either body mass index or fat mass, we observed a highly significant reduction in plasma leptin concentration after both the diet and the exercise interventions. There was no interaction between the interventions, suggesting a direct and additive effect of changes in diet and physical activity on plasma leptin concentrations.

Conclusion: Long-term changes in lifestyle consisting of decreased intake of dietary fat and increased physical activity reduced plasma leptin concentrations in humans beyond the reduction expected as a result of changes in fat mass.

Key Words: Leptin • diet • exercise • metabolic syndrome • men • weight loss • lifestyle

INTRODUCTION  
Robust biological mechanisms that resist changes in body fat content are responsible for the weight regain that usually follows weight loss, provided that food is available (1, 2). Several hormones play important roles in keeping body weight stable (2). Leptin is one of the newly discovered hormones that may be of marked importance in the regulation of body fat (3). This 16-kDa peptide is expressed and secreted in proportion to adipocyte size and number and circulates in plasma in a concentration highly correlated with body fat mass (3, 4–7). Administration of recombinant leptin to mice with mutations in the leptin gene indicated that leptin participates in the regulation of food intake and energy expenditure (8–10). However, because there are large variations in leptin concentrations among individuals with similar body compositions, it is likely that factors other than adipose mass influence plasma leptin concentrations (5, 6, 11–13). Potential modifiers of leptin concentrations are energy-yielding nutrients such as fatty acids, carbohydrates, proteins, and alcohol. Most studies published so far indicate that fasting and refeeding may change plasma leptin concentrations, whereas little is known about the effect of specific nutrients in humans (14).

Physical activity is important for long-term regulation of body weight, partly because it increases the resting metabolic rate (15, 16). Weight reduction after physical exercise is correlated with reductions in plasma leptin concentrations in obese middle-aged women (17). However, results regarding the effects of exercise on plasma leptin concentrations, independent of fat mass, are conflicting (18–20).

The aim of the present study was to examine whether improvement in the cardiovascular disease risk factor profile induced by changes in lifestyle among sedentary individuals (21–23) has any effects on plasma leptin concentrations. We measured plasma leptin concentrations in men with moderately elevated blood pressure and lipid concentrations who were assigned to the single or combined intervention of physical training and diet for 1 y.

SUBJECTS AND METHODS  
Study population and design
Samples and data were retrieved from the Oslo Diet and Exercise Study, a randomized, 2 x 2 factorial intervention trial (21, 22). The 2 interventions were physical exercise and dietary change, alone or in combination, lasting 1 y. The experimental design, recruitment of participants, and laboratory procedures are described in detail elsewhere (21, 22). The ethical principles of the Helsinki Declaration were followed and the trial was approved by the local ethical committee.

The trial included 186 men aged ( ± SD) 44.9 ± 2.5 y with mildly elevated diastolic blood pressure of 87.9 ± 8.1 mm Hg, plasma HDL-cholesterol concentration of 1.01 ± 0.17 mmol/L, triacylglycerol concentration of 2.28 ± 1.13 mmol/L, total cholesterol concentration of 6.3 ± 0.8 mmol/L, and body mass index (BMI; in kg/m²) of 28.6 ± 3.4 ( Dietary counseling was provided to the participants in the diet group and the combined diet and exercise group at the start of the study and then after 3 and 9 mo. The advice was individually tailored according to dietary habits and risk factor profile. Increased consumption of fish and fish products, vegetables, and fiber-rich products containing complex carbohydrates and reduced intake of saturated fat and cholesterol were recommended.

The exercise program entailed supervised endurance exercise, such as aerobics, circuit training, and fast walking and jogging, 3 times/wk. Each workout lasted 60 min. The exercise group and the combined diet and exercise group were not separated during training. Attendance at each workout was recorded, as was the physical activity some participants did at home. This corresponded to an average of 1.8 h/wk throughout the year. Furthermore, all participants were interviewed at the end of the trial about changes in physical activity habits. The diet group and the control group did not change their physical activity habits during the 1-y period.

