SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References

We examined whether birth weight (BW) predicts changes in body composition over a 6-year period in Swedish children and adolescents. For this purpose, a total of 247 children (55.5% girls) and 162 adolescents (60.5% girls) were included in the study and were followed up 6 years later. BW was obtained from parental records. We measured weight, height, waist circumference, and the bicep, tricep, subscapular, suprailiac, and medial calf skinfolds, and we calculated BMI, fat-free mass (FFM), and the sum of five skinfolds. Physical activity was assessed by accelerometry. Changes in pubertal status and baseline anthropometric estimates were used as confounders in all analysis. In the children cohort, we observed that BW was inversely associated with changes in BMI (β = −0.736, P = 0.002) and the sum of five skinfolds (β = −6.381, P = 0.009) regardless of confounders and physical activity, only in girls. We did not find any significant association between BW and adiposity gain estimates in the adolescent cohort. These findings give further support to the concept that low BW may have a programming effect of subsequent adiposity gain from childhood to adolescence. We also confirm the sex-related differences in the programming effect of body composition.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References

Worldwide the prevalence of obesity increased dramatically during the last decades.

Epidemiological studies, prospective birth cohorts and, more recently, genetic studies all indicate that the rapid weight gain trajectory to later obesity starts in the early infancy (1). The fetal period is considered as one of the critical windows for programming later fatness, cardiovascular disease and related morbidities in the long term (2,3,4,5,6). In this sense, there is increasing evidence that the origins of obesity may be very early in life (7). Body weight at birth weight (BW) is an established index of intrauterine environment (8).

Previous findings support the hypothesis that low BW would increase the risk of obesity development later in life by programming smaller proportions of lean body mass (9,10). A plausible explanation for this could be that individuals with lower lean tissue mass may have lower resting metabolic rate, which in the presence of an obesogenic environment (high energy-diets or sedentary lifestyle) may increase the risk of developing obesity later in life. There are few studies examining the programming effect of changes in body composition from childhood to adolescence, and none from adolescence to young adulthood, which are considered critical periods in the development of obesity (2).

The aim of the present study was to examine whether BW predicts changes in body composition over a 6-year period, i.e., from childhood to adolescence and from adolescence to young adulthood, in two groups of Swedish youth participants from the European Youth Heart Study (EYHS).

Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References

Study participants

The EYHS was designed to examine the interactions between personal, environmental, and lifestyle influences on the risk factors for future cardiovascular diseases (11). Study design, selection criteria, and sample calculations have been reported elsewhere (11).

Data collection was conducted in 1998–1999 (baseline data, EYHS-I) and 2004−2005 (follow-up data, EYHS-II). At baseline, 562 children (age ranging from 8.5 to 10.3 years, thereafter called “children cohort”) and 575 adolescents (age ranging from 14.7 to 16.7 years, thereafter called “adolescent cohort”) volunteered to participate in the EYHS-I. Six years later, 278 adolescents from the children cohort (152 female and 126 male) and 181 young adults from the adolescent cohort (111 female and 70 male) agreed to take part in the EYHS-II (Figure 1). The study was approved by the Research Ethics Committees of Örebro County Council (no. 690/98) and Huddinge University Hospital (no. 474/98). Written informed consent was provided by young participants (21 years) and by a parent or legal guardian by all children and adolescents.

image

Figure 1. Study design.

Download figure to PowerPoint

For this purpose, we identified a total of 247 children (∼9 years old, followed up when ∼15) and 162 adolescents (∼15 years old, followed up when ∼21) with complete data for BW and BMI at both baseline and at follow-up. Subjects from EYHS-I who participated in EYHS-II did not differ from those who did not in relation to anthropometric indices, physical activity, or fitness (12).

Newborn body weight

BW data were collected from parental recall by using a questionnaire to fill out at home. The validity of parents-reported BW data was verified in a randomly selected subset of the study sample by using available measured BWs from parent-held baby book directly obtained from hospital records (13).

Anthropometric variables

Height and weight were measured by standardized procedures, and BMI was calculated. Skinfold thickness were measured at the biceps, triceps, subscapular, suprailiaca, and calf areas on the left side of the body with a Holtain caliper (range: 0–40 mm; precision, 0.2 mm) according to Lohman's anthropometric standardization reference manual (14). Waist circumference was taken between the lower rib margin and the iliac crest, at the end of gentle expiration and waist-to-height ratio was calculated. We used the equations reported by Slaughter et al. to estimate body fat percentage (15) and thereafter fat-free mass (FFM) was calculated. The sum of five skinfolds was used as surrogates of total adiposity and waist circumference as surrogate of central adiposity. Measurements were made at baseline and at follow-up by the same trained investigators using essentially identical protocols.

Overweight and obesity status in children and adolescents was defined following the International Obesity Task Force that proposed sex- and age-adjusted BMI cutoff points (16). These cutoff points in childhood and adolescence are based on international data (97,876 males and 94,851 females from birth to 25 years of age) obtained from six large nationally representative cross sectional growth studies (Brazil, Great Britain, Hong Kong, the Netherlands, Singapore, and the United States) and linked to the widely accepted adult cutoff points of a BMI of 25 and 30 kg/m2 for adult overweight and obesity, respectively.

Pubertal stage was recorded by a researcher of the same gender as the child, after brief observation (Tanner stages from I to V) (17). Breast development in girls, and genital development in boys, were used for pubertal classification. The Tanner stage for the adolescent cohort at follow-up, i.e., young adulthood, was assumed to be V.

Physical activity

Physical activity was measured over 4 consecutive days (2 weekdays and at least 1 weekend day) with an activity monitor (MTI model WAM 7164; Manufacturing Technology, Shalimar, FL) attached on the right hip (18).

