The Effect of Intensive Robert S. Schwartz,
Endurance Exercise Training Young and Older Men
William P. Shuman,
on Body Fat Distribution
in
Valerie Larson, Kevin C. Cain, Gilbert W. Fellingham, James C. Beard,
Steven E. Kahn, John R. Stratton, Manuel D. Cerqueira, and ltamar B. Abrass Little is known about the effects of exercise interventions on the distribution of central and/or intra-abdominal (IA) fat, and until now there were no studies in the elderly. Therefore, in this study we investigated the effects of an intensive g-month endurance training program on overall body composition (hydrostatic weighing), fat distribution (body circumferences), and specific fat depots (computed tomography [Cl), in healthy young (n = 13; age, 26.2 2 2.4 years) and older (n = 15; age, 67.5 2 5.6 years) men. At baseline, overall body composition was similar in the two groups, except for a 9% smaller fat free mass in the older men (P < .05). The thigh and arm circumferences were smaller (P = .OOl and P < .05, respectively), while the waist to hip ratio (WHR) was slightly greater in the older men (0.92 + 0.04 v 0.97 + 0.04, P < .Ol). Compared with the relatively small baseline differences in body composition and circumferences, CT showed the older men to have a twofold greater IA fat depot (P < .OOl), 46% less thigh subcutaneous (SC) fat (P < .Ol), and 21% less thigh muscle mass (P < .OOl). Following endurance (jog/bike) trainiqg, both the young (+16%, P < .OOl) and the older men (+22%, P < .OOl) significantly increased their maximal aerobic power (Vo,max). This was associated with small but significant decrements in weight, percent body fat, and fat mass (all P < .OOl) only in the older men. Similarly, small decrements were noted in the waist (P < .OOl) and chest (P < .Ol) circumferences, as well as the WHR (P < .05) in the older men alone. On CT, the older men had greater than 20% decrements in the three central (IA, abdominal SC and chest SC) fat depots (all P < .OOl), and a 9% increment in thigh muscle mass (P < .Ol). The young men demonstrated significant decreases in IA (- 17%, P < .05), abdominal SC (- 10%. P < .05), and thigh SC (-20%. P < .Ol) fat depots. Except in the chest SC depot, the absolute change in a depot following endurance training was related to the initial size of the depot. We conclude that older men, who have a more central distribution of adiposity at baseline, had a preferential loss of fat from the central fat depots. It is possible, therefore, that endurance training will also allow preferential loss of central fat in other populations of subjects at risk for obesity-related metabolic complications and might produce impressive improvement in metabolic abnormalities, despite only a small loss of weight and fat. Copyright 0 1991 by WA Saunders Company
TH AGING there is a substantial decline in fat free mass; thus, an older individual who weighs the same as he/she did at a younger age, is likely to be “fatter.“’ Not only are the elderly fatter at any given weight or relative weight, they also have a more central distribution of adiposity.’ With rare exceptions,’ studies support the concept that the central distribution of fat in young and middle-aged individuals is a strong independent predictor of many obesity-related metabolic abnormalities, including (1) abnormal glucose tolerance, hyperinsulinemia, and diabetes mellitus4,5; (2) hyperlipidemia and reduced highdensity lipoprotein cholesterol concentration?; (3) hypertension’; (4) stroke’; (5) coronary artery disease7,Y;and (6) death.’ More recently, studies have focused on the intraabdominal (IA) fat depot as being uniquely important as a determinant of obesity-related complications.6.‘0 The importance of central obesity on the development of these same common age-related metabolic problems is less well studied. We have recently published data demonstrating that the absolute size of the IA fat depot, measured by computed tomography (CT), increases with age, and that a greater percentage of total adiposity resides in the IA depot even in healthy older men compared with young controls.” Despite the apparent importance of IA adiposity in obesity-related metabolic abnormalities and its possible importance in common age-related metabolic abnormalities, there is relatively little direct data on the effects of weight-reduction interventions (diet or exercise) on the IA fat depot.“,‘” In this study, we compared the effects of an intensive endurance exercise program on overall body composition, fat distribution, and specific fat depots in a group of 13 young male controls with a group of 15 healthy
w
Metabolism,
Vol40, No 5 (May), 1991:
pp 545-551
older men. We hypothesized that the previously noted increase in central and IA adiposity with aging was related to inactivity and would be preferentially reversed with endurance training. METHODS Subjects
Healthy, weight-stable, untrained young and older men taking no medication were recruited to participate in an intensive 6-month endurance training program. Subjects were screened with a complete medical history and physical examination, diet and exercise history, blood and urine chemistries (maxipanel, urinalysis, complete blood cell count), resting electrocardiogram, and a Bruce protocol maximal treadmill exercise test.” All of the older subjects had normal tomographic thallium 201 studies immediately and 3 hours following the Bruce treadmill protocol. All 17 young subjects who were screened passed the exercise treadmill test and entered From the Department of Medicine, Division of Gerontology and Geriattics, University of Washington, and the Seattle Veterans Affairs Medical Center, Seattle, WA. Supported by the National Institutes of Health and National Institute of Aging (PO1 AG06581). A portion of this work was also suppotted by the Geriatric Research, Education and Clinical Centerat the Seattle Veterans Affairs Medical Center and by the Medical Research Service of the Depatiment of Veterans Affairs. A portion of this work was conducted through the Clinical Research Centerfacility at the University of Washington Medical Center, supporred by the NIH (Grant No. RR-37). Address reprint requests to Robert S. Schwartz, MD, Division of Gerontology and Geriatlic Medicine, Harborview Medical Center, 325 Ninth Ave (.X4-87), Seattle, WA 98104. Copyright Q 1991 by W B. Saunders Company 00260495/9114005-0018$03.OOlO 545
SCHWARTZ ET AL
546
the program. Four of these were unable to complete the entire 6-month training program because of changes in job or school schedules. leaving 13 young subjects (age, 28.2 2 2.4 years; range, 24 to 31 years) who concluded the entire training program. Ten of the 13 young subjects who completed the entire study also had screening tomographic thallium studies and these were all normal. Forty-four of the older men initially recruited passed the laboratory screening. Despite no history of coronary artery disease, 22 (50%) of these otherwise healthy older men were found to have either fixed or reversible defects on their thallium exercise tests, or abnormal ST segment responses and were excluded from the study. Of the remaining 22 subjects: (1) two were eliminated because of significant ventricular ectopy during thallium exercise testing; (2) one was eliminated at the start of baseline testing when it was determined that he was much more active than he had originally admitted; (3) three potential subjects decided not to enter the program; and (4) one subject was dropped from the study after he developed a severe viral illness, associated with anorexia and weight loss, and was unable to maintain his usual training regimen. All of the remaining 15 older subjects (age, 67.5 + 5.8 years; range. 60 to 82 years) completed the entire h-month endurance training program. The data presented here represent the results of an endurance training intervention on a subset of subjects whose baseline data have been previously reported.”
Analytical Measurements Overall body composition was determined in the morning, following a 12-hour overnight fast, using the hydrostatic weighing technique.” With a minimum of six trials, the highest three weights that differed by less than 100 g were used. Residual volume was measured using the helium dilution technique. The average of three trials that differed by less than 150 mL was used. In our laboratory, underwater weighing to determine body composition (including the residual volume measurement) has a day-to-day coefficient of variation of 2.6%. Body circumferences were measured using the technique of Krotkiewski et al.’ The chest was measured at the nipple line with arms hanging at the sides during normal exhalation. The right upper arm was also measured at the nipple line. The waist circumference was measured at one third of the distance between the umbilicus and the xyphoid process (up from the umbilicus). The hip circumference was measured 4 cm below the superior, anterior aspect of the iliac crest. The right thigh circumference was measured one third of the way between the superior aspect of the patella and the superior anterior aspect of the iliac creast (up from the patella). The average of three separate measurements was used for all circumferential measurements. The day-to-day coefficients of variation of these circumference measurements are all less than 1% (0.3% to 0.9%). A limited (four slice) CT scan was performed to determine the size of specific subcutaneous (SC) and IA fat depots.‘” Scans were performed on a General Electric 9800 Scanner. with 120 kilovolt (peak), variable mA, 9.8-second scan time, lo-mm slice thickness, and large-body calibration settings. The scanner was calibrated before scans to insure a variation of less than ? 10 Hounsfield units (HU). Scans were made at the nipple line (chest SC), umbilicus (abdomen SC and IA), mid-thigh at a level halfway between the greater trochanter and the superior aspect of the patella (thigh SC), and greater trochanter (buttock SC). Because of problems with attenuation, the chest scan had to be performed with the arms held up over the head. Thus, scans of the arms were not obtained. The thigh SC fat depot measurement represents the sum of the right and left thighs together. With fat density defined as -50 to -250 HU, the cross-sectional area of fat in each depot was
determined using a density contour software program. Abdominal SC and IA (area confined within the transversalis fascia) depots were manually separated by the operator before reading the abdominal scan. Thigh muscle area was measured by subtracting the SC fat and bone areas from the total thigh area. Fat within the muscle was not evaluated separately, and thus was included in the thigh muscle area. The scans were all read by a single observer who was blinded to the age or condition (pre-post) of the subject. The coefficient of variation for reading the same scan on 10 separate days is 1.5%. Because of the radiation exposure, we have not specifically determined the day-to-day coefficient of variation of the CT measurements. Most of this variation is likely to be related to marking the level of the scan, a process that we have standardized with very good accuracy for the circumferential measurements. In the only published data on day-to-day reproducibility of the CT scan method, Sjostrom and Kvist, using double examinations, reported a variability less than l%.” Arterialized plasma eprinephrine and norepinephrine were measured in duplicate using a highly specific and sensitive single isotope radioenzymatic method.‘” The interassay coefficients of variation for the catecholamine assay in this laboratory are 6.5%’ and 12% at concentrations of 300 and 100 pg/mL. respectively. The intraassay coefficient of variation in less than 5% at either plasma concentrations.
