0021-972X/90/7106-1468$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1990 by The Endocrine Society
Vol. 71, No. 6 Printed in U.S.A.
Influence of Physical Activity on Insulin-Like Growth Factor-I in Healthy Younger and Older Men* ERIC T. POEHLMAN AND KENNETH C. COPELAND Division of Endocrinology, Metabolism, and Nutrition, Departments of Medicine and Pediatrics, University of Vermont College of Medicine, Burlington, Vermont 05405
fat distribution (waist to hip ratio, r = —0.45; waist to thigh
ABSTRACT. To examine the hypothesis that a lower level of physical activity influences the age-related decline in insulinlike growth factor-I (IGF-I), we measured serum concentrations in healthy nonobese younger and older men, characterized for maximal aerobic capacity (VO2 max) and energy expended in leisure time physical activity. To examine the independent influence of physical activity on IGF-I relative to other lifestyle variables, we also determined fat-free weight, percent body fat, body fat distribution (waist to hip and waist to thigh ratios), and habitual caloric intake in our population. IGF-I was 33% lower (P < 0.01) in older men than in younger men, inversely related to percent body fat (r = —0.55) and indices of upper body
ratio, r = —0.47), and positively related to VO2 max (r = 0.64)
and leisure time physical activity (r = 0.45; P < 0.01). IGF-I was not related to fat-free weight or daily caloric intake. After controlling for the effects of age by multiple regression analysis, VO2 max (r = 0.29) and leisure time physical activity (r = 0.24) were the sole factors independently related to IGF-I (P < 0.05). Our results suggest that multiple factors contribute to the agerelated decline in IGF-I. Lower levels of IGF-I in aging men are related at least in part to diminished physical activity. («/ Clin Endocrinol Metab 7 1 : 1468-1473, 1990)
P
LASMA levels of insulin-like growth factor-I (IGFI) decline with advancing age (1). The decline in IGF-I might reflect the effects of aging itself or the influences of age-associated changes in lifestyle. For example, coincident with aging, people show a decline in maximal aerobic capacity (2, 3), an increase in total and central deposition of body fat (4-6), a loss of fat-free weight (4), as well as a decline in total caloric intake (7). Short term periods of energy deficit are associated with a reduction in IGF-I (8-10), whereas short term energy surplus is associated with an increase in IGF-I (11). The macronutrient composition of the diet, particularly dietary protein, influences plasma concentrations of IGF-I (8, 10). We and other investigators have reported an inverse association of adiposity with IGF-I, independent of age (1,12). It is likely that lifestyle and environmental factors interact with the aging process to modulate IGFI, but it is unclear which factors independently contribute to the age-related decline in IGF-I. Of those few studies that have examined the relation Received May 3, 1990. Address all correspondence and requests for reprints to: Dr. Eric T. Poehlman, Division of Endocrinology, Metabolism, and Nutrition, Department of Medicine, University of Vermont College of Medicine, Burlington, Vermont 05405. * This work was supported by a grants from the NIA (AG-07857) and the American Association of Retired Persons Andrus Foundation (to E.T.P.), the General Clinical Research Center at the University of Vermont (RR-109), and Dr. Elliot Danforth, Jr. (DK-18535).
of IGF-I with adiposity and physical activity, none has relied on direct measures of body composition and maximal aerobic capacity (1, 12, 13). The current study was designed to examine the influence of physical activity on IGF-I. Because physical activity influences other lifestyle variables, it was also important to consider the influence of body composition, body fat distribution, and nutrient intake as potential modulators of the age-related decline in IGF-I within the context of the present study.
Materials and Methods Subjects Younger and older nonobese male participants were recruited through local newspapers and advertisements. Forty-two younger (18-36 yr) and 26 older (59-76 yr) men in excellent general health participated in this study. The effects of age and physical activity on norepinephrine kinetics and metabolic rate have previously been reported in this population (14, 15). Criteria for subject selection were as follows: no clinical symptoms or signs of heart disease, resting blood pressure below 140/90 mm Hg, normal resting electrocardiogram, normal electrocardiogram response to an exercise stress test, absence of any prescription or over the counter medication that could affect cardiovascular or metabolic function, no family medical history of diabetes or obesity, and weight stability (±2 kg) by medical history within the past year. All subjects were nonsmokers. The experimental protocol was approved by the Committee on Human Research for the Medical Sciences of the University of Vermont, and informed consent was obtained.
1468
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 22:25 For personal use only. No other uses without permission. . All rights reserved.
