Differences in the mean fat cell diameter of males between 1 and 48 months of age.код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOI.OGY 65.341-945 (19811 Differences in the Mean Fat Cell Diameter of Males Between 1 and 48 Months of Age FRANCIS E. JOHNSTON, MARIAN WESTON, SHORTIE MCKINNEY, JAMES COLEMAN, GILBERT0 PEREIRA, AND JEAN ROUNDS Department of Anthropology, University of Pennsylvania, Philadelphia, Pennsylvania 19104 (EE.J., M. W, Department of Nutrition, Drexel University, Philadelphia Pennsylvania 19104 (S.M.), Neonatology Department, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania 19104 (J.C., G.P., J.R.) KEY WORDS Fat cells, Growth, Body composition, Infancy ABSTRACT The mean fat cell diameter was determined from measurements of abdominal adipose cells, obtained during inguinal hernia repair, of 126 white and 95 black males ranging in age from 1through 48 months of age. The mean diameters of black and white subjects did not differ significantly, suggesting that differences in fatness among adults of these two ethnic groups have their origin beyond the age range of this study. The mean fat cell diameter increased through the 6-8-month age group, decreased until the end of the first year, and then levelled off through 48 months of age. Comparison of this curve with those for the triceps, subscapular, abdominal, and suprailiac skinfolds of the same subjects showed generally parallel courses except for the triceps, which continued to increase in size after the means of fat cell diameters and the other skinfolds had levelled off. Our data indicate that changes in body fatness on the trunk at least in the first 4 years of life may be accounted for by changes in fat cell size. The development, in the 1970s, of techniques for the characterization and quantification of adipose tissue cellularity stimulated a wide-ranging body of research into methodological issues, as well as applications of the findings to clinical and epidemiological areas of research. In particular, the discovery of increased adipose cell size and number among the obese (Bray, 1970; Hirsch and Knittle, 1970; Brook and Lloyd, 1973; Sjostrom and Bjorntorp, 1974)resulted in a surge of interest in the subject both in professional and popular publications. During this period, research and speculation led to the formulation of the “adipocyte number hypothesis,” which postulates that the number of adipocytes is fixed early in life due to environmental andor hereditary influences (Roche, 1981). The result of this early fixation is to “predestine” individuals towards a lifetime of obesity or leanness. While the implications of this hypothesis are clearly important and while they have enjoyed immense popularity, it is clear now that the adipocyte number hypothesis must be viewed with considerably more caution than was originally the case. 0 1984 ALAN R. LISS. INC. Sjostrom et al. (1972) have described differences among male and female adults of different ages, as well as among anatomical sites within the same individual. By transplanting cells of mice from site to site, Meade and Ashwell (1979) demonstrated that site differences resulted from location on the body rather than in the sample of fat tissue itself. In his review of age changes during development, Brook (1978) has suggested that, at present, methodological difficulties prevent any reliable estimates of adipocyte number in a n individual. Summarizing the available data, Roche (1981) concluded that, based upon present knowledge, the adipocyte number hypothesis is untenable. In large part, this is due t o methodological problems involved in the estimation of body fatness a t the whole body level, and to the extrapolation of measurements of adipocytes from a single region (or even from several regions) to the entire adipose organ. Additional data are required before the important questions raised by research conducted to date can be answered. Received January 9,1984;accepted July 31, 1984 342 F.E. JOHNSTON ET AL In this paper we present data on age differences in the mean size of abdominal adipose cells from a sample of 221 black and white male infants and young children. The information should contribute to our knowledge of the growth of adipocytes during this potentially crucial period of human development. MATERIALS AND METHODS The data analyzed in this paper are taken from a comprehensive mixed-longitudinal study of the growth and development of infants and children from the Greater Philadelphia area. Subjects were drawn from those males 48 months of age and younger who underwent surgery for repair of an inguinal hernia a t the Children’s Hospital of Philadelphia from August 1, 1979 through February 28, 1981. In view of the low incidence of hernia among females, we have restricted our sample to males. At the time of surgery, a sample of subcutaneous tissue was collected at the point of incision. Following the method of Sjostrom et al. (19711, the tissue was frozen after a brief