Exercise Testing Before entry into the training program, subjects underwent a standard Bruce treadmill exercise testI to determine their maxima1 aerobic power (Vo2max). The subjects had previously been introduced to this exercise testing protocol by walking on the treadmill (through stage 3) with a mouthpiece in place. In addition. each subject had already undergone a, similar treadmill exercise test as part of the screening process. Vozmax was calculated from the oxygen consumed and the carbon dioxide produced. Subjects exercised to subjective exhaustion. A good maximal effort is usually defined as a respiratory quotient (RQ) at maximum exercise of greater than or equal to 1.15. The mean RQ at maximum was greater than or equal to 1.21 ? 0.05 for both groups before and after training.
Endurance Training Program Subjects participated in an intensive endurance training program for 27 weeks. The training program consisted of five exercise sessions each week under the direct supervision of a masters level exercise physiologist. Sessions began with stretching and a lominute warm-up period, and ended with a lo-minute cool-down period. The subjects began their walk/jog exercise at 50% to 60% of their heart rate reserve (HRR; 0.6 [maximum pulse - resting pulse] + resting pulse) as calculated from their maximal exercise test. Subjects were gradually increased (at 2-weekly intervals) to exercising 45 minutes at 85% of HRR. All subjects reached this final level of exercise intensity by the fourth month of the training program. Pulse rates were continuously monitored electronically at each session during the entire training program (Exersentry. Computer Instruments, Hempstead, NY). The young subjects attended a mean of 3.99 2 0.60 of the maximum five sessions per week (80% compliance), compared with 4.44 2 0.43 session per week (89% compliance) for the older men (P < .05).
Diet Subjects in this project were participants in a much larger overall study assessing and comparing the cardiovascular and metabolic responses to endurance training in young and older men. Therefore, all subjects were weight-stabilized (as outpatients) for 21 days
547
EXERCISE AND FAT DISTRIBUTION IN MEN
before and at the end of the 6-month training program. The weight stabilization diet was composed of real food with 50%, 30%, and 20% of calories as carbohydrate, fat, and protein, respectively. Over the last 7 days of the diet the mean variance in weight was 180 g ( < 26 g/d) and was not different before versus after training. To avoid a detraining effect, subjects continued to exercise during the second weight stabilization period. During the 6-month training program, subjects ate their own food but were asked not to change their usual diet. A 3-day diet history taken before and at the end of the exercise period showed no changes in the macronutrient composition of the diet. The body composition/fat distribution studies (hydrostatic weighing and circumferential measurements) were completed on day 4 and the CT scan on day 6 of this weight stabilization period.
Statistics Each subject acted as his own control and preipost endurance training differences were compared using paired t tests (twotailed). Comparisons between groups were made using nonpaired t tests (two-tailed). Data are expressed as means + SD. Not all data were available on all subjects. When the number of subjects analyzed is less than the total (13 young, 15 older). it is specifically indicated in the appropriate table. Because of the relatively small number of subjects. and thus the possibility of a non-normal distribution, these data were also analyzed using nonparametric statistics (Wilcoxon signed rank test and Wilcoxon rank sum test) with no change in the results. RESULTS
At baseline both the young and the older men were mildly obese using body mass index (BMI) criteria (Table 1). The two groups were comparable at baseline with respect to relative weight and body composition, expect for a larger fat free mass in the young subjects. The central (waist and hip) body circumferences were similar in the two groups at baseline (Table 2) while the older men had 8% smaller arm (P < .05) and 12% smaller thigh (P = .OOl) circumferences. There was a small but significant difference in baseline WHR (P < .Ol), with the older men being 5% greater. Consistent with other recent studies,19 most of the difference in WHR appeared to be due to the observed difference in the waist circumference. While the subcutaneous fat depots measured by CT were similar between the groups at the chest, buttock, and abdomen (Table 3) the older men had half the SC fat area at thigh (P < .Ol) and
twice the IA fat area (P < .OOl) as the young group. This was not due to the difference in overall body size between the two groups, as the older men continued to have twice as much IA fat even when expressed per meter of height or per square meter of surface area. The IA fat was also more than twice as great in the older men when expressed relative to fat mass, total SC fat, or SC abdominal fat (all P < .OOl). The older men were found to have 20% less mid-thigh muscle area compared with the young men. The observed difference between the groups would only be somewhat lessened by including in the muscle area calculation the fat that exists within the muscle. As expected, at baseline the Vozmax in the older men was 34% lower than in the young controls (Table 1). Nonetheless, these older men were probably as or more fit than their age-matched peers. At the end of the 6-month endurance training program, there was a mean 22% + 18% increase in Vo,max expressed per kilogram body weight in the older men (P < ,001) compared with an 18% 2 10% improvement in the young men (P < .OOl). Neither the absolute increases in Vo,max nor the percent increase from baseline were significantly different between the groups. The results are similar when expressed per kilogram of fat free mass. The endurance training program was associated with small but statistically significant changes in body weight and body composition in the older men with a 2.5kg loss in weight (P < .OOl), a 2.3% decrease in percent body fat (P < .OOl), and a 2.4-kg decline in body fat mass (P < .OOl), but no change in fat free mass. The young men had no significant loss in weight, percent body fat, or fat mass, but had a significant l.O-kg gain of fat free mass (P < .05). The observed changes in body composition were different between the groups with respect to weight (P = .06), BMI (P < .05), and fat free mass (P < .05). Following the endurance training program there were no significant changes in circumferential measurements or WHR in the young subjects (Table 2). However, in the older men, waist (P < .OOl) and chest (P < .Ol) decreased by 3% and WHR decreased by 2% (P < .05). Only the change in chest circumference was noted to be significantly
Table 1. Physical Characteristics and Overall Body Composition Young (n = 13)
Older (n = 15) Within-Group
MeE.Llre
PE
Post
Differences
PX
Post
Within-Group
Between-Group
Differences
Differences
I/ ~~
Weight (kg) BMI (kg/m’)
85.1 r 15.0 26.0 + 3.5
84.6 ? 13.4 25.9 _+3.2
79.6 + 7.9 26.2 ? 2.7
77.1 + 7.8 25.4 ? 2.8
% Body fat
22.3 ? 7.4
20.8 t 5.9
24.7 + 3.8
22.4 + 3.5
Fat mass (kg) Fat-free mass (kg)
19.7 f 8.9 65.4 & 8.6*
18.1 t 7.1 66.4 ? 8.2
19.8 + 4.6 59.8 + 5.2
17.4 2 4.1 59.7 f 5.2
II
0
3,004 + 425t
3,183 % 365
/i
2,626 2 278
2.772 + 272
5
44.1 + 5.1*
51.9 + 5.9
29.1 ? 4.4
35.4 -c 3.6
II
Calories (kcalid) \io,max (mL/kg/min)
NOTE. “Calories” refers to calories required for weight stabilization. lP 2 .05, tP
d 0.01,
§P 5 .05, IIP 2 ,001;
W
5 0.001; baseline comparisons between groups.
pre/post comparisons within each group.
#P < .05, comparison of pre/post differences between groups.