Laboratory procedures
Blood samples were collected between 0800 and 1000 after the subjects had fasted overnight and abstained from smoking and after they were recumbent for 10 min. Furthermore, the participants were told to abstain from vigorous exercise for 4 d before blood sampling. Cardiovascular disease risk factors were assessed in each participant before and after the 1-y intervention by means of a standard clinical examination performed by a cardiologist.

Euglobulin clot lysis time was measured in fresh plasma. All other indexes were measured in batches at the end of the trial from samples that had been stored at -70°C (20). Insulin was measured by radioimmunoassay (Linco Research, St Charles, MO) (24). Intraassay CVs were estimated to be 8% for measurements in the range of 0–144 pmol, 7% for measurements between 144 and 576 pmol, and 9% for measurements >576 pmol. Interassay CVs were 11%, 14%, and 12% at low, medium, and high (>996 pmol) concentrations, respectively. Other components (glucose, lipids, and factor VII) were quantified by standard methods (21, 22). Plasma leptin was measured by competitive radioimmunoassay (Linco Research) with use of [¹²⁵I]leptin as a tracer (25). The intraassay CV was 5.5% and the interassay CV was 3.8%.

Blood pressure was measured by using automatic equipment (Vita-Stat blood pressure monitor; VitaStat Medical Services Inc, Bellevue, WA) after the subjects had rested supine for 10 min. Aerobic capacity was estimated directly by using a modified Balke test protocol (26). Body weights were measured by using a Lindel balance scale (Samhald, Klippan, Sweden) while the participants wore only underclothes. Heights were measured at the same time while subjects were not wearing shoes. Percentage body fat was measured by using a body-composition analyzer (Futurex-5000; Futurex Inc, Gaithersburg, MD) based on near-infrared interactance (27). The height, weight, frame size, and activity level of each person were entered into the body-composition analyzer. Subjects were seated with their right forearm supported on a table and optical density levels were measured at the arterial midline of the right biceps. Output, which was programmed by the manufacturer using their standard equation, was recorded as percentage body fat. Intrasubject variation was 1–2%. By analyzing the change in percentage body fat resulting from the intervention, we minimized the overestimation of body fat in lean subjects and underestimation in obese subjects. Maximal oxygen uptake ( Dietary assessment was accomplished by using an optical-mark-readable food-frequency questionnaire that has been extensively validated (28, 29). All participants completed the food-frequency questionnaire both before and after the intervention. On the basis of the given frequencies of use and the self-estimated portion sizes (according to given alternatives in household measures), daily intakes of energy and nutrients were calculated by using a food database and software system at the Institute for Nutrition Research, University of Oslo. Cod liver oil was included in the nutrient calculations. To assess compliance with diet, plasma concentrations of fatty acids were also measured at the start and end of the trial. Smoking habits were estimated by measurement of serum thiocyanate concentrations and by self-report.

Statistics
At baseline, the mean plasma leptin concentration was lower in smokers (8.2 ± 4.0 mg/L, n = 63) than in nonsmokers (10.3 ± 7.4 mg/L, n = 122) (P = 0.048). Because neither self-reported smoking status nor plasma thiocyanate concentrations changed significantly during the course of intervention, and because the smokers were randomly distributed in all 4 intervention groups, smokers were not treated differently from nonsmokers within the same intervention group.

