Monthly changes in the gain and loss of growth in weight of children living in Guatemala.код для вставкиСкачать
Monthly Changes in the Gain and Loss of Growth in Weight of Children Living in Guatemala BARRY BOGIN Department of Anthropology, Wayne State Uniuerslty, Detroit. MLchigan 48202 K E Y WORDS Seasonal growth . Weight growth . Guatemala ABSTRACT Monthly increments of weight growth for a sample of 246 Guatemala City private school children are analyzed for the presence of a seasonal pattern in rates of growth. Neither a seasonal pattern nor any other periodic rhythm is found. It is observed that a significantly greater number of children aged 5.0 to 6.9 years experience their minimum annual growth rate during the dry season, with up to 60% of them losing or not gaining weight in any one month. Patterns of diet, exercise and disease cannot explain this trend. A possible association between minimum weight growth and maximum growth in height is discussed. In a n earlier publication (Bogin, '781, a seasonal pattern in the rate of growth in height was described for a sample of children and youths living in Guatemala. It was found that preadolescent boys and girls and male youths aged 14 years and older grew a t a faster rate during the dry season, and a t a slower rate during the rainy season. These findings parallel those from previous investigations, conducted throughout the world, which find that the maximum annual rate of height growth usually coincides with the season of maximum sunlight availability (but see Marshall, '75, for an exception). The literature on seasonal variation in growth also reports t h a t weight gain and loss follow a seasonal pattern. Of the 29 studies reviewed by Bogin ('771, 22 find that fall or winter are the seasons of maximum gain. Four studies report summer to be the season of greatest growth and three find no seasonal variation. No study finds spring to be the season of maximum growth. Indeed, most report spring and summer to be the seasons of minimal growth. Several find actual weight loss to occur during the spring. Most of these investigations are based on samples of healthy, well nourished children living a t temperature latitudes north or south of the equator. In the tropical regions the traditional seasons-summer, fall, winter, spring-have little climatic meaning or biological relevance. AM. J. PHYS. ANTHROP. (1979) 51: 287-292. Dry and rainy seasons are more meaningful classifications. The three studies of seasonal differences in weight growth of children in the tropics report conflicting results. Vincent and Dierickx ('60) find the dry season to be the period of maximum gain for African and European children living in Leopoldville. This is also the season of maximum gain in height. Eveleth ('67-'68) also reports maximum gain in weight during the dry season for United States children living in Rio de Janeiro, Brazil. But, most of these children spent part of the Brazilian dry season vacationing in the U.S.A., so the seasonal effect on weight is difficult to interpret. Conversely, Billiwicz ('67) finds maximum weight gain of a group of rural Gambian children to occur in the rainy season. In this case diet may be the factor responsible for the seasonal pattern. Both minimal weight gain and food shortages occur during the dry season. Since the seasonal weight growth studies conducted in the tropics are not directly comparable with those from the temperate latitudes, a new study of tropical children was carried out and the results are reported in this paper. MATERIALS A N D METHODS The sample and research design used for this study are exactly the same as those used 287 288 BARRY BOGIN TABLE 1 Descriptive statistics for subsample size and age distributions Subsample code Subsample size Age range Mean age Standard deviation Boys Girls Boys M1 43 F1 42 M2 41 6.9 0.31 Boys Girls F2 39 11.0-13.9 12.3 12.0 0.46 0.44 5.0-6.9 6.0 0.36 Girls M3 46 F3 35 14.0-17.9 16.0 16.0 0.71 0.38 TABLE 2 Mean total gain in hezght and weight for each subsample, September, 1974-October, 1975 Height (cm) Subsample Meangain Standard deviation Minimum Maximum M1 6.47 F1 6.38 0.76 5.2 8.1 0.77 4.4 7.8 M2 Weight ckg) F1 F2 6.56 M3 1.83 F3 0.31 M1 3.46 3.20 2.16 1.91 3.9 3.0 12.1 10.4 1.32 0.0 6.4 0.48 0.0 1.7 1.47 1.4 7.9 1.05 1.3 6.3 7.17 for the study of seasonal variation in height growth described in Bogin (’78).Briefly, these are children of European, North American, and Guatemalan families living permanently in Guatemala. All are from the upper socioeconomic levels of Guatemala City and all are free of significant nutritional or disease stress. Six subsamples of children divided by age and sex were measured. Subsample structure is described in table 1. Three age groups were purposefully selected for study. Few researchers have considered age as a variable in either the design or analysis of seasonal growth