PHYSICAL ACTIVITY AND IGF-I Body composition, body fat distribution, and energy intake Body fat was estimated from body density by underwater weighing and simultaneous measurement of residual lung volume by helium dilution using the Siri equation (16). Fat-free weight (FFW) was estimated as total body weight minus fat weight. This measurement is highly reproducible in our laboratory (15), with a coefficient of variation of 4.9% performed under test-retest conditions in 28 individuals. Both the waist to hip circumference ratio (WHR) and the waist to thigh ratio (WTR), indices of upper body and lower body fat distribution, were measured, using a plastic measuring tape, in subjects standing erect and breathing quietly. The waist circumference measurements was taken as the ratio of the minimal circumference of the abdomen. Hip circumference was measured at the maximal gluteal protuberance of the buttocks. The thigh circumference was measured on the right leg at the level of the gluteal fold. All measurements were performed by the same investigator. Daily caloric and macronutrient intake was estimated from a 3-day food diary (2 weekdays, 1 weekend day), as previously described (17). Maximal aerobic capacity (V02 max) and leisure time physical activity (LTA) VO2 max was assessed by a progressive continuous exercise test to exhaustion on a motor-driven treadmill at a constant speed, as described previously (18). The reproducibility of VO2 max under test-retest conditions in 25 men has a coefficient of variation of 3.8% and an intraclass correlation of r = 0.94. LTA within the past year was assessed in a structured interview, using the Minnesota Leisure Time Physical Activity Questionnaire (19). Briefly, the energy expenditure during recreational activities is expressed as the ratio of the metabolic rate during activity divided by the basal metabolism, thus yielding an activity index expressed in kilocalories per day. The energy requirement of the activity index is then multiplied by the frequency, duration, and intensity of participation in that particular leisure time activity (i.e. walking, gardening, etc.) during the past year. Laboratory analysis Volunteers were admitted to the General Clinical Research Center the day before the measurement of IGF-I. Subjects did not engage in physical exercise for at least 36 h before measurements were performed. All volunteers were administered a weight-maintaining evening meal at 1800 h of mixed caloric composition (14% protein, 55% carbohydrate, and 31% fat). All measurements were performed the following morning, 12 h after the last meal, while the subject was recumbent. Serum IGF-I levels were measured by RIA after acid-ethanol extraction according to the method of Daughaday et al. (20), as described previously (21). IGF-I antiserum was obtained through the National Hormone and Pituitary Program (supplied by Drs. Van Wyk and Underwood, Chapel Hill, NC). IGFI (Amgen, Thousand Oaks, CA) was used as standard as well as tracer, after iodination with Na125I and purification with chloramine-T. The assay has a sensitivity of approximately 1 ng/mL, with intra- and interassay coefficients of variation of
1469
4% and 13%, respectively. This assay has been validated extensively (22). Values obtained after acid-ethanol extraction correlate highly with those obtained after G-50 acid chromatography and are approximately 2-fold higher than those obtained after assay of unextracted serum (21). Although acid-ethanol extraction does not eliminate all interference from circulating binding proteins, no subject included in the current study had renal or hepatic insufficiency, GH deficiency, or any other clinical condition known to be associated with alterations in plasma concentrations of IGF-I carrier proteins (23), and the assay methodology used was identical for all unknown samples. Statistical analysis An unpaired t test determined the significance of differences in the physical characteristics between the younger and older men. Pearson's correlation coefficients were used to assess the linear relationships among IGF-I, age, maximal aerobic capacity, body composition, fat distribution, and nutritional intake. Data were verified for normality of distribution and equality of variances, and no violations were noted. Multiple regression analysis was performed, in which the investigator controls the order of entry of the variables into the multiple regression model. The data were analyzed by entering age as the first independent variable and IGF-I as the dependent variable. After the entry of age into the regression model, partial correlations of IGF-I with other variables (e.g. VO2 max, percent body fat, caloric intake, etc.) were calculated to estimate their relationship with IGF-I independent of age. Similar analyses were performed by first entering VO2 max into the regression equation to estimate the relationship of IGF-I with the other variables independent of VO2 max. Partial correlations refer to the correlations between the residual values for one variable and residual variables for another, with both variables being predicted from a third variable (age or VO2 max in this study). All data are expressed as the mean ± SD.