#
#
#
548
SCHWARTZ ET AL
Table 2. Body Fat Distribution by Circumferences Young
(n = 13)
Older
(n = 15) Within-Group
Within-Group Pre
Post
Hip (cm)
89.7 k 11.4 97.1 * 9.5
87.9 ? 10.0 95.8 ? 7.0
Chest (cm)
98.5 ? 8.6
98.3 t 9.1
Thigh (cm)
54.3 ? 4.6*
55.0 * 4.4
Ml?aSUre Waist
(cm)
Differences
PLT
Post
91.6 + 6.7 96.3 + 6.4
II
100.6 ? 5.4
98.1 & 5.5
ll
47.9 + 4.7
47.5 ? 3.8
94.8 2 6.1 97.7 * 5.4 (n = 13)
Between-Group
Differences
Differences
# #
Arm (cm)
33.3 + 2.9*
33.0 + 3.5
30.6 + 2.3 (n = 13)
30.2 ? 2.3
WHR
0.92 + 0.04t
0.92 + 0.05
0.97 + 0.04
0.95 * 0.03
5
*P 5 .05, tP 5 .Ol, SP 5 ,001; baseline comparisons between groups. §P I .05, llP s .Ol, jlP 5 ,001; pre/post comparisons within each group. #P I
.05, comparisons of pre/post differences between groups.
different between the two groups following endurance training (P < .05). In contrast to the relatively small changes observed in weight, body composition, and body circumferences following training, substantial and preferential changes in fat distribution were noted by CT (Table 3; Figs 1 and 2). The older group was found to have greater than 20% decrements in the more central abdominal SC, chest SC, and IA fat depots (all P < .OOl). These changes were not significantly different from the 14% decrement noted in the SC buttock fat (P < ,001). However, all of the above fat depots were reduced significantly more than the SC thigh fat, where no change was noted in the older men following training. The young subjects had a somewhat different pattern of change following exercise training, with significant decrements in IA (-17%, P = .05), SC abdominal (-9%, P = .05), and SC thigh (-20%, P < .Ol) fat depots, a trend for decrements in the SC chest (-9%, P < .06) and SC buttock (-lo%, P < .09), and no change in SC chest fat. The amount of IA fat relative to total fat mass decreased significantly only in the older men (-14%. P < .0.5). Despite no change in overall fat free mass in the older men following training, mid-thigh muscle area, measured on CT, increased by 9% (P < .Ol). This contrasts with the young subjects who had a small increment in overall fat free mass, but no change in mid-thigh muscle area. The percent changes in tissue area after training were significantly different between the two groups at four CT sites. The young group had a greater loss of thigh SC fat (P < .Ol); the older group had significantly greater decrements in fat area at the SC abdomen (P < .05) and chest Table 3. Young
(P < .05), and the older men had a greater increment in mid-thigh muscle area (P < .05). In all of the fat depots except SC chest, there is a significant correlation between the pretraining size of the depot and the amount of fat lost from the depot after training, with correlations ranging from 0.45 to 0.60. This means that subjects who started with a large amount of fat in the IA, SC abdominal, SC buttock, or SC thigh lost more fat from that depot than did subjects who started with a small amount of fat. However, the correlations between the pretraining values and the percent changes in these four depots are not significantly different from zero. Because age-related abnormalities in sympathetically mediated lipolysis might play a role in the central distribution of fat noted in the elderly, arterialized plasma epinephrine and norepinephrine concentrations were measured before and at the end of the training period. At baseline, epinephrine concentrations were not different between the two groups (58.7 ? 16.2 v 67.8 + 16.6 pg/mL, P = NS), while norepinephrine was higher in the older men (190.9 2 48.3 v 273.3 ? 90.9, P < .Ol). Following intensive endurance training there were no significant changes in epinephrine or norepinephrine concentrations in either group.
DISCUSSION
The effects of endurance training on body fat distribution (as measured by skinfolds and circumferences) have been previously reported in young subjects. Despres et al”” initially reported that a 20-week endurance training pro-
Body Fat Depot Areas by CT
(n = 13)
Older
(n = 15)
Within-Group Ml??BlNe
Pre
Post
Differences
Within-Group Pre
Post
Differences
54.8 ? 33.6
5
144.5 + 49.4
109.0 + 44.9
II
Abdominal SC (cm’) Chest SC (cm’)
218.7 2 119.5 92.9 ? 56.1 (n = 12)
197.9 + 103.9 90.1 -f 56.7
§
172.8? 41.8 115.3 ? 32.3
137.9 2 41.7 91.3 zk 34.5
1~
Buttock SC (cm’) Thigh SC (cm2)
182.3 ? 78.8 (n = 12) 154.6 + 75.8t
164.3 + 68.6 123.4 t 54.5
157.6 + 39.3 (n = 14) 80.5 ? 29.7
135.4 2 35.7 80.3 + 35.5
II
Thigh muscle (cm’)
345.8 + 56.7*
345.3 * 39.0
274.5 2 30.0
299.7 * 37.3
n
IA (cm’)
66.3 -t 37.1*
11
*P 5 .05, tP 5 .Ol, SP 5 ,001; baseline comparisons between groups. §P I .05, llP 5 .Ol, IIP 5 ,001; pre/post comparisons within each group. #P 5 .05, ##P
2 .Ol; comparison of pre/post differences between groups,
Between-Group Differences
##
#
EXERCISE AND FAT DISTRIBUTION
3 G 2
z .l-5
549
IN MEN
0
-20
-40
0
Older
-60 14
Abd-SO
Chest-SQ
Butt-Xl
Thigh-W
Thagh-Mus
Fig 1. Absolute changes in tissue areas measured by CT, following a B-month endurance training program in young (n = 13) and older (n = 15) men. Data are expressed as the mean k SEM. SD = SC. lP 5 .05. l*P 5 .Ol, l**p I ,001 pre/post comparisons within each group. ‘P < .05, ‘*P < .Ol comparison of pre/post differences between groups. Comparisons are adjusted for differences between groups in initial values.
gram in young non-obese men produced a 2.6-kg fat loss. This was associated with a 22% decrease in trunk skinfolds and only a 12.5% decrease in extremity skinfolds. This study noted that the changes in skinfolds were related to the initial size of the skinfold. In a 15week, high-intensity exercise training study, in non-obese young men,” these investigators again found a significantly greater reduction in truncal compared with extremity fat (27% v 15%) and thus a decrease in the ratio of extremity to truncal fat. Following a 1,000 kcalid exercise-induced energy deficit in moderately obese men, they noted a 6.8-kg loss of fat mass, a preferential loss of truncal versus extremity fat, and a greater reduction in total fat mass than SC fat, suggesting a loss of deep fat.‘j Another group has recently reported a signifi-
r
#
120 m
1
n q
IA
Young Older
Abd-SO
CL
Chest-SO Butt-SO Thigh-SD Thigh-Mus
Fig 2. Percent changes in tissue areas, measured by CT, following a B-month endurance training program in young (n = 13) and older (n = 15) men. Data are expressed as means t SEM. SO = SC. ‘P < 65, ‘“P < .Ol comparison between groups are adjusted for differences between groups in initial values.