studies. Marshall (‘15) measured preadolescent children so that the rapid acceleration and deceleration of growth during the adolescent “spurt” in growth would not mask the seasonal effect. Conversely, Thompson (’42) presents data in which the seasonal differences in rate of height growth are most pronounced during the adolescent years. No data for weight growth is given. Kemsley (’53) reports that seasonal variation in body weight, independent of diet and exercise, persists into adult life. Due t o the conflict of opinion and the lack of sufficient data in the literature it was decided to measure rates of growth at several different ages. The data consist of 13 monthly increments of weight (from September, 1974 t o October, 1975). Weight was taken immediately following the height measurement. All measurements were taken by the author using a Health-0-Meter balance capable of recording t o the nearest half pound. The pound measure- M2 6.21 FZ 6.63 M3 3.45 F3 1.25 2.28 3.16 1.4 -4.0 10.7 15.3 2.37 -2.1 9.9 2.74 -6.4 7.5 ments were converted t o kilograms for analysis. The subjects were weighed barefoot, dressed in minimal indoor clothing. An assistant recorded and checked all measurements for reliability. Any individual missing more than one measurement was excluded from the study. RESULTS Table 2 presents descriptive statistics for the mean total gain in height and weight of the children in each of the subsamples during the 14 months of study. Growth for the M1 and F1 subsamples is relatively homogeneous and representative of the annual growth rate expected for clinically healthy children of this age (Tanner and Whitehouse, ’76). Growth of the youths in the M2 and F2 subsamples is heterogeneous. The range of variability is due to the inclusion of both preadolescent and adolescent individuals in the age group. Similar heterogeneity in growth of the M3 and F3 subsamples is due to the inclusion of adolescent and post-adolescent youths and some individuals who have stopped growing in height. But, weight growth, both positive and negative, is still pronounced, and seasonally related fluctuations may still be discernible. The total gain in weight was divided between the mean amounts achieved during the dry and rainy seasons. Table 3 presents these data. During the time of study the dry season began in October and the rainy season began in April (Bogin, ’78). Weight growth during the dry season is greater than the rainy season in every subsample. A test for the significance MONTHLY CHANGES IN WEIGHT GROWTH TABLE 3 Mean gain in weight achieued during the dry semon and rainy season ~ Dry season ~~ Rainy season Subsample M1 F1 M2 F2 M3 F3 n jigain s.d. n .?gain 41 38 40 37 46 35 2.09 1.90 3.11 3.78 2.27 0.80 0.94 0.95 1.83 2.86 1.85 2.88 33 33 35 33 42 33 1.42 1.20 3.08 2.46 1.37 0.51 8.d. 1.27 1.56 1.87 1.53 1.61 1.93 ' The difference &tween dry and rainy season means is significant at the 0.05 level. of the difference for all subsamples combined is highly significant (F,.,,, = 121.38, p < 0.001), but post-hoc comparisons of the means indicate signficance only for the F2 and M 3 subsamples. Since the growth rates of individual children may differ markedly from the mean seasonal trend (Bransby, '45; Marshall, '711, individual sequences of monthly changes were examined. Table 4 presents the month in which individual children, expressed as a percentage of the total subsample size, attained their maximum or minimum rate of growth. As can be seen the timimg of the maximum or minimum rates is variable and no obvious pattern emerges. To look for a pattern the monthly growth rates were analyzed by season. The growth rates, grouped by dry and rainy season, are presented in table 5. The number and percentage of individual children reaching maximum or minimum growth rate 2 89 in each season is given for each subsample. The statistical test for the difference between seasonal percentages for all subsamples combined it-test for percentages, Snedecor, '56) is not significant for maximum weight growth but is highly significant for minimum weight growth. Comparisons of the seasonal percentages of minimum growth for each subsample indicates significant differences for the M1, F1, M2 and F3 subsamples. Some of these minimum weight gains were negative. Other children experienced zero gains. Table 6 presents the percentage of children losing or not gaining weight during each month. The youngest children (subsamples M1 and Fl) show the strongest relation of zero or negative gain with the dry season. The relation is weak in the M2 subsample and absent in the others. It is also apparent that zero or negative gain is not confined to either season for any subsample. Table 5 also shows that many children experience both maximum and minimum growth rates in the same season. This is especially apparent for the dry season percentages; approximately half the children