Results The physical characteristics of the younger and older men are displayed in Table 1. Younger men were taller than older men (P < 0.05). No differences in total body weight were noted between groups. Percent body fat was lower in younger men (P < 0.01), whereas FFW was higher in younger men (P < 0.01) than in older men. VO2 max, expressed per kilogram of body weight, was higher (P < 0.01) in younger men than in older men, and this difference persisted when expressed in liters per min or per kg FFW (P < 0.01; data not shown). Calories expended in LTA were greater in younger than in older men (P < 0.01). Older men displayed higher waist to hip and waist to thigh ratios (P < 0.01). Daily energy intake was approximately 17% lower in older men compared to younger men. Percent intake of dietary fat was lower in younger men (P < 0.01), whereas percent intake of carbohydrates was higher in younger men (P < 0.01) than in older men. The serum levels of IGF-I were approximately 33%
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 22:25 For personal use only. No other uses without permission. . All rights reserved.
POEHLMAN AND COPELAND
1470
J C E & M • 1990 Vol 71 • No 6
TABLE 1. Physical characteristics of the younger and older men Younger (n = 42)
Variables Age (yr) Ht (cm) Wt (kg) % Body fat FFW (kg) VO2 max (mL/kg-min) LTA (kcal/day) WHR WTR Energy intake (Cal/day) Dietary fat (%) Dietary protein {%) Dietary carbohydrate (%)
25.4 ± 5.6 178.2 ± 6.5° 75.5 ± 10.7 10.2 ± 5.06 67.5 ± 7.96 55.5 ± 6.6* 424 ± 162* 0.86 ± 0.056 1.46 ± 0.09" 2931 ± 696" 30 ±6* 13.9 ± 3
54 ±8 6
Range 18-36 163-191 60-111 4-24.3 54.2-87.2 40.0-71.2 150-916 0.78-1.08 1.31-1.68 1277-4276 12-40 9-25 39-70
Older (n = 26) 67.0 ± 175.0 ± 78.4 ± 20.9 ± 61.7 ± 34.7 ± 286 ± 0.93 ± 1.74 ± 2436 ±
4.8 6.7 9.9 5.5 6.2 8.6 168 0.07 0.18 470
34 ± 5 14.3 ± 2
48 ±6
Range 59-76 165-189 59.2-99 10.1-31.3 49.2-74.3 19.1-51.0 22-560 0.81-1.07 1.39-2.14 1851-3452 22-41 11-18 35-61
Values are the mean ± SD. °P
Vol. 71, No. 6 Printed in U.S.A.
Influence of Physical Activity on Insulin-Like Growth Factor-I in Healthy Younger and Older Men* ERIC T. POEHLMAN AND KENNETH C. COPELAND Division of Endocrinology, Metabolism, and Nutrition, Departments of Medicine and Pediatrics, University of Vermont College of Medicine, Burlington, Vermont 05405
fat distribution (waist to hip ratio, r = —0.45; waist to thigh
ABSTRACT. To examine the hypothesis that a lower level of physical activity influences the age-related decline in insulinlike growth factor-I (IGF-I), we measured serum concentrations in healthy nonobese younger and older men, characterized for maximal aerobic capacity (VO2 max) and energy expended in leisure time physical activity. To examine the independent influence of physical activity on IGF-I relative to other lifestyle variables, we also determined fat-free weight, percent body fat, body fat distribution (waist to hip and waist to thigh ratios), and habitual caloric intake in our population. IGF-I was 33% lower (P < 0.01) in older men than in younger men, inversely related to percent body fat (r = —0.55) and indices of upper body
ratio, r = —0.47), and positively related to VO2 max (r = 0.64)
and leisure time physical activity (r = 0.45; P < 0.01). IGF-I was not related to fat-free weight or daily caloric intake. After controlling for the effects of age by multiple regression analysis, VO2 max (r = 0.29) and leisure time physical activity (r = 0.24) were the sole factors independently related to IGF-I (P < 0.05). Our results suggest that multiple factors contribute to the agerelated decline in IGF-I. Lower levels of IGF-I in aging men are related at least in part to diminished physical activity. («/ Clin Endocrinol Metab 7 1 : 1468-1473, 1990)
P
LASMA levels of insulin-like growth factor-I (IGFI) decline with advancing age (1). The decline in IGF-I might reflect the effects of aging itself or the influences of age-associated changes in lifestyle. For example, coincident with aging, people show a decline in maximal aerobic capacity (2, 3), an increase in total and central deposition of body fat (4-6), a loss of fat-free weight (4), as well as a decline in total caloric intake (7). Short term periods of energy deficit are associated with a reduction in IGF-I (8-10), whereas short term energy surplus is associated with an increase in IGF-I (11). The macronutrient composition of the diet, particularly dietary protein, influences plasma concentrations of IGF-I (8, 10). We and other investigators have reported an inverse association of adiposity with IGF-I, independent of age (1,12). It is likely that lifestyle and environmental factors interact with the aging process to modulate IGFI, but it is unclear which factors independently contribute to the age-related decline in IGF-I. Of those few studies that have examined the relation Received May 3, 1990. Address all correspondence and requests for reprints to: Dr. Eric T. Poehlman, Division of Endocrinology, Metabolism, and Nutrition, Department of Medicine, University of Vermont College of Medicine, Burlington, Vermont 05405. * This work was supported by a grants from the NIA (AG-07857) and the American Association of Retired Persons Andrus Foundation (to E.T.P.), the General Clinical Research Center at the University of Vermont (RR-109), and Dr. Elliot Danforth, Jr. (DK-18535).