cant decrease in the WHR in premenopausal women following a 6-month endurance training program.*’ Until now, no published study has evaluated the effects of endurance exercise on specific fat depots as measured by CT. In the present study, we noted a small but significant loss of body weight and total fat mass in the older men following an intensive 6-month endurance exercise program. In addition, the older men had a small but significant decrease in central (chest and abdominal) circumferences, while the more peripheral (arm, hip, and thigh) circumferences did not change. Contrary to the relatively small changes in body weight, total fat, and body circumferences, impressive (>20%) decrements were found in all three central fat depots with CT (SC chest and abdomen, and IA). Thus, there is a preferential loss of adiposity from central depots following intensive endurance training in older men. The pattern of exercise associated change in fat distribution in young men appears to be different than that observed in the older men, with a smaller loss of overall body fat and central fat, and a greater decrement in peripheral (thigh) fat. Some of the differences noted between the groups may be explained by the less central pattern of adiposity at baseline in the young. This is consistent with the previous finding that women, who have a more peripheral distribution of adiposity, appear resistant to exercise-induced changes in adiposity.” It is unlikely that the differences in the effects of training between the groups can be explained by differences in the training itself, since similar absolute and relative changes in Vo?max occurred in the two groups. Furthermore, the changes in caloric requirements for weight stabilization (Table 1) were similar in the two groups. The reasons for the preferential deposition of central fat with aging is unclear. Unlike premenopausal women who appear to have a strong relationship between central adiposity and testosterone concentrations,” androgenicity is unlikely to be a major determinant of the baseline central adiposity in older men, since androgen levels decrease with age.24 Catecholamines are important regulators of fat cell size and might play a role in the age-related predisposition toward central adiposity. In man, lipolysis is controlled by both p- (lipolytic) and (Y?-(antilipolytic) adrenergic receptor activity.‘5 The elevated catecholamine concentrations noted with aging in this study and previously” appear to cause homologous desensitization of the p-,” but not at,-adrenergi? receptor responsiveness. This could lead to a decrease in lipolytic activity or even an antilipolytic effect. Since normally IA fat appears to be very lipolytic,‘Y this depot might be most greatly affected by a relative decrease in p- to a,-receptor responsiveness, and a decrease in lipolysis. Another potential explanation for the central adiposity in older men is that adipose tissue lipoprotein lipase-induced triglyceride storage in central depots increases with aging. At present, there is no available data concerning this possibility. The reasons for the observed preferential loss of fat from central depots following endurance training are also unclear. It might be secondary to reversal of either an abnormal lipolytic or triglyceride storage pattern in the
SCHWARTZ ET AL
550
elderly. It is unlikely that exercise-associated changes in plasma catecholamine concentrations could explain the observed changes in fat depots, because neither epinephrine nor norepinephrine levels changed in either group with training. It is possible that a decrease in androgenicity with training may play a role in the observed decrement in central adiposity. A reduced testosterone level has been described in well-trained endurance athletes.‘” We noted relatively large changes in the central fat depots on CT (especially in the elderly), despite very small changes in overall adiposity. Although some of this may be accounted for by measurement error, the accuracy of our measures makes it unlikely that this can totally explain the difference. We hypothesize a redistribution of adiposity to areas not specifically measured on our limited number of scans. It is clear from the data of Sjostrom and Kvist” that changes in adiposity can be quite different in adjacent areas such as in the upper and lower abdomen. Whatever the cause of the decrement in central adiposity following endurance training, the magnitude of loss appears to be related to a more central distribution of adiposity at baseline. Since subjects with a more central
distribution of fat are most likely to suffer from obesityassociated metabolic abnormalities, it is possible that endurance exercise training may produce the greatest potential benefit in this at risk group, despite relatively small amounts of total weight loss. Although dietary weight loss studies in older individuals have not been published, most studies in young and middle-aged women show little or no change in central fat distribution (WHR) with dieting.“,” Therefore, exercise may be the preferred treatment for inactive patients with central obesity. Further support for this differential treatment concept must await more studies evaluating the comparative effects of weight loss diets and endurance exercise on fat distribution in different subject populations and the relationship between the observed changes in fat distribution and metabolic abnormalities.
ACKNOWLEDGMENT The authors would like to recognize the expert technical assistance of Leo Jaeger and Jan Wilson, as well as the skilled secretarial assistance of David Redick and Maria Limtiaco, in the preparation of the manuscript.
REFERENCES 1. Brozek J: Changes of body composition in man during maturity and their nutritional implications. Fed Proc 11:7X4-793, 1952 2. Borkan GA, Hults DE, Gerzof SG, et al: Age changes in body composition revealed by computed tomography. J Gerontol38:673677,1983 3. Garn SM, Sullivan TV, Hawthorne V: Evidence against functional differences between central and peripheral fat. Am J Clin Nutr 47:836-839, 1988
(eds): Fat Distribution During Growth and Later Health Outcome. New York, NY, Liss, 1988, pp 313-332 13. Despres J-P, Tremblay A, Nadeau A, et al: Physical training and changes in regional adipose tissue distribution. Acta Med Stand Suppl723:205-212,1988 14. Bruce RA, Kusumi F, Hosmer D: Maximal oxygen intake and normographic assessment of functional aerobic impairment in cardiovascular disease. Am Heart J 85:546-562, 1073
4. Kissebah AH, Vydelingum N, Murray R, et al: Relation of body fat distribution to metabolic complications of obesity. J Clin Endocrinol Metab 54:254-259, 1982
15. Goldman RF, Buskirk EK: A method for underwater weighing and determination of body density, in Brozek J, Henschel A (eds): Techniques for Measuring Body Composition. Washington, DC, National Academy of Sciences, 1961. pp 78-79
5. Krotkiewski M, Bjorntorp P, Smith U: Impact of obesity on metabolism in men and women. Importance of regional adipose tissue distribution. J Clin Invest 72:1150-1162, 1983
16. Shuman WP, Newell-Morris LL. Leoneth DL. et al: Abnormal body fat distribution detected by computed tomography in diabetic men. Invest Radio1 21:483-487, 1986
6. Peiris AN, Sothmann MS. Hoffmann RG, et al: Adiposity, fat distribution, and cardiovascular risk. Ann Intern Med 110:867-872. 1989
17. Sjostrom L. Kvist H. Regional body fat measurements with CT scan and evaluation of anthropometric predictions. Acta Med Stand SuppI723:169-177.1988
7. Gillum RF: The association of body fat distribution with hypertension, hypertensive heart disease, coronary disease, diabetes and cardiovascular risk factors in men and women aged 18-79 years. J Chronic Dis 40:421-428. 1987
18. Evans MI, Halter JB. Porte D Jr: Comparison of double- and single-isotope enzymatic methods for measuring catecholamines in human plasma. Clin Chem 24:567-570, 1978
8. Welin L, Svardsudd K, Wilhelmsen L. et al: Analysis of risk factors for stroke in a cohort of men born in 1913. N Engl J Med 317:521-526, 1987 9. Larsson B, Svardsudd K, Welin L, et al: Abdominal adipose tissue distribution, obesity and risk of cardiovascular disease and death: 13 year follow-up of participants in a study of men born in 1913. Br Med J 288:1401-1404,1984 10. Seidell JC, Bjorntorp P, Sjostrom L, et al: Regional distribution of muscle and fat mass in men-New insight into the risk of abdominal obesity using computed tomography. Int J Obesity 13:289-303, 1989 11. Schwartz RS, Shuman WP, Bradbury VL. et al: Body fat distribution in healthy young and older men. J Gerontol 45:M181M185,1990 12. Himes JH: Alteration in distribution of adipose tissue in response to nutritional intervention, in Bouchard C, Johnston FE
19. Shimokata H. Tobin JD, Muller DC, et al: Studies in the distribution of body fat: I. Effects of age, sex and obesity. J Gerontol44:M66-M73,1989 20. Despres J-P, Bouchard C. Tremblay A. et al: Effects of aerobic training on fat distribution in male subjects. Med Sci Sports Exert 17:113-118, 1985 21. Tremblay fat distribution Fat Distribution York, NY, Liss,
A, Despres J-P, Bouchard C: Alteration in body with exercise, in Bouchard C, Johnston FE (eds): During Growth and Later Health Outcomes. New 1988, pp 297-312
22. Krotkiewski M, Bjorntorp P: Muscle tissue in obesity with different distribution of adipose tissue. Int J Obesity 10:311-341. 1986 23. Peiris AN, Mueller RA, Struve MF, et al: Relationship of androgenic activity to splanchnic insulin metabolism and peripheral glucose utilization in premenopausal women. J Clin Endocrinol Metab 64:162-169. 1987
EXERCISE AND FAT DISTRIBUTION
IN MEN
24. Tenover JS, Matsumoto AM, Plymate SR, et al: The effects of aging in normal men on bioavailable testosterone and luteinizing hormone secretion: Response to clomiphene citrate. J Clin Endocrinol Metab 65:1118-1126, 1987 25. Mauriege P, Galitzky J, Berlan M, et al: Heterogeneous distribution of beta- and alphas-adrenoreceptor binding sites in human fat cells from various fat depots: Functional consequences. Eur J Clin Invest 17:156-165, 1987 26. Veith RC, Featherstone JA, Linares OA, et al: Age differences in plasma norepinephrine kinetics in humans. J Gerontol 41:319-324, 1986 27. Stiles GL, Caron MG, Letkowitz RJ: B-Adrenergic receptors. Biochemical mechanisms of physiologic regulation. Physiol Rev 64:661-743,1984
551
28. Pfeifer MA, Ward K, Malpass T. et al: Variations in circulating catecholamines fail to alter human platelet alpha?adrenergic receptor number or affinity for yohimbine or [‘HI dihydroergocryptine. J Clin Invest 74:1063-1068,1984 29. Rebuffe-Strive M, Anderson B, Olbve L, et al: Metabolism of adipose tissue in intraabdominal depots of nonobese men and women. Metabolism 38:453-458, 1989 30. Hackney AC: Endurance training and testosterone levels. Sports Med 8:117-127, 1989 31. Garrow JS: Is body fat distribution changed by dieting? Acta Med Stand Suppl723:199-203,1988 32. Wadden TA, Stunkard AJ, Johnston FE, et al: Body fat distribution in adult obese women. II. Changes in fat distribution accompanying weight reduction. Am J Clin Nutr 47~229-234, 1986
Endurance Exercise Training Young and Older Men
William P. Shuman,
on Body Fat Distribution
in
Valerie Larson, Kevin C. Cain, Gilbert W. Fellingham, James C. Beard,