Because leptin values were skewed to the left, the values were logarithmically transformed. Fat mass (defined as body weight multiplied by percentage body fat) was skewed to the right and thus transformed values were also used for this variable. Response variables (leptin and fat mass) were defined as the differences between log values at baseline and after 1 y of intervention. Analysis of leptin response adjusted for BMI or fat mass was done by the following per subject calculation:

RESULTS  
Effect of intervention on plasma leptin concentration
After dietary intervention (n = 101: diet group and combined diet and exercise group), plasma leptin concentrations were reduced (Table 1). The change in plasma leptin concentration after dietary intervention was significantly greater than that after 1 y of nondietary intervention (ie, in the control group and the exercise group; P < 0.001). Likewise, the exercise intervention (n = 105: exercise group and combined diet and exercise group) reduced plasma leptin concentrations and the change in plasma leptin concentration after exercise intervention was significantly different from that after nonexercise intervention (P = 0.002). Logarithmic transformation of the plasma leptin concentration did not alter the conclusions for either intervention (dietary compared with nondietary or exercise compared with nonexercise).

View this table:
TABLE 1. Plasma leptin concentrations, body mass index, and body fat mass at baseline and the change in these variables after a 1-y intervention¹  
BMI was lower after dietary intervention than after nondietary intervention (P < 0.001) and after exercise intervention than after nonexercise intervention (P = 0.005). Fat mass was also lower after dietary intervention (P < 0.001, calculated as percentage fat mass and kg body wt) and after exercise intervention [P = 0.002 (%) and P = 0.003 (kg)].

Change in the plasma leptin concentration correlated with change in BMI after both types of interventions (r = 0.397 for dietary intervention and r = 0.485 for exercise intervention, P < 0.001 for both). Interestingly, change in the plasma leptin concentration correlated with change in percentage body fat after exercise intervention (r = 0.229, P = 0.02) but not after dietary intervention (r = 0.156). After adjustment for BMI and fat mass, either as percentage fat or in kg body wt, plasma leptin concentrations were reduced after both dietary intervention (P < 0.001 for all adjustments) and exercise intervention (P < 0.001 for BMI and fat mass in kg; P = 0.002 for percentage fat).

There were no significant interactions between the dietary and exercise interventions when calculated as simple responses on a logarithmic scale for leptin (P = 0.32) and fat mass (P = 0.89) or when calculated as responses adjusted for BMI (P = 0.32), percentage fat (P = 0.37), and fat mass (P = 0.35). This finding indicates that both types of interventions had independent effects on plasma leptin concentrations.

Dietary intake
Baseline values for intakes of energy and nutrients and changes after 1 y of intervention are shown in Table 2. Dietary advice promoted reduced intakes of total energy, protein, fat (including most types of fatty acids), cholesterol, alcohol, and sugar in the diet group and the combined diet and exercise group. The reported reductions in intake of saturated fatty acids, monounsaturated fatty acids, and polyunsaturated fatty acids in the diet group were not significantly different from those in the combined diet and exercise group. The intake of 20–22-carbon n-3 fatty acids increased in both dietary intervention groups. The only significant change in dietary intake in the exercise group was a lower intake of fat as a percentage of total energy intake; intakes of other nutrients remained unchanged in the exercise and control groups during the 1-y intervention period.

View this table:
TABLE 2. Intakes of energy and nutrients¹  
Correlation between leptin and dietary intake
Neither baseline intake nor change in intakes of total energy, protein, carbohydrates, and cholesterol correlated with the plasma leptin concentration or with change in the plasma leptin concentration in any intervention group. At baseline, the plasma leptin concentration correlated positively with the intake of vitamin A (r = 0.152, P = 0.04) and negatively with the intake of alcohol (r = -0.155, P = 0.04) for all subjects (n = 180). The reduced intake of fat and fatty acids, including saturated, monounsaturated, and polyunsaturated fatty acids, correlated positively with change in the plasma leptin concentration in the diet group (data not shown).

Reductions in intakes of total energy and dietary fat correlated with change in the plasma leptin concentration after both dietary intervention and nonexercise intervention. Reductions in intakes of total energy and carbohydrate, but not change in fat intake, correlated with change in the plasma leptin concentration after exercise intervention (Table 3).

