INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THERMOREGULATORY MECHANISMS are susceptible to modification by early postnatal experience (17, 22). Animals reared at lower environmental temperatures exhibit improved defense on subsequent exposure to cold than those reared at higher temperatures (4, 12, 15). Conversely, animals reared at elevated temperatures are more tolerant of exposure to heat than those reared at lower temperatures (13, 14). Similar effects of early temperature exposures have also been noted in humans (5, 11). Effects of temperature exposure in early life on thermoregulation appear to be qualitatively different from those of cold and heat acclimatization induced by environmental exposures in adult animals (5, 8, 10, 17). The mechanisms underlying the effects of early temperature exposure are unknown, but because the functional properties of other portions of the central nervous system appear to be programmed by sensory experience during critical periods of postnatal life (2, 21), the possibility exists that the development of thermoregulatory function is likewise influenced by the temperature of the postnatal environment.

In addition to these functional effects, animals reared at warm and cool temperatures also display different morphological characteristics. Animals reared under warm conditions have longer extremities and tails and larger salivary glands, all changes that would facilitate dissipation of heat (17). Cold-reared animals grow more hair (17). Also, cold-rearing increases the mass and functional capacity of brown adipose tissue as well as the density of the sympathetic innervation in the tissue (1, 19).

In conjunction with studies examining the impact of rearing temperature on sympathetic nervous system (SNS) development, the following experiments were designed to assess the effect, if any, of early temperature exposure on body weight, fat pad weight in gonadal and retroperitoneal sites, and food intake. Because thermogenic responses to cold exposure share many similarities with those to increased caloric intake (16), these studies tested the hypothesis that animals reared in a cool environment (18°C), which possess greater capacity for thermogenesis, would be less susceptible to weight gain when presented with an opportunity to overeat (10% sucrose to drink) than animals reared at a warm temperature (30°C). In striking contrast to these expectations, however, 18°C-reared animals, both male and female, were more prone to gain weight and to accumulate fat in intra-abdominal locations than animals reared at 30°C.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animals. One-day-old male or female CD rats with foster mothers or timed pregnant CD rats were obtained from Charles River Breeding Laboratories (Wilmington, MA). On the day after arrival or the day after delivery, litters were culled to 10 pups each. Plastic cages containing individual mothers and litters were then placed into one of two temperature-controlled chambers set at 18 and 30°C (±0.2°C). Both chambers (model ST 50 GC/M; Sure-Temp, Apex, NC; 50 ft³ internal volume) were equipped with double glass doors so that illumination was provided by room lighting as well as by a timer-controlled internal light set to coincide with the 14:10-h light-dark cycle of the room. Pups were weaned at 21-22 days and housed 4 or 5/cage within the chambers. Animals were removed from these chambers at 60 days of age, unless otherwise specified, and were housed 2 or 3/cage for the duration of the studies in a room maintained at 21 ± 2°C. Animals used in this study were maintained in accordance with the guidelines and with the approval of the Animal Care and Use Committee of Northwestern University Medical School.

Feeding protocol. Unless otherwise specified, animals were provided free access to water and standard laboratory chow (Prolab R-M-H 3000; Agway, Syracuse, NY). In the overfeeding studies, animals were housed in pairs and received either chow or chow plus a 10% (wt/vol) solution of sucrose. Bottles containing sucrose were replaced daily, and both groups received water to drink as well. In studies in which food intake was monitored, chow consumption was measured for each cage over a 4-day period from Monday to Friday, and sucrose consumption was measured daily during the 4-day interval. Caloric intake was calculated on the basis of digestible energy of the chow (3.7 kcal/g) and of sucrose (3.94 kcal/g).

Assessment of abdominal fat. At the end of each experiment, animals were killed by CO₂ inhalation. In male rats, epididymal and retroperitoneal fat pads were removed and weighed separately, whereas in female rats parametrial and retroperitoneal fat pads were removed and weighed together. Epididymal fat was separated from epididymis, epididymal vessels, and vas deferens, and parametrial fat was freed from salpinx, ovary, and uterus. Retroperitoneal fat was removed bilaterally from a region bounded caudally by the inguinal region, medially by the midline, cranially by the diaphragm, and laterally as far as fat was apparent. In female rats, parametrial and retroperitoneal fat were taken together. Abdominal fat was taken as the sum of retroperitoneal and gonadal fat pads. Dissection of adipose tissue depots was carried out by the same individual in all studies.