of IGF-I with adiposity and physical activity, none has relied on direct measures of body composition and maximal aerobic capacity (1, 12, 13). The current study was designed to examine the influence of physical activity on IGF-I. Because physical activity influences other lifestyle variables, it was also important to consider the influence of body composition, body fat distribution, and nutrient intake as potential modulators of the age-related decline in IGF-I within the context of the present study.
Materials and Methods Subjects Younger and older nonobese male participants were recruited through local newspapers and advertisements. Forty-two younger (18-36 yr) and 26 older (59-76 yr) men in excellent general health participated in this study. The effects of age and physical activity on norepinephrine kinetics and metabolic rate have previously been reported in this population (14, 15). Criteria for subject selection were as follows: no clinical symptoms or signs of heart disease, resting blood pressure below 140/90 mm Hg, normal resting electrocardiogram, normal electrocardiogram response to an exercise stress test, absence of any prescription or over the counter medication that could affect cardiovascular or metabolic function, no family medical history of diabetes or obesity, and weight stability (±2 kg) by medical history within the past year. All subjects were nonsmokers. The experimental protocol was approved by the Committee on Human Research for the Medical Sciences of the University of Vermont, and informed consent was obtained.
1468
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 22:25 For personal use only. No other uses without permission. . All rights reserved.
PHYSICAL ACTIVITY AND IGF-I Body composition, body fat distribution, and energy intake Body fat was estimated from body density by underwater weighing and simultaneous measurement of residual lung volume by helium dilution using the Siri equation (16). Fat-free weight (FFW) was estimated as total body weight minus fat weight. This measurement is highly reproducible in our laboratory (15), with a coefficient of variation of 4.9% performed under test-retest conditions in 28 individuals. Both the waist to hip circumference ratio (WHR) and the waist to thigh ratio (WTR), indices of upper body and lower body fat distribution, were measured, using a plastic measuring tape, in subjects standing erect and breathing quietly. The waist circumference measurements was taken as the ratio of the minimal circumference of the abdomen. Hip circumference was measured at the maximal gluteal protuberance of the buttocks. The thigh circumference was measured on the right leg at the level of the gluteal fold. All measurements were performed by the same investigator. Daily caloric and macronutrient intake was estimated from a 3-day food diary (2 weekdays, 1 weekend day), as previously described (17). Maximal aerobic capacity (V02 max) and leisure time physical activity (LTA) VO2 max was assessed by a progressive continuous exercise test to exhaustion on a motor-driven treadmill at a constant speed, as described previously (18). The reproducibility of VO2 max under test-retest conditions in 25 men has a coefficient of variation of 3.8% and an intraclass correlation of r = 0.94. LTA within the past year was assessed in a structured interview, using the Minnesota Leisure Time Physical Activity Questionnaire (19). Briefly, the energy expenditure during recreational activities is expressed as the ratio of the metabolic rate during activity divided by the basal metabolism, thus yielding an activity index expressed in kilocalories per day. The energy requirement of the activity index is then multiplied by the frequency, duration, and intensity of participation in that particular leisure time activity (i.e. walking, gardening, etc.) during the past year. Laboratory analysis Volunteers were admitted to the General Clinical Research Center the day before the measurement of IGF-I. Subjects did not engage in physical exercise for at least 36 h before measurements were performed. All volunteers were administered a weight-maintaining evening meal at 1800 h of mixed caloric composition (14% protein, 55% carbohydrate, and 31% fat). All measurements were performed the following morning, 12 h after the last meal, while the subject was recumbent. Serum IGF-I levels were measured by RIA after acid-ethanol extraction according to the method of Daughaday et al. (20), as described previously (21). IGF-I antiserum was obtained through the National Hormone and Pituitary Program (supplied by Drs. Van Wyk and Underwood, Chapel Hill, NC). IGFI (Amgen, Thousand Oaks, CA) was used as standard as well as tracer, after iodination with Na125I and purification with chloramine-T. The assay has a sensitivity of approximately 1 ng/mL, with intra- and interassay coefficients of variation of