View this table:
TABLE 3. Correlation between change in nutrient intake and change in the plasma leptin concentration after 1 y of intervention¹  
Correlation between leptin and cardiovascular disease risk factors
We found a strong correlation at baseline between the plasma concentrations of leptin and insulin and between the plasma concentrations of leptin and glucose (both fasting and after a standard 75-g oral glucose load). No significant correlation was observed between plasma concentrations of HDL₂, HDL₃, LDL cholesterol, or lipoprotein(a) and leptin, either at baseline or after the intervention (data not shown).

Analysis by 2 x 2 factorial design showed a significant correlation between changes in the plasma concentrations of leptin and insulin (fasting) in all groups (Table 4). After exercise intervention, changes in the plasma insulin concentration (after a glucose load) and in cholesterol failed to correlate with change in the plasma leptin concentration, in contrast with the finding in the other groups. Change in the plasma concentration of glucose (fasting), the plasma concentration of apolipoprotein A-I, and
View this table:
TABLE 4. Correlation between change in hemostatic variables and plasma lipid concentrations and change in the plasma leptin concentration¹  

DISCUSSION  
This is the first study to investigate the effect of long-term changes in lifestyle components such as diet and exercise, and in particular the combination of the 2, on plasma leptin concentrations. Both interventions had a strong effect, reducing the plasma leptin concentration beyond the effect expected as a result of changes in body weight and fat mass.

After dietary intervention, leptin, BMI, and fat mass were reduced, and change in the plasma leptin concentration correlated with change in BMI. However, change in the plasma leptin concentration failed to correlate with change in percentage body fat resulting from dietary intervention, suggesting a direct effect of dietary fat on the plasma leptin concentration. Several explanations are possible for this effect of dietary fat. We assumed that there was no time bias in reported dietary intake because there was a 1-y period between each report; thus, participants were unlikely to have remembered the details of their first report when completing the food-frequency questionnaire the second time. During this study period, however, the participants might have been influenced by being participants in the study and may have underestimated their dietary intakes at the end of the intervention. The mechanism by which the accumulation of triacylglycerol influences the expression of leptin in adipose tissue is not known. In humans and in rodents, concentrations of circulating leptin are reduced after fasting and increased after overfeeding (30). During relatively short-term fasting and refeeding experiments, Kolaczynski et al (30) found that BMI values were not markedly altered, suggesting that the leptin concentration may be regulated by factors other than body fat. Leptin concentrations do not change after normal meal consumption (6), but dietary changes over a period of 1 y may modify plasma leptin concentrations and energy balance.

Reductions in intakes of fat and fatty acids in the combined diet and exercise group were similar to those in the diet group (Table 2); however, no correlation was found between change in the plasma leptin concentration and fat intake (as shown by the nondietary and exercise interventions; Table 3). This may suggest that increased physical activity overrides the correlation between the plasma leptin concentration and dietary fat intake. Physical activity may reduce leptin messenger RNA expression in rats (31) and lower the abdominal tissue leptin production rate in humans (32). In the present study, we observed a reduction in plasma leptin concentrations, BMI, and fat mass after exercise intervention; however, the effect on leptin concentration was also strongly significant after adjustment for either BMI or fat mass.

Lifestyle changes such as those in this randomized trial were reported previously to improve carbohydrate metabolism, reduce insulin resistance (33), and reduce blood pressure in persons with hypertension (23). Considine (34) hypothesized that changes in energy intake or expenditure may be detected by the adipocyte and thus influence synthesis of leptin via insulin, corticoids, and epinephrine. A training program that improves insulin sensitivity could alter leptin concentrations independently of adipose tissue mass (34). Exercise is often added to energy restriction in the treatment of obesity and it also has preventive effects on the development of diabetes (35, 36). The combined diet and exercise intervention reduced insulin resistance in our patients with metabolic syndrome (33) and we observed a strong correlation between change in the plasma leptin concentration and changes in insulin and glucose. Pasman and Saris (19) studied the effect of long-term exercise training on leptin concentration and concluded that regular exercise allows "resetting" of the leptin concentration so that a lower concentration can be maintained at a certain body fat content. In a study of well-trained runners, short-term exercise was found to have no detectable effect on serum leptin concentrations (18). Physical activity is known to affect sympathetic nerve signals (37). An increase in sympathetic nerve activity promotes down-regulation of plasma leptin concentrations (38), but the dynamics between changes in energy metabolism and leptin is not understood. We conclude that long-term diet and exercise interventions may have direct effects on the plasma leptin concentration beyond the effect expected due to changes in fat mass.