Data analysis. Data are displayed as means ± SE, unless otherwise noted. Statistical analyses of variance (ANOVA) and covariance were performed using either Data Desk 5.0 statistical software (Data Description, Ithaca, NY) or BMDP program 2V (6). Post hoc, pair-wise comparisons following ANOVA utilized Scheffé's test. Analyses of caloric intake included weekly body weight measurements as covariates in a repeated-measures analysis of the weekly estimates of caloric intake. Analyses of correlation and partial correlation employed BMDP program 6R (7).

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Effect of rearing temperature on body weight and epididymal fat. Body and epididymal fat pad weights of male rats at 60 and 120 days of age are presented in Fig. 1. After the end of 2-mo period of differential temperature exposure, rats reared at 18°C were 16% heavier than those reared at 30°C (351 ± 8 vs. 304 ± 6 g, P < 0.0001). Epididymal fat pad weights were also increased in the 18°C-reared rats (3.71 ± 0.45 vs. 2.64 ± 0.14 g, P = 0.0139). The relation between fat pad weight and body weight, however, differed in the two groups (Fig. 1). Not only was fat pad weight increased in relation to body weight in the 18°C-reared animals (P = 0.0057), but the slope of the line relating fat pad to body weight was steeper as well (P = 0.0061). At 120 days of age, 60 days after removal of the animals from the temperature-controlled chambers, 18°C-reared rats remained heavier (546 ± 9 vs. 500 ± 9 g, P = 0.0009) and displayed larger epididymal fat pads in absolute weight (12.11 ± 1.01 vs. 8.89 ± 0.48 g, P = 0.0041) and in relation to body weight (P = 0.0009). The slope of the relationship between epididymal fat pad and body weight also remained elevated in the 18°C-reared rats (P = 0.0008). These data imply that as 18°C-reared male rats gain weight more of that weight gain represents epididymal fat pad enlargement than in 30°C-reared rats. Moreover, this effect persists for at least 60 days when both groups are housed at a common temperature.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Effect of rearing temperature on relation between epididymal fat and body weight in 60- (A) and 120-day-old (B) male rats. , Animals housed from 1 to 60 days at 18°C; , animals housed at 30°C.

Effect of rearing temperature on responses to sucrose in male rats. In a separate experiment, 18- and 30°C-reared animals were subdivided into two equal-weight groups 1 wk after removal from the temperature-controlled chambers and provided access either to chow alone or to chow plus a 10% sucrose solution to drink. Body weights at the beginning and end of the 7-wk feeding regimen and weight gain in grams and as percent of initial weight in the four rearing temperature (T_rear):diet groups are presented in Table 1. Overall, 18°C-reared rats weighed more than 30°C-reared animals by repeated-measures ANOVA. Sucrose-fed rats gained more weight than their chow-fed counterparts. This effect of sucrose was more pronounced in 18- than in 30°C-reared rats. Animals reared at 18°C given sucrose gained 51% more weight than chow-fed controls (254 ± 21 vs. 168 ± 10 g, P = 0.0006), whereas 30°C-reared animals gained only 16% more weight with sucrose than with chow (210 ± 15 vs. 181 ± 8 g, P = NS). As percentage of initial body weight, sucrose feeding increased weight gain in 18°C-reared rats from 42.1 ± 1.6 to 61.6 ± 3.2% (P = 0.0002), whereas in 30°C-reared animals weight gain rose from 52.6 ± 1.8 to 60.6 ± 3.7% (P = 0.0456) with sucrose feeding. Epididymal fat pad weights from the animals in this experiment are shown in Fig. 2. As can be seen, fat pad weight was increased, both in absolute weight and in relation to body weight (P = 0.0058 and 0.0024, respectively, for T_rear × diet interaction) in the 18°C-reared, sucrose-fed animals. Thus, under room-temperature conditions, the increments in weight gain and fat accumulation in epididymal fats pads observed in sucrose-fed animals compared with chow-fed controls were greater in the animals reared at 18°C than in those raised at 30°C.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Effect of rearing temperature on relation between epididymal fat and body weight in male rats fed chow or chow plus sucrose. Epididymal fat pads were obtained at the conclusion of the study presented in Table 1. Data are presented as means ± SE for 5 or 6 animals in each group. Left: epididymal fat pad weight in grams. Right: the same data as percent of body weight (%BWt). Open bars, animals fed chow; solid bars, animals fed sucrose in addition to chow.