experience maximum weight gains and three quarters their minimum gain. However, since minimum growth rates do cluster into significant groupings for some subsamples it is possible the maximum growth rates also cluster, but not according to season of the year. In an effort to detect other patterns for maximum weight growth a periodic regression and harmonic analysis (Bliss, '70) was performed. The value of this analysis is that the complete month to month data for each child are TABLE 4 Percentage of mmimum and minimum growth rates of weight observed each month Subsamples M1 Month October November December January February March April May June July August September October F1 F2 M2 Max Min Max Min Max Min 7.0 20.9 7.0 7.0 14.0 2.3 2.3 0 7.0 20.9 2.3 2.3 7.0 0 2.3 2.3 4.7 4.7 18.6 16.3 37.2 0 0 4.7 2.4 19.0 7.1 14.3 16.7 2.4 2.4 0 4.8 7.1 4.7 2.3 2.3 14.3 4.8 17.1 17.1 4.9 9.8 14.6 2.4 2.4 7.3 0 14.6 2.4 7.3 0 2.4 9.8 4.9 4.9 19.5 12.2 12.2 7.3 4.9 7.3 4.9 4.9 4.9 4.8 7.1 2.4 7.1 7.1 11.9 9.5 26.5 4.8 2.4 7.1 7.1 7.1 Max 5.1 10.3 10.3 7.7 10.3 5.1 2.6 12.8 7.7 7.7 5.1 5.1 10.3 F3 M3 Min Max Mln Max Min 5.1 15.4 2.6 2.6 17.9 5.1 10.3 5.1 7.7 2.6 7.7 5.1 12.8 6.5 13.0 6.5 2.2 6.5 4.3 4.3 6.5 19.6 2.2 23.9 2.2 2.2 4.3 10.9 4.3 6.5 13.0 15.2 6.5 8.7 6.5 6.5 2.2 8.7 6.5 0 8.6 14.3 8.6 20.0 2.9 8.6 2.9 2.9 5.7 11.4 8.6 5.7 0 20.0 2.9 5.7 8.6 11.4 11.4 11.4 5.7 8.6 0 5.7 8.6 290 BARRY BOGIN TABLE 5 sonal rhythm absent, but that no other periodic rhythm for weight growth was present. Number and percentage of individuals attaining maximum or minrmum velocities in monthly rates of weight growth during the dry and rainy seasons ’ Subsample Dry seaeon n , a F2 M3 F3 M1 F1 M2 F2 M3 F3 Minimum growth rates 6 18.0 37 90.2 30 78.9 12 38.4 12 34.3 29 72.5 16 48.5 23 62.2 30 65.3 16 38.1 10 30.3 25 71.4 312 p % Maximum growth rates 23 56.1 20 60.6 26 68.4 16 48.5 24 60.0 17 48.6 23 62.1 16 48.5 20 43.5 16 38.1 12 36.4 23 65.7 M1 F1 I Y Rainy season DISCUSSION 0.001 0.001 0.05 0.50 0.10 0.001 Overall test of significance: maximum growth rates t = 5.72, p = 0.001. = 1.66, p = 0.10;minimum growth rates t TABLE 6 Percentage of children losing or not gaining weight during each month of study Subsamples Month M1 October November December January February March April ,May June July August September October FI M2 F2 M3 F3 11 8 5 11 30 45 7 5 10 22 30 32 55 17 11 14 16 31 42 60 12 12 20 42 32 32 14 18 14 12 19 13 12 6 30 16 29 32 37 14 25 49 37 31 38 47 38 50 5 8 19 21 16 19 10 13 15 25 14 12 18 6 3 18 9 15 20 33 22 35 22 51 51 34 36 45 6 3 treated separately to determine if a periodic rhythm in growth rate exists. Each individual is then compared with the mean trend of growth for his or her subsample. Using this technique both seasonal and non-seasonal periodic rhythms of growth for individual children can be detected. Because of its length a full discussion of this procedure is not presented here (the reader is referred to Bogin, ’77). The results of the analysis show that the only significant pattern is that for the non-periodic scatter of weight gains and losses from month to month. This means that not only was a sea- The mean gain in weight for each subsample of children is greater during the dry season then during the rainy season. For this group pattern to be truly seasonal it is necessary that the growth rates of individual children conform to the mean trend and that each child‘s rate of change in growth rate be periodic, regularly increasing to a maximum and decreasing to a minimum (Luce, ’71; Bogin, ’77). When individual sequences of growth are analyzed for the distribution of maximum and minimum weight gains no seasonal rhythm is found. In fact, i t is observed that between 25 and 50% of the children in all subsamples experience both maximum and minimum weight gains during the dry season. Periodic regression and harmonic analysis suggest that the observed monthly weight changes are essentially random when compared with the expected periodic pattern. The observed weight fluctuations are probably not associated with obvious weight-influencing factors such as food availability, exercise or disease. It was previously shown that food availability and exercise patterns are virtually constant throughout the year for these children and that the influence of disease is negligible (Bogin, ’77, ’78). The apparent randomness in weight growth of the older children and youths (subsamples M2-F3) may be largely due to the effects of adolescence. No attempt was made to assess the developmental status of any child but i t is reasonable to assume that some of these individuals, aged 11.0 to 17.0 years, would be experiencing the adolescent growth “spurt.” Included in this group would be some adolescents accelerating in growth rate, some a t or near peak weight velocity, and others decelerating in growth velocity. For the American School population the mean age of