period of fixation in formaldehyde, sectioned, and mounted for photomicrography. The photomicrographs were enlarged to a standard size and a n average of 56 cells were traced for each individual. The circumferences of the cells were then measured with a planimeter, converted by computer to diameters, and expressed for analysis in microns. Systematic scanning procedures were followed to prevent bias in selecting cells for tracing. Virtually all of the 12,000+ cells were traced by one research assistant and all circumferences were measured by another. Inter- and intraobserver reliability for both tracing and measurement were analyzed and found to be quite high. Usable adipocyte measurements were obtained from 221 subjects, 126 white and 95 black, ranging in age from 1through 48 months of age. A battery of additional information and measurements was collected a t the time of surgery. This included anthropometry (growth and body composition), sociodemographic data, and information on dietary intake and preferences. Those subjects under 12 months of age were enrolled in the longitudinal component of the study, to be visited in their homes at 6 (if appropriate), 12, 18, and 36 months of age. At these follow-up visits anthropometric data were collected, as well as similar measurements of siblings and the subject’s mother. In addition we have collected dietary recalls and histories, food preferences of the mother, maternal ratings of child temperament at 6,18, and 36 months (Carey, 1970; Carey and McDevitt, 1978), and, in a subsample at 18 months, records of heart beats and sleeplwake patterns over a %-hour period. In view of the increased incidence of shorter gestational ages associated with inguinal hernia (Czeizel, 1980),all ages have been adjusted for gestational age as verified from the hospital records at birth. For this paper, subjects have been grouped into eight groups, as shown in Table 1. These particular age groupings were chosen for two reasons. First, we optimized our age distribution of subjects, giving us sufficient numbers per cell. Second, the groups correspond in general to the rate of postnatal growth, providing larger numbers a t ages when growth velocities are highest. For this paper we are reporting only on the analysis of fat cell diameters and anthropometry at the first, i.e., the surgery, visit. Subsequent publications will utilize the longitudinal component as well as the other data collected as part of this study. RESULTS We examined the shapes of the distributions of fat cell means in each of our subjects. This was done visually as well as by the calculation of the 3rd and 4th moments about the mean. As might be expected, a range of variability was found in skewness and kurtosis. However, the majority of distributions did not depart significantly from the normal in skewness and, as a result, we have chosen to utilize the mean and standard deviation for description and parametric statistics for analysis. Table 1presents the number of subjects, by race, in each age group, and, for the fat cell diameters, their means, standard deviations, and coefficients of variation. An analysis of black and white subjects, adjusted for age, revealed no significant differences in the means (F’ = 0.79, p = 0.37) and, as a result, we have pooled the two racial groups to increase sample size and resulting power. There is considerable variation within any single age group, with coefficients of variation ranging from 11.4 to 22.2%. The smallest standard deviation, 7.86, occurs among FAT CELL SIZE AND GROWTH 343 TABLE 1. Mean fat cell diameter in Philadelphia male infants and children Age group (mo) Black Sample size White B-2 11 3-5 6-8 9-11 12-17 18-23 24-35 36-48 34 12 9 10 7 5 20 22 8 8 18 13 18 19 Total 95 126 I the youngest subjects, this value being significantly less than the next smallest SD, 11.95 a t 3-5 months (F = 2.31, p = .004). The means show a n increase in fat cell diameter from the youngest group, birth to 2 months, through the 9-11-month group. The difference between the B-2 and 3-5 month means is significant (t = 2.65, df = 85, p < .01) as is that between the 6-8- and 9-11month means (t = 2.18, df = 43, p <.05). Following the 9-11-month group, the means decrease and level off, showing no significant differences among themselves. The changes with age may be verified by calculating, within age groups, the correlation of mean fat cell diameter and age. To do so, we obtain the following in the first 2 years of life: B-5 months, +0.17; 6-11 months, 0.00; 12-17 months, -0.81, and 18-23 months, -0.32. Figure 1presents graphically the mean fat cell diameter by age group. In addition we have plotted the means for each of the four skinfolds which were measured within 24 hours of surgery. The shapes of the curves for the subscapular, abdominal, and suprailiac skinfolds generally parallel that for the mean fat cell diameter, with increasing values in the first year of life, followed by a decrease and a levelling-off by 18 months. From 24 to 48 months, the means of these four skinfolds show, overall, a slight decrease, something which is not seen in the fat cell measurements. The triceps skinfold means do not conform to the above observations. While there is the increase in the first year and a subsequent decrease, there is a general increase in the means from the 12-17-month group through the age range of the study. Total Mean 31 56 20 17 28 20 25 24 68.5 74.8 73.8 77.0 68.0 70.2 70.1 69.6 22 1 Fat Cell Diameter (pm) S.D. cv (%) 7.86 11.95 16.39 13.09 13.90 14.03 11.86 13.80 11.4 16.0 22.2 17.0 20.5 20.0 16.9 19.8 10 u mm. 2. 