Effect of rearing and housing temperatures on responses to sucrose in male rats. To assess the continuing importance of environmental temperature, a separate experiment was conducted in which temperature was varied during the feeding portion of the study. As in the previous experiment, this study compared the impact of T_rear on body weight and fat pad weight responses to sucrose feeding. In addition, animals in each of the four T_rear:diet groups were housed at either room temperature (23°C) or at 18°C from 60 to 120 days of age. The feeding protocol began 1 wk after animals were transferred to their new environments. Body weights at the beginning and end of the 8-wk feeding protocol and weight gain in grams and as percent of initial weight in the eight T_rear:housing temperature (T_hous):diet groups are presented in Table 2. As in the previous experiment, 18°C-reared animals were heavier overall (P < 0.0001). Under both T_hous levels, animals reared at 18°C gained on average 6.9 g more than rats reared at 30°C (P = NS, T_rear × week interaction). T_rear did, however, affect weight gain as a function of diet (P = 0.0515, T_rear × diet × week interaction; T_rear × diet interaction: P = 0.0515 and P = 0.0452 for weight gain by gram and by percent body weight, respectively). Rats reared at 18°C gained 69.8 g more with sucrose feeding than with chow (P < 0.0001 by gram and by percent body weight), whereas in 30°C-reared animals the weight gain differential was only 27.5 g (P = 0.0733 by gram and P = 0.0167 by percent body weight). In addition, animals housed at room temperature gained 25.1 g more weight than those housed at 18°C (P = 0.0217, T_hous × week interaction). T_hous also influenced weight gain as a function of diet (P = 0.0045, T_hous × diet × week interaction; T_hous × diet interaction: P = 0.0045 and P = 0.0003 for weight gain by gram and by percent body weight, respectively). Rats housed at room temperature gained 80.2 g more with sucrose feeding than with chow (P < 0.0001), whereas in animals housed at 18°C the difference in weight gain was only 17.1 g (P = NS).

Epididymal and retroperitoneal fat pads were removed and weighed at the end of the experiment and the results (expressed as percent of body weight) are presented in Fig. 3. Overall, in animals reared at 18°C epididymal fat pads were 22% larger (in relation to body weight) and retroperitoneal fat pads 26% heavier than comparable fat pads in rats reared at 30°C (P = 0.0024 and 0.0007, respectively). Animals fed sucrose displayed a 23% enlargement in epididymal fat and a 39% increase in retroperitoneal fat over fat pads obtained from chow-fed animals (P = 0.0015 and P < 0.0001, respectively). Although T_hous did not affect epididymal fat pad weight, retroperitoneal fat pads were 26% larger in rats housed at room temperature than in those maintained at 18°C (P = 0.0009), a difference that was greater in sucrose-fed than chow-fed animals (T_hous × diet interaction, P = 0.0213).

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Effect of rearing and housing temperature on epididymal (A) and retroperitoneal (B) fat weight in relation to body weight in male rats fed chow or chow plus sucrose. Epididymal fat pads were obtained at the conclusion of the study presented in Table 2. Data are presented as means ± SE for 8 animals in each group. Open bars, animals housed from 60 to 120 days at 18°C and fed chow; solid bars, animals housed at 18°C and fed sucrose in addition to chow; hatched bars, animals housed at 23°C and fed chow; crosshatched bars, animals housed at 23°C and fed sucrose in addition to chow.

Estimates of caloric intake averaged over the 8 wk of sucrose feeding are presented in Table 3. Caloric intake was increased in the 18°C-reared animals compared with 30°C-reared rats (P = 0.0002), in animals housed at 18°C compared with those housed at room temperature (P = 0.0006), and in sucrose-fed compared with chow-fed animals (P = 0.0099). Although the effects of T_hous and diet on caloric intake remained statistically significant when body weight or body weight^0.75 was included as a covariate in the analysis, the effect of T_rear did not. The mean increase in caloric intake (adjusted for body weight^0.75) in animals housed at 18°C rather than at room temperature, however, was greater in animals previously reared at 18°C (+61 kcal · cage¹ · day¹) than in those reared at 30°C (+42 kcal · cage¹ · day¹, T_rear × T_hous interaction, P = 0.0456). Without adjustment for body weight, caloric intake during the 8-wk period declined by 12% on average (P < 0.0001) despite a mean increase in body weight of 61%, indicating that body weight per se was not the sole determinant of caloric intake.