Steven E. Kahn, John R. Stratton, Manuel D. Cerqueira, and ltamar B. Abrass Little is known about the effects of exercise interventions on the distribution of central and/or intra-abdominal (IA) fat, and until now there were no studies in the elderly. Therefore, in this study we investigated the effects of an intensive g-month endurance training program on overall body composition (hydrostatic weighing), fat distribution (body circumferences), and specific fat depots (computed tomography [Cl), in healthy young (n = 13; age, 26.2 2 2.4 years) and older (n = 15; age, 67.5 2 5.6 years) men. At baseline, overall body composition was similar in the two groups, except for a 9% smaller fat free mass in the older men (P < .05). The thigh and arm circumferences were smaller (P = .OOl and P < .05, respectively), while the waist to hip ratio (WHR) was slightly greater in the older men (0.92 + 0.04 v 0.97 + 0.04, P < .Ol). Compared with the relatively small baseline differences in body composition and circumferences, CT showed the older men to have a twofold greater IA fat depot (P < .OOl), 46% less thigh subcutaneous (SC) fat (P < .Ol), and 21% less thigh muscle mass (P < .OOl). Following endurance (jog/bike) trainiqg, both the young (+16%, P < .OOl) and the older men (+22%, P < .OOl) significantly increased their maximal aerobic power (Vo,max). This was associated with small but significant decrements in weight, percent body fat, and fat mass (all P < .OOl) only in the older men. Similarly, small decrements were noted in the waist (P < .OOl) and chest (P < .Ol) circumferences, as well as the WHR (P < .05) in the older men alone. On CT, the older men had greater than 20% decrements in the three central (IA, abdominal SC and chest SC) fat depots (all P < .OOl), and a 9% increment in thigh muscle mass (P < .Ol). The young men demonstrated significant decreases in IA (- 17%, P < .05), abdominal SC (- 10%. P < .05), and thigh SC (-20%. P < .Ol) fat depots. Except in the chest SC depot, the absolute change in a depot following endurance training was related to the initial size of the depot. We conclude that older men, who have a more central distribution of adiposity at baseline, had a preferential loss of fat from the central fat depots. It is possible, therefore, that endurance training will also allow preferential loss of central fat in other populations of subjects at risk for obesity-related metabolic complications and might produce impressive improvement in metabolic abnormalities, despite only a small loss of weight and fat. Copyright 0 1991 by WA Saunders Company
TH AGING there is a substantial decline in fat free mass; thus, an older individual who weighs the same as he/she did at a younger age, is likely to be “fatter.“’ Not only are the elderly fatter at any given weight or relative weight, they also have a more central distribution of adiposity.’ With rare exceptions,’ studies support the concept that the central distribution of fat in young and middle-aged individuals is a strong independent predictor of many obesity-related metabolic abnormalities, including (1) abnormal glucose tolerance, hyperinsulinemia, and diabetes mellitus4,5; (2) hyperlipidemia and reduced highdensity lipoprotein cholesterol concentration?; (3) hypertension’; (4) stroke’; (5) coronary artery disease7,Y;and (6) death.’ More recently, studies have focused on the intraabdominal (IA) fat depot as being uniquely important as a determinant of obesity-related complications.6.‘0 The importance of central obesity on the development of these same common age-related metabolic problems is less well studied. We have recently published data demonstrating that the absolute size of the IA fat depot, measured by computed tomography (CT), increases with age, and that a greater percentage of total adiposity resides in the IA depot even in healthy older men compared with young controls.” Despite the apparent importance of IA adiposity in obesity-related metabolic abnormalities and its possible importance in common age-related metabolic abnormalities, there is relatively little direct data on the effects of weight-reduction interventions (diet or exercise) on the IA fat depot.“,‘” In this study, we compared the effects of an intensive endurance exercise program on overall body composition, fat distribution, and specific fat depots in a group of 13 young male controls with a group of 15 healthy
w
Metabolism,
Vol40, No 5 (May), 1991:
pp 545-551
older men. We hypothesized that the previously noted increase in central and IA adiposity with aging was related to inactivity and would be preferentially reversed with endurance training. METHODS Subjects
Healthy, weight-stable, untrained young and older men taking no medication were recruited to participate in an intensive 6-month endurance training program. Subjects were screened with a complete medical history and physical examination, diet and exercise history, blood and urine chemistries (maxipanel, urinalysis, complete blood cell count), resting electrocardiogram, and a Bruce protocol maximal treadmill exercise test.” All of the older subjects had normal tomographic thallium 201 studies immediately and 3 hours following the Bruce treadmill protocol. All 17 young subjects who were screened passed the exercise treadmill test and entered From the Department of Medicine, Division of Gerontology and Geriattics, University of Washington, and the Seattle Veterans Affairs Medical Center, Seattle, WA. Supported by the National Institutes of Health and National Institute of Aging (PO1 AG06581). A portion of this work was also suppotted by the Geriatric Research, Education and Clinical Centerat the Seattle Veterans Affairs Medical Center and by the Medical Research Service of the Depatiment of Veterans Affairs. A portion of this work was conducted through the Clinical Research Centerfacility at the University of Washington Medical Center, supporred by the NIH (Grant No. RR-37). Address reprint requests to Robert S. Schwartz, MD, Division of Gerontology and Geriatlic Medicine, Harborview Medical Center, 325 Ninth Ave (.X4-87), Seattle, WA 98104. Copyright Q 1991 by W B. Saunders Company 00260495/9114005-0018$03.OOlO 545
SCHWARTZ ET AL
546
the program. Four of these were unable to complete the entire 6-month training program because of changes in job or school schedules. leaving 13 young subjects (age, 28.2 2 2.4 years; range, 24 to 31 years) who concluded the entire training program. Ten of the 13 young subjects who completed the entire study also had screening tomographic thallium studies and these were all normal. Forty-four of the older men initially recruited passed the laboratory screening. Despite no history of coronary artery disease, 22 (50%) of these otherwise healthy older men were found to have either fixed or reversible defects on their thallium exercise tests, or abnormal ST segment responses and were excluded from the study. Of the remaining 22 subjects: (1) two were eliminated because of significant ventricular ectopy during thallium exercise testing; (2) one was eliminated at the start of baseline testing when it was determined that he was much more active than he had originally admitted; (3) three potential subjects decided not to enter the program; and (4) one subject was dropped from the study after he developed a severe viral illness, associated with anorexia and weight loss, and was unable to maintain his usual training regimen. All of the remaining 15 older subjects (age, 67.5 + 5.8 years; range. 60 to 82 years) completed the entire h-month endurance training program. The data presented here represent the results of an endurance training intervention on a subset of subjects whose baseline data have been previously reported.”
Analytical Measurements Overall body composition was determined in the morning, following a 12-hour overnight fast, using the hydrostatic weighing technique.” With a minimum of six trials, the highest three weights that differed by less than 100 g were used. Residual volume was measured using the helium dilution technique. The average of three trials that differed by less than 150 mL was used. In our laboratory, underwater weighing to determine body composition (including the residual volume measurement) has a day-to-day coefficient of variation of 2.6%. Body circumferences were measured using the technique of Krotkiewski et al.’ The chest was measured at the nipple line with arms hanging at the sides during normal exhalation. The right upper arm was also measured at the nipple line. The waist circumference was measured at one third of the distance between the umbilicus and the xyphoid process (up from the umbilicus). The hip circumference was measured 4 cm below the superior, anterior aspect of the iliac crest. The right thigh circumference was measured one third of the way between the superior aspect of the patella and the superior anterior aspect of the iliac creast (up from the patella). The average of three separate measurements was used for all circumferential measurements. The day-to-day coefficients of variation of these circumference measurements are all less than 1% (0.3% to 0.9%). A limited (four slice) CT scan was performed to determine the size of specific subcutaneous (SC) and IA fat depots.‘” Scans were performed on a General Electric 9800 Scanner. with 120 kilovolt (peak), variable mA, 9.8-second scan time, lo-mm slice thickness, and large-body calibration settings. The scanner was calibrated before scans to insure a variation of less than ? 10 Hounsfield units (HU). Scans were made at the nipple line (chest SC), umbilicus (abdomen SC and IA), mid-thigh at a level halfway between the greater trochanter and the superior aspect of the patella (thigh SC), and greater trochanter (buttock SC). Because of problems with attenuation, the chest scan had to be performed with the arms held up over the head. Thus, scans of the arms were not obtained. The thigh SC fat depot measurement represents the sum of the right and left thighs together. With fat density defined as -50 to -250 HU, the cross-sectional area of fat in each depot was
determined using a density contour software program. Abdominal SC and IA (area confined within the transversalis fascia) depots were manually separated by the operator before reading the abdominal scan. Thigh muscle area was measured by subtracting the SC fat and bone areas from the total thigh area. Fat within the muscle was not evaluated separately, and thus was included in the thigh muscle area. The scans were all read by a single observer who was blinded to the age or condition (pre-post) of the subject. The coefficient of variation for reading the same scan on 10 separate days is 1.5%. Because of the radiation exposure, we have not specifically determined the day-to-day coefficient of variation of the CT measurements. Most of this variation is likely to be related to marking the level of the scan, a process that we have standardized with very good accuracy for the circumferential measurements. In the only published data on day-to-day reproducibility of the CT scan method, Sjostrom and Kvist, using double examinations, reported a variability less than l%.” Arterialized plasma eprinephrine and norepinephrine were measured in duplicate using a highly specific and sensitive single isotope radioenzymatic method.‘” The interassay coefficients of variation for the catecholamine assay in this laboratory are 6.5%’ and 12% at concentrations of 300 and 100 pg/mL. respectively. The intraassay coefficient of variation in less than 5% at either plasma concentrations.