REFERENCES  

1. Leibel RL, Rosenbaum, Hirsch J. Changes in energy expenditure resulting from altered body weight. N Engl J Med 1995;332:621–8.
2. Campfield LA, Smith FJ, Burn P. Strategies and potential molecular targets for obesity treatment. Science 1998;280:1383–7.
3. Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM. Potential cloning of the mouse gene and its human homologue. Nature 1994;372:425–32.
4. Murakami T, Shima K. Cloning of rat obese cDNA and its expression in obese rats. Biochem Biophys Res Commun 1995;209:944–52.
5. Maffei M, Halaas J, Ravussin E, et al. Leptin levels in human and rodent: measurements of plasma leptin and ob RNA in obese and weight-reduced subjects. Nat Med 1996;1:1155–61.
6. Considine RV, Sinha MK, Heiman ML, et al. Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N Engl J Med 1996;334:292–5.
7. Funahashi T, Shimomura I, Hiraoka H, et al. Enhanced expression of rat obese (ob) gene in adipose tissue of ventromedial hypothalamus (VMH)-lesioned rats. Biochem Biophys Res Commun 1995;211: 469–75.
8. Pelleymounter M, Cullen MJ, Baker MB, et al. Effects of the obese gene product on body weight regulation in ob/ob mice. Science 1995;269:540–3.
9. Halaas JL, Gajiwala KS, Maffei M, et al. Weight-reducing effects of the plasma protein encoded by the obese gene. Science 1995;269: 543–6.
10. Campfield LA, Smith FJ, Guisez Y, Devos R, Burn P. Recombinant mouse OB protein: evidence for a peripheral signal linking adiposity and central neural networks. Science 1995;269:546–9.
11. McGregor GP, DeSaga JF, Ehlenz K, et al. Radioimmunological measurement of leptin in plasma of obese and diabetic human subjects. Endocrinology 1996;137:1501–4.
12. Hosoda K, Masuzaki H, Ogawa Y, et al. Development of a radioimmunoassay for human leptin. Biochem Biophys Res Commun 1996;221:234–9.
13. Weigle DS, Ganter S, Kuijper JL, Leonetti DL, Boyko EJ, Fujimoto WY. Effect of regional fat distribution and Prader-Willi syndrome on plasma leptin levels. J Clin Endocrinol Metab 1997;82:566–70.
14. Coleman RA, Herrmann TS. Nutritional regulation of leptin in humans. Diabetologia 1999;42:639–46.
15. Mæhlum S, Grandmontage M, Newsholme E, Sejersted OM. Magnitude and duration of excess postexercise oxygen consumption in healthy young subjects. Metabolism 1986;35:423–9.
16. Bahr R. Excess postexercise oxygen consumption—magnitude, mechanisms and practical implications. Acta Physiol Scand 1992; 144(suppl):1–70.
17. Kohrt WM, Landt M, Birge SJ Jr. Serum leptin levels are reduced in response to exercise training, but not hormone replacement therapy, in older women. J Clin Endocrinol Metab 1996;81:3980–5.
18. Hickey MS, Considine RV, Israel RG, et al. Leptin is related to body fat content in male distance runners. Am J Physiol 1996;271:E938–40.
19. Pasman WJ, Saris WHM. The relation of insulin and OB protein in trained and control, weight-reduced males. Obes Res 1996;4:14s (abstr).
20. Perusse L, Collier G, Gagnon J, et al. Acute and chronic effects of exercise on leptin levels in humans. J Appl Physiol 1997;83:5–10.