Effect of T_rear on body weight and abdominal fat in female rats. Studies similar to those described above in male animals were also carried out in female rats. Body weight and weight of abdominal (parametrial plus retroperitoneal) fat pads of female rats at 60 and 122 days of age are presented in Fig. 4. After 2 mo of exposure, rats reared at 18°C were 11% heavier than those reared at 30°C (253 ± 11 vs. 229 ± 6 g, P = 0.0825). Moreover, in contrast to the effects of T_rear in male rats, abdominal fat pad weights were decreased in the 18°C-reared rats (3.08 ± 0.33 vs. 4.27 ± 0.38 g, P = 0.0349). The disparity in fat pad size was even greater when expressed in relation to body weight (1.20 ± 0.10 %body wt in 18°C-reared animals vs. 1.86 ± 0.16 %body wt in 30°C-reared rats, P = 0.0030). Moreover, the relation between fat pad weight and body weight did not differ in the two groups (Fig. 4). At 122 days of age, 18°C-reared rats were significantly heavier (392 ± 20 vs. 327 ± 11 g, P = 0.0138) and displayed larger abdominal fat pads in absolute weight (29.09 ± 4.41 vs. 15.17 ± 1.73 g, P = 0.0108) and in relation to body weight (P = 0.0134). The slope of the relationship between fat pad weight and body weight was slightly, but not significantly, elevated in the 18°C-reared rats (P = 0.0928). These data indicate that female rats while living at 18°C accumulate less abdominal fat than their 30°C-reared counterparts, but that once they are placed at a common, intermediate temperature, they increase abdominal fat to a greater extent than 30°C-reared rats.

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Effect of rearing temperature on relation between abdominal fat and body weight in 60- (A) and 122-day-old (B) female rats. , Animals housed from 1 to 60 days at 18°C; , animals housed at 30°C.

Effect of T_rear on responses to sucrose in female rats. After removal from the temperature-controlled chambers, 18- and 30°C-reared female rats were divided into two groups 1 wk later and given access to either chow alone or to chow plus 10% sucrose. Initial and final body weights and weight gain over the 8-wk feeding period are presented in Table 4. Overall, body weight in 18°C-reared female rats was greater than in 30°C-reared animals (P = 0.0015). Moreover, 18°C-reared rats also gained more weight, either in absolute terms or in relation to initial weight, than 30°C-reared females (P = 0.0064 and 0.0182, respectively). Weight gain was greater in sucrose-fed than chow-fed rats (diet × week interaction, P = 0.0006; diet effect on weight gain: P = 0.0006 and P = 0.0002 for weight gain by gram and by percent body weight, respectively). In contrast to responses in male rats, these effects of T_rear and diet were additive in female animals (i.e., the T_rear × diet × week interaction on body weight and the T_rear × diet interaction on weight gain were not statistically significant). Weight gain in 18°C-reared animals given sucrose was 78 g and 28.3% greater than in chow fed controls (P = 0.0033 and P = 0.0022 by gram and percent body weight, respectively), and in 30°C-reared animals weight gain was 55 g and 22.1% greater than in chow-fed controls (P = 0.0324 and P = 0.0136 by gram and percent body weight, respectively). Thus T_rear affects weight gain in female as in male rats, an effect that is additive with that of diet.

Abdominal fat pad weights from the animals in this experiment are shown in Fig. 5. As can be seen, T_rear and diet also exerted additive effects on fat pad size, both in absolute weight (T_rear, P = 0.0008; diet, P = 0.0044) and in relation to body weight (T_rear, P = 0.0011; diet, P = 0.0039). Estimates of caloric intake in this experiment are presented in Table 5. Caloric intake was increased in the 18°C-reared animals compared with 30°C-reared rats and in the sucrose-fed animals compared with the chow-fed controls. These differences were not observed when body weight or body weight^0.75 was included as a covariate in the analysis. Without adjustment for body weight, however, caloric intake declined by 19% overall during the period of observation (P < 0.0001) and to a greater extent in sucrose-fed (25%) than in chow-fed animals (10%, P = 0.0006).

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Effect of rearing temperature on relation between abdominal fat and body weight in female rats fed chow or chow plus sucrose. Epididymal fat pads were obtained at the conclusion of the study presented in Table 4. Data are presented as means ± SE for 8 animals in each group. Left: abdominal fat pad weight in grams. Right: the same data as percent of body weight. Open bars, animals fed chow; solid bars, animals fed sucrose in addition to chow.