1. mg. One particular observation distinguishes the fat cell means from each of the skinfold thicknesses. The mean fat cell diameter is greatest in the 9-11-month group. This value is, as noted above, significantly greater than the subsequent 12-17-month mean, but not different from the preceding 6-8-month value (t = 0.66, df = 35, p > .40). In contrast to this, the largest mean for each of the skinfolds occurred in the 6-8-month group. DISCUSSION As noted earlier, in recent years, doubt has been cast upon the validity of current measures of adipose cell number as indices of hypercellularity in individuals. For example, in a study of the adipose cells of 80 obese and 27 nonobese individuals, Jung et al. (1978) found a slight increase in the estimated num- 344 F.E. JOHNSTON El' AI,. ber of fat cells of the obese; however, no relationship existed between cell number and the age at onset of obesity. Based upon differences in the mean size of subcutaneous and omental cells, they concluded that the techniques for counting which are used a t present result in a n underestimate of cell number. This suggests that the obese may be able to accommodate increased amounts of fat without adding new cells. Kirtland and Gurr (19791, after reviewing the published evidence, reached a similar conclusion. While recognizing that the obese do have more fat cells than the nonobese, they suggest that this is a consequence, and not a cause, of the obesity. Based upon the literature, we have chosen to focus our study upon cell size, a variable which can be measured with much greater reliability as well as validity. Even so, the measurement of adipose cell size presents methodological difficulties. Of particular concern is the degree to which a sample of adipose tissue from one site is representative of other sites throughout the body. Issues related to the protection of human subjects, especially infants and children, precluded our taking concurrent samples from other sites. However, some data on variation among sites are available in the literature. In a study of 24 adults in their 20s, Sjostrom et al. (1972) found, among 11males, correlations between weights of fat cells from different anatomical regions ranging from 0.43 to 0.56. The exception to this range in their study was the correlation of epigastric and hypogastric fat, at 0.81. Except for the last, these coefficients were not significant because of the very small sample size. If we assume that the values would remain the same, they would achieve significance at the .05 level with a sample size of 21. Even so, the correlations are only moderate and leave considerable variance unexplained. The age-associated patterns of our data are generally consistent with the findings of Boulton et al. (1978)in a study of a sample of 43 infants and fetuses through 28 months of age. Using the modal cell diameter for analysis, they found a n increase until 6-8 months of age, with no further changes. The mean diameters of our subjects increased in a similar fashion, but showed somewhat smaller means from 12 through 48 months. The age-related patterns of the fat cell means agreed quite well with the patterns for the skinfolds which were measured, the least so with the triceps. This general agreement suggests that the increase in body fatness between early infancy and 48 months of age, as indicated by skinfold thicknesses, can be accounted for by an increase in adipose cell size. However, two exceptions to this observation exist. The first is the increase in the means of the triceps fold after the 12-17month group, which was not noted for fat cell diameter. One possible explanation is that this reflects differences in limb fat, estimated by the triceps fold, and trunk fat, from which the adipose sample was taken. The second exception to the above observation is in the disagreement between the skinfolds and the fat cell means in the age at their peak value. For all skinfolds, this occurred in the 6-8-month group while the 911-monthmean was the highest fat cell mean. Sampling or measurement error are always possibilities. Some studies have found the peak values for skinfold thinkness during infancy to occur somewhat earlier, while still others have failed to note any peak at all (Johnston, 1978). Because of the lack of comparative information, we cannot speculate on the significance in our data, if