1469
4% and 13%, respectively. This assay has been validated extensively (22). Values obtained after acid-ethanol extraction correlate highly with those obtained after G-50 acid chromatography and are approximately 2-fold higher than those obtained after assay of unextracted serum (21). Although acid-ethanol extraction does not eliminate all interference from circulating binding proteins, no subject included in the current study had renal or hepatic insufficiency, GH deficiency, or any other clinical condition known to be associated with alterations in plasma concentrations of IGF-I carrier proteins (23), and the assay methodology used was identical for all unknown samples. Statistical analysis An unpaired t test determined the significance of differences in the physical characteristics between the younger and older men. Pearson's correlation coefficients were used to assess the linear relationships among IGF-I, age, maximal aerobic capacity, body composition, fat distribution, and nutritional intake. Data were verified for normality of distribution and equality of variances, and no violations were noted. Multiple regression analysis was performed, in which the investigator controls the order of entry of the variables into the multiple regression model. The data were analyzed by entering age as the first independent variable and IGF-I as the dependent variable. After the entry of age into the regression model, partial correlations of IGF-I with other variables (e.g. VO2 max, percent body fat, caloric intake, etc.) were calculated to estimate their relationship with IGF-I independent of age. Similar analyses were performed by first entering VO2 max into the regression equation to estimate the relationship of IGF-I with the other variables independent of VO2 max. Partial correlations refer to the correlations between the residual values for one variable and residual variables for another, with both variables being predicted from a third variable (age or VO2 max in this study). All data are expressed as the mean ± SD.
Results The physical characteristics of the younger and older men are displayed in Table 1. Younger men were taller than older men (P < 0.05). No differences in total body weight were noted between groups. Percent body fat was lower in younger men (P < 0.01), whereas FFW was higher in younger men (P < 0.01) than in older men. VO2 max, expressed per kilogram of body weight, was higher (P < 0.01) in younger men than in older men, and this difference persisted when expressed in liters per min or per kg FFW (P < 0.01; data not shown). Calories expended in LTA were greater in younger than in older men (P < 0.01). Older men displayed higher waist to hip and waist to thigh ratios (P < 0.01). Daily energy intake was approximately 17% lower in older men compared to younger men. Percent intake of dietary fat was lower in younger men (P < 0.01), whereas percent intake of carbohydrates was higher in younger men (P < 0.01) than in older men. The serum levels of IGF-I were approximately 33%
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 22:25 For personal use only. No other uses without permission. . All rights reserved.
POEHLMAN AND COPELAND
1470
J C E & M • 1990 Vol 71 • No 6
TABLE 1. Physical characteristics of the younger and older men Younger (n = 42)
Variables Age (yr) Ht (cm) Wt (kg) % Body fat FFW (kg) VO2 max (mL/kg-min) LTA (kcal/day) WHR WTR Energy intake (Cal/day) Dietary fat (%) Dietary protein {%) Dietary carbohydrate (%)
25.4 ± 5.6 178.2 ± 6.5° 75.5 ± 10.7 10.2 ± 5.06 67.5 ± 7.96 55.5 ± 6.6* 424 ± 162* 0.86 ± 0.056 1.46 ± 0.09" 2931 ± 696" 30 ±6* 13.9 ± 3
54 ±8 6
Range 18-36 163-191 60-111 4-24.3 54.2-87.2 40.0-71.2 150-916 0.78-1.08 1.31-1.68 1277-4276 12-40 9-25 39-70
Older (n = 26) 67.0 ± 175.0 ± 78.4 ± 20.9 ± 61.7 ± 34.7 ± 286 ± 0.93 ± 1.74 ± 2436 ±
4.8 6.7 9.9 5.5 6.2 8.6 168 0.07 0.18 470
34 ± 5 14.3 ± 2
48 ±6
Range 59-76 165-189 59.2-99 10.1-31.3 49.2-74.3 19.1-51.0 22-560 0.81-1.07 1.39-2.14 1851-3452 22-41 11-18 35-61
Values are the mean ± SD. °P