Comparison of effects of T_rear on abdominal fat in male and female rats. Because the previous experiments in male and female rats raised the possibility of a gender difference in the impact of T_rear on abdominal fat accumulation, an additional study was performed comparing directly the effect of T_rear in male and female rats. In this study, mothers and pups were placed in the temperature-controlled chambers 1 day postpartum. At 60 days of age, animals were removed from the chambers and measurements were made of abdominal fat in male and female rats. The results are shown in Fig. 6. Body weights were greater in males than females (280 ± 12 vs. 229 ± 5 g, P < 0.0001) and greater in 18°C- than 30°C-reared animals (280 ± 13 vs. 231 ± 5 g, P < 0.0001). In addition, the differential effect of T_rear on body weight was greater in males (99.4 g) than in females (33.0 g, P = 0.0002). With respect to abdominal fat, the effect of T_rear on weight of gonadal and retroperitoneal fat pads was exclusively evident in male rats and was not at all apparent in female animals (T_rear × gender, P < 0.0001). In relation to body weight, however, 18°C-reared males rats displayed more abdominal fat than 30°C-reared males (P = 0.0058), whereas 18°C-reared females exhibited less fat than 30°C-reared females (P = 0.0487). Thus this study of abdominal fat accumulation in male and female littermates confirmed previous suggestions that gender influences the impact of T_rear on abdominal fat accumulation.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Comparison of effect of rearing temperature on relation between abdominal fat and body weight in male and female littermates fed chow or chow plus sucrose. Abdominal fat pads were obtained at 60 days of age. Data are presented as means ± SE. Left: abdominal fat pad weight in grams. Right: the same data as percent of body weight. Open bars, 18°C-reared animals; solid bars, 30°C-reared rats. Numbers in each bar refer to the number of animals in each group.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The studies described in this report examined the effect of environmental temperature during the first 2 mo of life on subsequent changes in body weight and in abdominal fat pad weights. The findings show that rats reared at 18°C gained more weight and accumulated more fat in gonadal and retroperitoneal depots than animals reared at 30°C when both were housed in an environment of intermediate temperature. Moreover, ad libitum access to a 10% sucrose solution in addition to the lab chow diet exacerbated weight gain and fat accumulation in the 18°C-reared animals. Differences between male and female rats were also evident, as 30°C-reared male rats were relatively resistant to the effects of sucrose compared with 18°C-reared males, whereas female rats appeared comparably susceptible whether reared at 18 or 30°C. In addition, exposure to 18°C increased fat deposition in abdominal fat pads of male rats, but decreased it in female animals compared with males and females reared at 30°C (Fig. 6). Consequently, raising rats at an environmental temperature of 18°C for the first 2 mo of life promotes weight gain and expansion of abdominal fat stores in both male and female rats when they are subsequently housed at a warmer room temperature.

Data for body weight were analyzed as gain in weight both in gram and as percent change over the 7- or 8-wk feeding period. This use of both additive and proportional models for data analysis was necessitated by the difference in body weights between 18- and 30°C-reared rats (male > female) at the start of the feeding studies in 9- to 10-wk-old rats. Because 18°C-reared male rats weighed, on average, 50-80 g more than the 30°C-reared males before first exposure to the 10% sucrose, it was necessary to demonstrate that greater weight gain with sucrose feeding was not merely a consequence of the higher initial weight in the 18°C-reared animals. Analysis of percent change in body weight revealed that 18°C-reared rats fed sucrose gained more weight in grams and slightly more as a percentage of initial weight than either diet group of 30°C-reared rats, but substantially more than 18°C-reared rats fed chow alone. Consequently, the differential effect of sucrose ingestion on body weight was greater in 18- than in 30°C-reared male rats. In female rats, weight gain in grams and as percent of initial weight was greater in both sucrose- and chow-fed 18°C-reared rats than in comparable groups of 30°C-reared animals.

Differences in weight gain between 18- and 30°C-reared animals could be consequent to alterations in food intake, in energy metabolism, or in both. In two of three feeding studies, attempts were made to monitor food intake for 4 days each week over the 8-wk feeding protocol (Tables 3 and 5). These attempts were also confounded by differences in body weight, both between rearing groups and between diet groups over the course of the study. Although in both experiments 18°C-reared rats ingested more energy than 30°C-reared animals, these differences were not apparent when the weekly estimates of food intake were related to the weekly measurements of body weight by covariance analysis. Thus in a study of the current design it is not possible retrospectively to disentangle the effect of body weight on food intake (larger animals presumably eat more) from the effect of food intake on body weight (greater food intake promotes increased weight gain).