Exercise Testing Before entry into the training program, subjects underwent a standard Bruce treadmill exercise testI to determine their maxima1 aerobic power (Vo2max). The subjects had previously been introduced to this exercise testing protocol by walking on the treadmill (through stage 3) with a mouthpiece in place. In addition. each subject had already undergone a, similar treadmill exercise test as part of the screening process. Vozmax was calculated from the oxygen consumed and the carbon dioxide produced. Subjects exercised to subjective exhaustion. A good maximal effort is usually defined as a respiratory quotient (RQ) at maximum exercise of greater than or equal to 1.15. The mean RQ at maximum was greater than or equal to 1.21 ? 0.05 for both groups before and after training.
Endurance Training Program Subjects participated in an intensive endurance training program for 27 weeks. The training program consisted of five exercise sessions each week under the direct supervision of a masters level exercise physiologist. Sessions began with stretching and a lominute warm-up period, and ended with a lo-minute cool-down period. The subjects began their walk/jog exercise at 50% to 60% of their heart rate reserve (HRR; 0.6 [maximum pulse - resting pulse] + resting pulse) as calculated from their maximal exercise test. Subjects were gradually increased (at 2-weekly intervals) to exercising 45 minutes at 85% of HRR. All subjects reached this final level of exercise intensity by the fourth month of the training program. Pulse rates were continuously monitored electronically at each session during the entire training program (Exersentry. Computer Instruments, Hempstead, NY). The young subjects attended a mean of 3.99 2 0.60 of the maximum five sessions per week (80% compliance), compared with 4.44 2 0.43 session per week (89% compliance) for the older men (P < .05).
Diet Subjects in this project were participants in a much larger overall study assessing and comparing the cardiovascular and metabolic responses to endurance training in young and older men. Therefore, all subjects were weight-stabilized (as outpatients) for 21 days
547
EXERCISE AND FAT DISTRIBUTION IN MEN
before and at the end of the 6-month training program. The weight stabilization diet was composed of real food with 50%, 30%, and 20% of calories as carbohydrate, fat, and protein, respectively. Over the last 7 days of the diet the mean variance in weight was 180 g ( < 26 g/d) and was not different before versus after training. To avoid a detraining effect, subjects continued to exercise during the second weight stabilization period. During the 6-month training program, subjects ate their own food but were asked not to change their usual diet. A 3-day diet history taken before and at the end of the exercise period showed no changes in the macronutrient composition of the diet. The body composition/fat distribution studies (hydrostatic weighing and circumferential measurements) were completed on day 4 and the CT scan on day 6 of this weight stabilization period.
Statistics Each subject acted as his own control and preipost endurance training differences were compared using paired t tests (twotailed). Comparisons between groups were made using nonpaired t tests (two-tailed). Data are expressed as means + SD. Not all data were available on all subjects. When the number of subjects analyzed is less than the total (13 young, 15 older). it is specifically indicated in the appropriate table. Because of the relatively small number of subjects. and thus the possibility of a non-normal distribution, these data were also analyzed using nonparametric statistics (Wilcoxon signed rank test and Wilcoxon rank sum test) with no change in the results. RESULTS
At baseline both the young and the older men were mildly obese using body mass index (BMI) criteria (Table 1). The two groups were comparable at baseline with respect to relative weight and body composition, expect for a larger fat free mass in the young subjects. The central (waist and hip) body circumferences were similar in the two groups at baseline (Table 2) while the older men had 8% smaller arm (P < .05) and 12% smaller thigh (P = .OOl) circumferences. There was a small but significant difference in baseline WHR (P < .Ol), with the older men being 5% greater. Consistent with other recent studies,19 most of the difference in WHR appeared to be due to the observed difference in the waist circumference. While the subcutaneous fat depots measured by CT were similar between the groups at the chest, buttock, and abdomen (Table 3) the older men had half the SC fat area at thigh (P < .Ol) and
twice the IA fat area (P < .OOl) as the young group. This was not due to the difference in overall body size between the two groups, as the older men continued to have twice as much IA fat even when expressed per meter of height or per square meter of surface area. The IA fat was also more than twice as great in the older men when expressed relative to fat mass, total SC fat, or SC abdominal fat (all P < .OOl). The older men were found to have 20% less mid-thigh muscle area compared with the young men. The observed difference between the groups would only be somewhat lessened by including in the muscle area calculation the fat that exists within the muscle. As expected, at baseline the Vozmax in the older men was 34% lower than in the young controls (Table 1). Nonetheless, these older men were probably as or more fit than their age-matched peers. At the end of the 6-month endurance training program, there was a mean 22% + 18% increase in Vo,max expressed per kilogram body weight in the older men (P < ,001) compared with an 18% 2 10% improvement in the young men (P < .OOl). Neither the absolute increases in Vo,max nor the percent increase from baseline were significantly different between the groups. The results are similar when expressed per kilogram of fat free mass. The endurance training program was associated with small but statistically significant changes in body weight and body composition in the older men with a 2.5kg loss in weight (P < .OOl), a 2.3% decrease in percent body fat (P < .OOl), and a 2.4-kg decline in body fat mass (P < .OOl), but no change in fat free mass. The young men had no significant loss in weight, percent body fat, or fat mass, but had a significant l.O-kg gain of fat free mass (P < .05). The observed changes in body composition were different between the groups with respect to weight (P = .06), BMI (P < .05), and fat free mass (P < .05). Following the endurance training program there were no significant changes in circumferential measurements or WHR in the young subjects (Table 2). However, in the older men, waist (P < .OOl) and chest (P < .Ol) decreased by 3% and WHR decreased by 2% (P < .05). Only the change in chest circumference was noted to be significantly
Table 1. Physical Characteristics and Overall Body Composition Young (n = 13)
Older (n = 15) Within-Group
MeE.Llre
PE
Post
Differences
PX
Post
Within-Group
Between-Group
Differences
Differences
I/ ~~
Weight (kg) BMI (kg/m’)
85.1 r 15.0 26.0 + 3.5
84.6 ? 13.4 25.9 _+3.2
79.6 + 7.9 26.2 ? 2.7
77.1 + 7.8 25.4 ? 2.8
% Body fat
22.3 ? 7.4
20.8 t 5.9
24.7 + 3.8
22.4 + 3.5
Fat mass (kg) Fat-free mass (kg)
19.7 f 8.9 65.4 & 8.6*
18.1 t 7.1 66.4 ? 8.2
19.8 + 4.6 59.8 + 5.2
17.4 2 4.1 59.7 f 5.2
II
0
3,004 + 425t
3,183 % 365
/i
2,626 2 278
2.772 + 272
5
44.1 + 5.1*
51.9 + 5.9
29.1 ? 4.4
35.4 -c 3.6
II
Calories (kcalid) \io,max (mL/kg/min)
NOTE. “Calories” refers to calories required for weight stabilization. lP 2 .05, tP
d 0.01,
§P 5 .05, IIP 2 ,001;
W
5 0.001; baseline comparisons between groups.
pre/post comparisons within each group.
#P < .05, comparison of pre/post differences between groups.