21. The ODES investigators. The Oslo Diet and Exercise Study (ODES): design and objectives. Control Clin Trials 1993;14:229–43.
22. Anderssen SA, Haaland A, Hjermann I, Urdal P, Gjelsdal K, Holme I. Oslo Diet and Exercise Study: a one year randomized intervention trial: effect on haemostatic variables and other coronary risk factors. Nutr Metab Cardiovasc Dis 1995;5:189–200.
23. Anderssen SA, Holme I, Urdal P, Hjermann I. Diet and exercise intervention have favourable effects on blood pressure in mild hypertensives: The Oslo Diet and Exercise Study (ODES). Blood Pressure 1995;4:343–9.
24. Birkeland KI, Torjesen PA, Eriksson J, Vaaler S, Groop L. Hyperinsulinaemia of type II diabetes is not present prior to the development of hyperglycemia. Diabetes Care 1994;17:1307–10.
25. Ma Z, Gingerich RL, Santiago JV, Klein S, Smith CH, Landt M. Radioimmunoassay of leptin in human plasma. Clin Chem 1996;42: 942–6.
26. Balke B. Optimale körperliche leistungsfähigkeit, ihre messung und veranderung infolge arbeitsermüdung. (The optimal physical capacity, measurements and changes following a working load.) Arbeitsphysiologie 1954;15:311–23 (in German).
27. McLean KP, Skinner JS. Validity of Futurex-5000 for body composition determination. Med Sci Sports Exerc 1992;24:253–8.
28. Nes M, Andersen LF, Solvoll K, et al. Accuracy of a quantitative food frequency questionnaire applied in elderly Norwegian women. Eur J Clin Nutr 1992;46:809–21.
29. Solvoll K, Lund-Larsen K, Søyland E, Sandstad B, Drevon CA. A quantitative food frequency questionnaire evaluated in a group of dermatologic outpatients. Scand J Nutr 1993;37:50–5.
30. Kolaczynski JW, Ohannesian JP, Considine RV, Marco CC, Caro JF. Response of leptin to short-term and prolonged overfeeding in humans. J Clin Endocrinol Metab 1996;81:4162–5.
31. Friedman JE, Ferrara CM, Aulak KS, et al. Exercise training down-regulates ob gene expression in the genetically obese SHHF/Mcc-fa^cp rat. Horm Metab Res 1997;29:214–9.
32. Racette SB, Coppack SW, Landt M, Klein S. Leptin production during moderate-intensity aerobic exercise. J Clin Endocrinol Metab 1997;82:2275–7.
33. Torjesen PA, Birkeland KI, Anderssen SA, Hjermann I, Holme I, Urdal P. Lifestyle changes may reverse development of the insulin resistance syndrome. Diabetes Care 1997:20:26–31.
34. Considine RV. Invited editorial on "Acute and chronic effects of exercise on leptin levels in humans." J Appl Physiol 1997;83:3–4.
35. Bogardus C, Ravussin E, Robbins DC, Wolfe RR, Horton ES, Sims EAH. Effects of physical training and diet therapy on carbohydrate metabolism in patients with glucose intolerance and non-insulin-dependent diabetes mellitus. Diabetes 1984;33:311–8.
36. Eriksson KF, Lindegarde F. Prevention of type 2 (non-insulin dependent) diabetes mellitus by diet and physical exercise. The 6-year Malmö feasibility study. Diabetologia 1991;34:891–8.
37. Björntorp P. Effects of physical training on blood pressure in hypertension. Eur Heart J 1987;8:71–6.
38. Slieker LJ, Sloop KW, Surface PL, et al. Regulation of expression of ob mRNA and protein by glucocorticoids and cAMP. J Biol Chem 1996;271:5301–4.