The available data regarding weight gain and energy intake, however, indicate that differences in energy intake are not sufficient to explain greater weight gain either in 18- versus 30°C-reared rats or in sucrose- versus chow-fed animals. For example, in the 30°C-reared female rats (Tables 4 and 5), energy intake was only 20% greater (without weight adjustment) in sucrose-fed compared with chow-fed rats and yet sucrose-fed rats gained 71% more weight. After adjustment for body weight, energy intake in the 30°C-reared females was identical in sucrose- and chow-fed groups; nonetheless, sucrose-fed rats gained proportionally more weight than chow-fed animals and accumulated more of that weight in abdominal fat pads (Fig. 5). Consequently, these and other similar disparities between group differences in weight gain and fat accumulation and in energy intake strongly suggest that environmental temperature during the first 2 mo of life affects the regulation of energy metabolism, although the nature of this effect is unclear at the present time. Because body composition was not specifically measured in these studies, it is possible, however, that group differences in weight gain may not fully represent changes in energy storage.

These studies were designed in conjunction with others examining the effect of T_rear on SNS function. In both male and female rats, sympathetic innervation to brown fat is increased in 18- compared with 30°C-reared animals, a difference that persists for up to 4 mo after removal from the temperature-controlled chambers (unpublished observations). Activity of these sympathetic nerves to brown fat, assessed in unanesthetized animals housed at room temperature using the technique of [³H]norepinephrine turnover, is greater in both male and female rats reared at 18°C than in those raised at 30°C (unpublished observations). The relation of these developmental alterations in brown fat sympathetic function to the differences in weight gain and fat accumulation noted here is presently unclear. Because sucrose ingestion stimulates sympathetic activity in brown fat (23), further studies are required to determine whether T_rear modifies the SNS response to dietary carbohydrate.

Effects of environmental temperature and diet have been shown previously to affect development of gonadal and retroperitoneal fat pads differently. Exposure of young, growing rats to a temperature of 5°C induced lipolytic and hyperplastic responses in epididymal, but not retroperitoneal, fat pads (18). High-fat feeding, by contrast, induced greater hyperplasia in retroperitoneal than gonadal fat pads, although hypertrophic responses were similar (9). In the current studies, retroperitoneal fat pad weights (in relation to body weight) were 9% greater in male rats housed at 23°C when fed chow and 39% greater when fed sucrose than those housed at 18°C, whereas in the same animals epididymal fat pad weights were 1% smaller in chow-fed and 11% larger in sucrose-fed rats. Consequently, retroperitoneal fat pads appear more responsive than epididymal fat to the synergistic effects of diet and concurrent environmental temperature. Similar comparisons were not available from female rats because no attempt was made to distinguish between gonadal and retroperitoneal fat.

Although the studies described here were carried out in laboratory animals, data are available to suggest that early exposure to cooler environmental temperatures may increase body weight in humans as well. In 1955, Newman and Munro (20) published their analysis of height and weight data from 15,216 young, healthy Caucasian males inducted into the US Army between 1946 and 1953. These investigators divided their subjects into subsets by state of birth and analyzed the relations between state means for height and weight and mean state temperature in the months of July and January (obtained from US Weather Bureau for 1952). This relationship has been reanalyzed here using body mass index (BMI, kg/m²) from the state averages for height and weight contained in APPENDIX A of their paper (20). Analyses of correlation and partial correlation were weighted by the number of subjects for each state. The relation between BMI and the mean January temperature is illustrated in Fig. 7; the weighted correlation coefficient is 0.77. Partial correlation between BMI and mean January temperature (holding mean July temperature constant) is also high (r = 0.59). A similar analysis of BMI and mean July temperature yields a simple, weighted correlation of 0.61, whereas weighted partial correlation between these two variables (with mean January temperature constant) is only 0.06. Because the weighted correlation between mean January and July temperatures in their data set was 0.76 (r = 0.78, unweighted), these data suggest that the relationship between BMI and temperature is stronger with January cold than with July heat. Newman and Munro (20) considered and then rejected the possibility that differences in ethnic populations might confound their analysis of Caucasian, American males, a conclusion that cannot be reexamined readily at this time. Although the data of Newman and Munro do not provide definitive proof of the lasting impact of thermal conditions during rearing in a human population, they are nonetheless consistent with the current findings in a laboratory model of early temperature exposure.

View larger version (13K): [in this window] [in a new window]   Fig. 7.   Relation between body mass index (BMI) of young Caucasian males and the mean January temperature of the state in which they were born. BMI was calculated from state averages for height and weight from the data of Newman and Munro (20).

Perspectives