any, of the difference between the ages at which the peaks values for mean fat cell diameters and mean skinfold thicknesses occur. Based upon a longitudinal study of adipocyte development, Hager et al. (1977) suggested that increased body fatness can be explained by an increase in fat cell size from 1-12 months, but by fat cell number from 12-18 months. Our cross-sectional data, if correct, suggest that the increase in trunk fatness during the first 4 years of life may be accounted for by a n increase in fat cell size, with the possible exception of the last part of the first year of life. Our data indicate no difference between fat cell diameter of black and white subjects. Neither were there differences in skinfold thicknesses. While our sample sizes are small, this is in agreement with the study of Johnston and Beller (19761, who found no differences in the triceps and subscapular skinfolds of black, white, or Puerto Rican newborns. Since it is well-established that the levels of fatness of black American children, youth, and adults are less than those of their white age and sex peers (Johnston et al., 1972, 19741, it seems clear that factors responsible for these differences are not expressed until after early childhood. With respect t o fat cell size, the findings of our study FAT CEI,I, SIZE A N D GROWI’H regarding racial differences are consistent with anthropometric data. Finally, we must emphasize that our findings may be generalized to normal-weight infants and children, and not to the obese. None of our subjects could be classed unambiguously as obese. Consequently, any conclusions or suggestions growing out of this analysis may not be applied to obese infants or children, whose fat cells may develop within a different set of environmental andi or hereditary determinants. ACKNOWLEDGMENTS This research was supported in part by USPHS Grant HD-12480. The assistance of Adrienne Rogers, Judy Ricci, Marc Hurowitz, and Charles White is gratefully recognized. This paper was presented a s part of a symposium, “Adiposity Through the Life Cycle,” organized by L.S. Adair and F.E. Johnston, at the annual meeting of the American Association of Physical Anthropologists, Indianapolis, April 8, 1983. LITERATURE CITED Boulton, TJC, Dunlop, M, and Court, J M (1978) The growth a n d development of fat cells in infancy. Pediatr Res 12:908-911. Bray, GA (1970) Measurement of subcutaneous fat cells from obese patients. Ann. Intern. Med. 73.365-569. Brook, CGD (1978) Cellular growth:adipose tissue. In F Falkner and JM Tanner (eds):Human Growth, 2. Postnatal Growth, New York: Plenum, pp.21-34. Brook, CGD, a n d Lloyd, JK(1973) Adipose cell size and glucose tolerance in obese children and effects of diet. Arch. Dis. Child. 48:301-304. Carey, WB (1970) A simplified method for measuring infant temperature. J Pediatr 77:188-194. Carey, WB, a n d McDevitt, SC (1978) Revision of the 345 infant temperament questionnaire. Pediatrics 61.735739. Czeizel, A (1980) Epidemiologic characteristics of congenital hernia. Helv. Pediatr. 35:57-60. Hager, A, Sjostrom, L, Arvidsson, P, Bjorntorp, P, and Smith, U (1977) Body fat and tissue cellularity in infants: a longitudinal study. Metabolism 26r607-614. Hirsch, J, and Knittle, JL. (1970) Cellularity of obese and nonobese human adipose tissue. Fed. Proc. 2.915161521. Johnston, F E (1978) Somatic growth of the infant a nd preschool child. In F Falkner and JM Tanner (eds): H uma n Growth, 2. Postnatal Growth. New York: Plenum. pp. 91-116. Johnston, FE, a nd Beller, A (1976) Anthropometric evaluation of the body composition of black, white, and Puerto Rican newborns. Am. J. Clin. Nutr. 29t61-65. Johnston, F E , Hamill, PVV, and Lemeshow, S (1972) Skinfold Thickness of Children 6-11 Years. United States. DHEW Pub (HSM) 73-1602. Washington DC: U S Govt Printing Office. Johnston, FE, Hamill, PVV, and Lemeshow, S (1974) Skinfold Thickness of Youths 12-17 Years, United States. DHEW Pub (HRA) 74-1614. Washinbeon DC: U S Govt Printing Office. Junlr. WT. Gurr. MI. Robinson. MP. a nd James. WPT (19’iS) Does adipocyte hyperceliularity in obesity exist? Br. Med. J. 2:319-321. Kirtland, J, and Gurr, MI (1979) Adipose tissue cellularity, a review. 2. The relationship between cellularity and obesity. Int. J. Obesity 3:15-55. Meade, CJ, a nd Ashwell, M (1979)Site differences in fat cell size-a transplantation study. Proc. Nutr. Soc. 38:62A. Roche, A F (1981) The adipocyte number hypothesis. Child Dev. 5 2 3 - 4 3 . Sjostrom, I,, and Bjorntorp, P (1974) Body composition and adipose tissue cellularity in human obesity. Acta Med. Scand. 2%5:20 1-2 11. Sjostrom, L, Bjorntorp, P, a n d Vrana, J (1971) Microscopic fa t cell size measurements of frozencut adipose tissue in comparison with automatic determination of osmium-fixed fat cells. J. Lipid Res. 12:521-530. Sjostrom, L, Smith, U, Krotkiewksi, and Bjiirntorp, P (1972) Cellularity in different regions of adipose tissue in young men a nd women. Metabolism 21:1143-1153.