#
#
#
548
SCHWARTZ ET AL
Table 2. Body Fat Distribution by Circumferences Young
(n = 13)
Older
(n = 15) Within-Group
Within-Group Pre
Post
Hip (cm)
89.7 k 11.4 97.1 * 9.5
87.9 ? 10.0 95.8 ? 7.0
Chest (cm)
98.5 ? 8.6
98.3 t 9.1
Thigh (cm)
54.3 ? 4.6*
55.0 * 4.4
Ml?aSUre Waist
(cm)
Differences
PLT
Post
91.6 + 6.7 96.3 + 6.4
II
100.6 ? 5.4
98.1 & 5.5
ll
47.9 + 4.7
47.5 ? 3.8
94.8 2 6.1 97.7 * 5.4 (n = 13)
Between-Group
Differences
Differences
# #
Arm (cm)
33.3 + 2.9*
33.0 + 3.5
30.6 + 2.3 (n = 13)
30.2 ? 2.3
WHR
0.92 + 0.04t
0.92 + 0.05
0.97 + 0.04
0.95 * 0.03
5
*P 5 .05, tP 5 .Ol, SP 5 ,001; baseline comparisons between groups. §P I .05, llP s .Ol, jlP 5 ,001; pre/post comparisons within each group. #P I
.05, comparisons of pre/post differences between groups.
different between the two groups following endurance training (P < .05). In contrast to the relatively small changes observed in weight, body composition, and body circumferences following training, substantial and preferential changes in fat distribution were noted by CT (Table 3; Figs 1 and 2). The older group was found to have greater than 20% decrements in the more central abdominal SC, chest SC, and IA fat depots (all P < .OOl). These changes were not significantly different from the 14% decrement noted in the SC buttock fat (P < ,001). However, all of the above fat depots were reduced significantly more than the SC thigh fat, where no change was noted in the older men following training. The young subjects had a somewhat different pattern of change following exercise training, with significant decrements in IA (-17%, P = .05), SC abdominal (-9%, P = .05), and SC thigh (-20%, P < .Ol) fat depots, a trend for decrements in the SC chest (-9%, P < .06) and SC buttock (-lo%, P < .09), and no change in SC chest fat. The amount of IA fat relative to total fat mass decreased significantly only in the older men (-14%. P < .0.5). Despite no change in overall fat free mass in the older men following training, mid-thigh muscle area, measured on CT, increased by 9% (P < .Ol). This contrasts with the young subjects who had a small increment in overall fat free mass, but no change in mid-thigh muscle area. The percent changes in tissue area after training were significantly different between the two groups at four CT sites. The young group had a greater loss of thigh SC fat (P < .Ol); the older group had significantly greater decrements in fat area at the SC abdomen (P < .05) and chest Table 3. Young
(P < .05), and the older men had a greater increment in mid-thigh muscle area (P < .05). In all of the fat depots except SC chest, there is a significant correlation between the pretraining size of the depot and the amount of fat lost from the depot after training, with correlations ranging from 0.45 to 0.60. This means that subjects who started with a large amount of fat in the IA, SC abdominal, SC buttock, or SC thigh lost more fat from that depot than did subjects who started with a small amount of fat. However, the correlations between the pretraining values and the percent changes in these four depots are not significantly different from zero. Because age-related abnormalities in sympathetically mediated lipolysis might play a role in the central distribution of fat noted in the elderly, arterialized plasma epinephrine and norepinephrine concentrations were measured before and at the end of the training period. At baseline, epinephrine concentrations were not different between the two groups (58.7 ? 16.2 v 67.8 + 16.6 pg/mL, P = NS), while norepinephrine was higher in the older men (190.9 2 48.3 v 273.3 ? 90.9, P < .Ol). Following intensive endurance training there were no significant changes in epinephrine or norepinephrine concentrations in either group.
DISCUSSION
The effects of endurance training on body fat distribution (as measured by skinfolds and circumferences) have been previously reported in young subjects. Despres et al”” initially reported that a 20-week endurance training pro-
Body Fat Depot Areas by CT
(n = 13)
Older
(n = 15)
Within-Group Ml??BlNe
Pre
Post
Differences
Within-Group Pre
Post
Differences
54.8 ? 33.6
5
144.5 + 49.4
109.0 + 44.9
II
Abdominal SC (cm’) Chest SC (cm’)
218.7 2 119.5 92.9 ? 56.1 (n = 12)
197.9 + 103.9 90.1 -f 56.7
§
172.8? 41.8 115.3 ? 32.3
137.9 2 41.7 91.3 zk 34.5
1~
Buttock SC (cm’) Thigh SC (cm2)
182.3 ? 78.8 (n = 12) 154.6 + 75.8t
164.3 + 68.6 123.4 t 54.5
157.6 + 39.3 (n = 14) 80.5 ? 29.7
135.4 2 35.7 80.3 + 35.5
II
Thigh muscle (cm’)
345.8 + 56.7*
345.3 * 39.0
274.5 2 30.0
299.7 * 37.3
n
IA (cm’)
66.3 -t 37.1*
11
*P 5 .05, tP 5 .Ol, SP 5 ,001; baseline comparisons between groups. §P I .05, llP 5 .Ol, IIP 5 ,001; pre/post comparisons within each group. #P 5 .05, ##P
2 .Ol; comparison of pre/post differences between groups,
Between-Group Differences
##
#
EXERCISE AND FAT DISTRIBUTION
3 G 2
z .l-5
549
IN MEN
0
-20
-40
0
Older
-60 14
Abd-SO
Chest-SQ
Butt-Xl
Thigh-W
Thagh-Mus
Fig 1. Absolute changes in tissue areas measured by CT, following a B-month endurance training program in young (n = 13) and older (n = 15) men. Data are expressed as the mean k SEM. SD = SC. lP 5 .05. l*P 5 .Ol, l**p I ,001 pre/post comparisons within each group. ‘P < .05, ‘*P < .Ol comparison of pre/post differences between groups. Comparisons are adjusted for differences between groups in initial values.
gram in young non-obese men produced a 2.6-kg fat loss. This was associated with a 22% decrease in trunk skinfolds and only a 12.5% decrease in extremity skinfolds. This study noted that the changes in skinfolds were related to the initial size of the skinfold. In a 15week, high-intensity exercise training study, in non-obese young men,” these investigators again found a significantly greater reduction in truncal compared with extremity fat (27% v 15%) and thus a decrease in the ratio of extremity to truncal fat. Following a 1,000 kcalid exercise-induced energy deficit in moderately obese men, they noted a 6.8-kg loss of fat mass, a preferential loss of truncal versus extremity fat, and a greater reduction in total fat mass than SC fat, suggesting a loss of deep fat.‘j Another group has recently reported a signifi-
r
#
120 m
1
n q
IA
Young Older
Abd-SO
CL
Chest-SO Butt-SO Thigh-SD Thigh-Mus
Fig 2. Percent changes in tissue areas, measured by CT, following a B-month endurance training program in young (n = 13) and older (n = 15) men. Data are expressed as means t SEM. SO = SC. ‘P < 65, ‘“P < .Ol comparison between groups are adjusted for differences between groups in initial values.