peak weight velocity (PWV) is estimated to be 13.6 years for boys and 12.1 years for girls (Johnston et al., ’73). This is also an age when weight reduction via “dieting” is a common practice, and may have influenced the findings. The weight growth of the oldest females (subsample F3) is further complicated by the fact that about 75%of these young women did not grow in height during the study. Thus Thompson’s (’42) observation of a greater sea- MONTHLY CHANGES IN WEIGHT GROWTH sonal effect on growth rates during adolescence is not found in the present study for growth in weight. Nor was it observed for growth in height (Bogin, '78). Kemsley's ('53) report of a seasonal rhythm in weight growth after height growth has stopped was not repeated in the present study. The non-periodic changes in growth of the youngest children cannot be an effect of adolescence. It is also unlikely that children 5.0 to 6.9 years old were significantly restricting their food intake in order to reduce weight. The greater percentage of children at this age losing or not gaining weight must be otherwise explained. The annual school holiday coincides with the beginning of the dry season, but the largest number of zero or negative weight gains occur four months after the children return to school. Several sources from the literature on seasonal growth suggest that, for groups of children, rate of weight growth is a t or near minimum when rate of height growth is a t maximum (Nylin, '29; Orr and Clark, '30; Fitt, '41; Bransby, '45, Takahashi, '66). Growth in height of the M1 and F1 subsamples follows a periodic rhythm with about 75% of the individuals experiencing maximum gains during the dry season and minimum gains during the rainy season (Bogin, '78). Although the weight growth of these children is not seasonal or periodic, a significantly greater number of individuals experienced minimum gains during the dry season. In the last three months of this season between 30 and 60%of the children either lost or did not gain weight. An association between maximum growth in height and minimum growth in weight is suggested by these data for the youngest children. Detailed confirmation of this trend remains to be demonstrated. ACKNOWLEDGMENTS Research was supported in part by the American School of Guatemala and the Universidad del Valle de Guatemala, and conducted while the author was a fellow with the Latin American Teaching Fellowship 291 Program of Tufts University. The author thanks the editor and referees of this journal for helpful criticisms. LITERATURE CITED Billewicz, W. Z. 1967 A note on body weight measurements and seasonal variation. Human Biology, 39: 241-250. Bliss, C. I. 1970 Statistics in Biology. Vol. 2. McGraw-Hill, New York. Bogin, B. A. 1977 Periodic Rhythm in the Rates of Growth in Height and Weight of Children and its Relation to Season of t h e Year. Doctoral dissertation. University Microfilms, Ann Arbor. 1978 Seasonal pattern in the rate of growth in height of children living in Guatemala. Am. J. Phys. Anthrop., 49: 205-210. Bransby, E. R. 1945 The seasonal growth of children. Medical Officer, 73: 149-151. Eveleth, P. B. 1967-68 Physical growth of American children in the tropics. Revista de Antropologia. 1.5/26; 14-25. Fitt, A. B. 1941 Seasonal Influence on Growth. Function and Inheritance. New Zealand Council For Educational Research, Wellington. Johnston, F. E., M. Borden and R. B. MacVean 1973 Height, weight and their growth velocities in Guatemalan private school children of high socioeconomic class. Human Biology, 45: 627-641. Kemsley, W. F. F. 1953 Changes in body weight from 19431950. Annals of Eugenics, 18; 22-42. Luce, G. G. 1971 Biological Rhythms in Human and A n i ~ ma1 Physiology. Dover, New York. Marshall, W. A. 1971 Evaluation in growth rates in height over periods of less than one year. Archives of Disease in Childhood, 46: 414-420. 1975 The relationship of variations in children's growth rates to seasonal climatic variation. Annals of Human Biology, 2: 243-250. Nylin, G. 1929 Periodical variation in growth, standard metabolism and oxygen capacity of the blood in children. Acta Medica Scandinavia, 31: 1-207. Orr, J. B., and M. L. Clark 1930 Seasonal variation i n t h e growth of school children. Lancet, 2: 365-367. Snedecor. G. W. 1956 Statistical Methods. Fifth ed. Iowa State College, Ames. Takahashi, E. 1966 Growth and environmental factors in Japan. Human Biology, 38: 112-130. Tanner, J. M., a n d R . H. Whitehouse 1976 Clinical longitudinal standards for height, weight, height velocity, weight velocity and the stages of puberty. Archives of Disease in Childhood, 51: 170-179. Thompson, DArcy W. 1942 On Growth and Form. Second ed. Cambridge University, London. Vincent, M., and J. Dierickx 1960 Etude sur la croissance saisonniere des ecoliers de Leopoldville. Annales de la Societe Belge de Medecine Tropicale, 40: 837-844.