cant decrease in the WHR in premenopausal women following a 6-month endurance training program.*’ Until now, no published study has evaluated the effects of endurance exercise on specific fat depots as measured by CT. In the present study, we noted a small but significant loss of body weight and total fat mass in the older men following an intensive 6-month endurance exercise program. In addition, the older men had a small but significant decrease in central (chest and abdominal) circumferences, while the more peripheral (arm, hip, and thigh) circumferences did not change. Contrary to the relatively small changes in body weight, total fat, and body circumferences, impressive (>20%) decrements were found in all three central fat depots with CT (SC chest and abdomen, and IA). Thus, there is a preferential loss of adiposity from central depots following intensive endurance training in older men. The pattern of exercise associated change in fat distribution in young men appears to be different than that observed in the older men, with a smaller loss of overall body fat and central fat, and a greater decrement in peripheral (thigh) fat. Some of the differences noted between the groups may be explained by the less central pattern of adiposity at baseline in the young. This is consistent with the previous finding that women, who have a more peripheral distribution of adiposity, appear resistant to exercise-induced changes in adiposity.” It is unlikely that the differences in the effects of training between the groups can be explained by differences in the training itself, since similar absolute and relative changes in Vo?max occurred in the two groups. Furthermore, the changes in caloric requirements for weight stabilization (Table 1) were similar in the two groups. The reasons for the preferential deposition of central fat with aging is unclear. Unlike premenopausal women who appear to have a strong relationship between central adiposity and testosterone concentrations,” androgenicity is unlikely to be a major determinant of the baseline central adiposity in older men, since androgen levels decrease with age.24 Catecholamines are important regulators of fat cell size and might play a role in the age-related predisposition toward central adiposity. In man, lipolysis is controlled by both p- (lipolytic) and (Y?-(antilipolytic) adrenergic receptor activity.‘5 The elevated catecholamine concentrations noted with aging in this study and previously” appear to cause homologous desensitization of the p-,” but not at,-adrenergi? receptor responsiveness. This could lead to a decrease in lipolytic activity or even an antilipolytic effect. Since normally IA fat appears to be very lipolytic,‘Y this depot might be most greatly affected by a relative decrease in p- to a,-receptor responsiveness, and a decrease in lipolysis. Another potential explanation for the central adiposity in older men is that adipose tissue lipoprotein lipase-induced triglyceride storage in central depots increases with aging. At present, there is no available data concerning this possibility. The reasons for the observed preferential loss of fat from central depots following endurance training are also unclear. It might be secondary to reversal of either an abnormal lipolytic or triglyceride storage pattern in the
SCHWARTZ ET AL
550
elderly. It is unlikely that exercise-associated changes in plasma catecholamine concentrations could explain the observed changes in fat depots, because neither epinephrine nor norepinephrine levels changed in either group with training. It is possible that a decrease in androgenicity with training may play a role in the observed decrement in central adiposity. A reduced testosterone level has been described in well-trained endurance athletes.‘” We noted relatively large changes in the central fat depots on CT (especially in the elderly), despite very small changes in overall adiposity. Although some of this may be accounted for by measurement error, the accuracy of our measures makes it unlikely that this can totally explain the difference. We hypothesize a redistribution of adiposity to areas not specifically measured on our limited number of scans. It is clear from the data of Sjostrom and Kvist” that changes in adiposity can be quite different in adjacent areas such as in the upper and lower abdomen. Whatever the cause of the decrement in central adiposity following endurance training, the magnitude of loss appears to be related to a more central distribution of adiposity at baseline. Since subjects with a more central
distribution of fat are most likely to suffer from obesityassociated metabolic abnormalities, it is possible that endurance exercise training may produce the greatest potential benefit in this at risk group, despite relatively small amounts of total weight loss. Although dietary weight loss studies in older individuals have not been published, most studies in young and middle-aged women show little or no change in central fat distribution (WHR) with dieting.“,” Therefore, exercise may be the preferred treatment for inactive patients with central obesity. Further support for this differential treatment concept must await more studies evaluating the comparative effects of weight loss diets and endurance exercise on fat distribution in different subject populations and the relationship between the observed changes in fat distribution and metabolic abnormalities.
ACKNOWLEDGMENT The authors would like to recognize the expert technical assistance of Leo Jaeger and Jan Wilson, as well as the skilled secretarial assistance of David Redick and Maria Limtiaco, in the preparation of the manuscript.
REFERENCES 1. Brozek J: Changes of body composition in man during maturity and their nutritional implications. Fed Proc 11:7X4-793, 1952 2. Borkan GA, Hults DE, Gerzof SG, et al: Age changes in body composition revealed by computed tomography. J Gerontol38:673677,1983 3. Garn SM, Sullivan TV, Hawthorne V: Evidence against functional differences between central and peripheral fat. Am J Clin Nutr 47:836-839, 1988
(eds): Fat Distribution During Growth and Later Health Outcome. New York, NY, Liss, 1988, pp 313-332 13. Despres J-P, Tremblay A, Nadeau A, et al: Physical training and changes in regional adipose tissue distribution. Acta Med Stand Suppl723:205-212,1988 14. Bruce RA, Kusumi F, Hosmer D: Maximal oxygen intake and normographic assessment of functional aerobic impairment in cardiovascular disease. Am Heart J 85:546-562, 1073
4. Kissebah AH, Vydelingum N, Murray R, et al: Relation of body fat distribution to metabolic complications of obesity. J Clin Endocrinol Metab 54:254-259, 1982
15. Goldman RF, Buskirk EK: A method for underwater weighing and determination of body density, in Brozek J, Henschel A (eds): Techniques for Measuring Body Composition. Washington, DC, National Academy of Sciences, 1961. pp 78-79
5. Krotkiewski M, Bjorntorp P, Smith U: Impact of obesity on metabolism in men and women. Importance of regional adipose tissue distribution. J Clin Invest 72:1150-1162, 1983
16. Shuman WP, Newell-Morris LL. Leoneth DL. et al: Abnormal body fat distribution detected by computed tomography in diabetic men. Invest Radio1 21:483-487, 1986
6. Peiris AN, Sothmann MS. Hoffmann RG, et al: Adiposity, fat distribution, and cardiovascular risk. Ann Intern Med 110:867-872. 1989
17. Sjostrom L. Kvist H. Regional body fat measurements with CT scan and evaluation of anthropometric predictions. Acta Med Stand SuppI723:169-177.1988
7. Gillum RF: The association of body fat distribution with hypertension, hypertensive heart disease, coronary disease, diabetes and cardiovascular risk factors in men and women aged 18-79 years. J Chronic Dis 40:421-428. 1987
18. Evans MI, Halter JB. Porte D Jr: Comparison of double- and single-isotope enzymatic methods for measuring catecholamines in human plasma. Clin Chem 24:567-570, 1978
8. Welin L, Svardsudd K, Wilhelmsen L. et al: Analysis of risk factors for stroke in a cohort of men born in 1913. N Engl J Med 317:521-526, 1987 9. Larsson B, Svardsudd K, Welin L, et al: Abdominal adipose tissue distribution, obesity and risk of cardiovascular disease and death: 13 year follow-up of participants in a study of men born in 1913. Br Med J 288:1401-1404,1984 10. Seidell JC, Bjorntorp P, Sjostrom L, et al: Regional distribution of muscle and fat mass in men-New insight into the risk of abdominal obesity using computed tomography. Int J Obesity 13:289-303, 1989 11. Schwartz RS, Shuman WP, Bradbury VL. et al: Body fat distribution in healthy young and older men. J Gerontol 45:M181M185,1990 12. Himes JH: Alteration in distribution of adipose tissue in response to nutritional intervention, in Bouchard C, Johnston FE
19. Shimokata H. Tobin JD, Muller DC, et al: Studies in the distribution of body fat: I. Effects of age, sex and obesity. J Gerontol44:M66-M73,1989 20. Despres J-P, Bouchard C. Tremblay A. et al: Effects of aerobic training on fat distribution in male subjects. Med Sci Sports Exert 17:113-118, 1985 21. Tremblay fat distribution Fat Distribution York, NY, Liss,
A, Despres J-P, Bouchard C: Alteration in body with exercise, in Bouchard C, Johnston FE (eds): During Growth and Later Health Outcomes. New 1988, pp 297-312
22. Krotkiewski M, Bjorntorp P: Muscle tissue in obesity with different distribution of adipose tissue. Int J Obesity 10:311-341. 1986 23. Peiris AN, Mueller RA, Struve MF, et al: Relationship of androgenic activity to splanchnic insulin metabolism and peripheral glucose utilization in premenopausal women. J Clin Endocrinol Metab 64:162-169. 1987
EXERCISE AND FAT DISTRIBUTION
IN MEN
24. Tenover JS, Matsumoto AM, Plymate SR, et al: The effects of aging in normal men on bioavailable testosterone and luteinizing hormone secretion: Response to clomiphene citrate. J Clin Endocrinol Metab 65:1118-1126, 1987 25. Mauriege P, Galitzky J, Berlan M, et al: Heterogeneous distribution of beta- and alphas-adrenoreceptor binding sites in human fat cells from various fat depots: Functional consequences. Eur J Clin Invest 17:156-165, 1987 26. Veith RC, Featherstone JA, Linares OA, et al: Age differences in plasma norepinephrine kinetics in humans. J Gerontol 41:319-324, 1986 27. Stiles GL, Caron MG, Letkowitz RJ: B-Adrenergic receptors. Biochemical mechanisms of physiologic regulation. Physiol Rev 64:661-743,1984
551
28. Pfeifer MA, Ward K, Malpass T. et al: Variations in circulating catecholamines fail to alter human platelet alpha?adrenergic receptor number or affinity for yohimbine or [‘HI dihydroergocryptine. J Clin Invest 74:1063-1068,1984 29. Rebuffe-Strive M, Anderson B, Olbve L, et al: Metabolism of adipose tissue in intraabdominal depots of nonobese men and women. Metabolism 38:453-458, 1989 30. Hackney AC: Endurance training and testosterone levels. Sports Med 8:117-127, 1989 31. Garrow JS: Is body fat distribution changed by dieting? Acta Med Stand Suppl723:199-203,1988 32. Wadden TA, Stunkard AJ, Johnston FE, et al: Body fat distribution in adult obese women. II. Changes in fat distribution accompanying weight reduction. Am J Clin Nutr 47~229-234, 1986