Statistical analysis

Statistical analyses were performed using “Statistical Package for the Social Sciences (SPSS)” software 17.0 (SPSS, Chicago, IL), and the level of significance was set at 0.05. Data are presented as means and s.d., unless otherwise stated.

Regression analysis was used to examine the association of BW with anthropometric variables at baseline and at follow-up (i.e., BMI, FFM, sum of five skinfolds, and waist circumference). The associations between BW and FFM, sum of five skinfolds and waist circumference (at baseline and at follow-up) were additionally adjusted for height (at baseline and at follow-up, respectively).

The relationships between BW and body composition changes over a 6-year period were also examined by linear regression analysis in an extended model approach. Model I included the predictor (BW) and the dependent variable (i.e., changes in BMI, FFM, sum of five skinfolds, and waist circumference) controlling for changes in Tanner stage number (Tanner stage at follow-up—Tanner stage at baseline). For model II, analysis where controlled for the confounding effect of the corresponding value of body composition at baseline. For model III, physical activity level at baseline was added. Changes in FFM, the sum of five skinfolds and waist circumference were additionally adjusted for changes in height. All the analyses were performed for the whole sample and separately by sex, and by sex and age group (children and adolescents cohorts).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References

Descriptive characteristics of the participants at baseline and at follow-up are shown in Table 1. In the children cohort, 15.4% were overweight at baseline and 14.2% at follow-up (P < 0.001). The percentage of overweight girls decreased from childhood to adolescence (P < 0.001). In the adolescent cohort, 9.9% were overweight at baseline and 19.1% at follow-up (P < 0.001); the percentage of overweight increased from adolescence to adulthood in both males and females (P > 0.001).

Table 1.  Characteristics of participants at baseline and at follow-up
inline image

Associations of BW with body composition at baseline and at follow-up

In girls from the children cohort, the results showed significant relationships between BW and both BMI and FFM at baseline (all P < 0.05). In boys from the children cohort, BW was significantly associated with both BMI and FFM at baseline and at follow-up (all P < 0.05), and with waist circumference at follow-up (adjusted P < 0.02, Table 2).

Table 2.  Regression coefficients (β) and standard errors (s.e.) showing relations between birth weight (kg) and body composition estimates at baseline and at follow-up
inline image

In the adolescent cohort, we did not find any significant association between BW and both BMI and body composition estimates in either women or men (Table 2).

Associations of BW with changes in body composition over a 6-year period

In girls from the children cohort, the relationships between BW and changes in either BMI or the sum of five skinfolds were statistically significant, regardless of the corresponding baseline total adiposity estimate (model II, Table 3) and physical activity at baseline (model III, Table 3). Thus, we observed that a decrease in 1 kg in BW predicted an increase of 0.72 kg/m2 in BMI (adjusted P = 0.001) and 7.0 mm in the sum of five skinfolds (adjusted P = 0.007), regardless of change in pubertal status, the corresponding baseline adiposity estimate and physical activity level at childhood (model III, Table 3). On the other hand, BW was negatively associated with changes in central adiposity (P < 0.05), but this relationship became nonsignificant when waist circumference at baseline was entered into the model (model II, Table 3) and the results did not change after further adjustment for physical activity level (model III, Table 3).

Table 3.  Regression coefficients (β) and standard errors (s.e.) showing relations between birth weight (kg) and changes (follow-up — baseline values) in body composition
inline image

In boys from the children cohort, BW was significantly associated with changes in FFM (P < 0.01) (model I, Table 3), but this relationship became nonsignificant when FFM at baseline was entered into the model (model II, Table 3) and the results did not change after further adjustment for physical activity level (model III, Table 3).

In the adolescent cohort, we did not find any significant relationship between BW and BMI or body composition estimates changes in neither men nor women.

The results were not altered when using follow-up anthropometric indices as outcomes, instead of using changes (follow-up-baseline) values and adjusting for the corresponding baseline adiposity values (data not shown).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References

The main objective of the present study was to investigate the relationship between BW and changes in body composition over two different developmental periods which are considered critical windows for the onset of obesity. We observed an inverse association between BW and subsequent total adiposity gain in girls from the children cohort, regardless of baseline adiposity estimate, change in pubertal status, or physical activity levels. These findings support the notion that low BW is a significant risk factor for later adiposity gain from childhood to adolescence and give further support to the sex-related differences in the programming effect of body composition.

In our study, BW predicted changes in total adiposity, as measured by the sum of five skinfolds, in girls from childhood (∼9 years) to adolescence (∼15 years), but not from adolescence (∼15 years) to young adulthood (∼21 years). One possible explanation for this difference could be that in the first transition period, from childhood to adolescence, environmental and behavioral factors have had less time to substantially modify the programming effect, than in the second period. Most of body composition changes take place during the first transition period, which includes pubertal maturation. Total body fat, lean body mass, and bone mass increase during pubertal maturation with considerable sexual dimorphism. Likewise, body fat percentage increase is higher in girls than in boys, whereas boys accrue larger amounts and proportions of lean body mass than girls over the same period. However, changes in body proportions and in body composition are less pronounced from adolescence to young adulthood. Also of note is that between 50 and 76% of the girls and boys from the adolescents cohort belonged to the Tanner stage V, indicating that most of the participants followed from adolescence to young adulthood was sexually mature already at baseline. This fact may contribute to explain the results observed. Nevertheless, we cannot exclude the possibility that the results could be reflecting some of the cultural and environmental background differences in the two cohorts at the baseline. Other environmental factors, such as consumption of high energy-diets or sedentary lifestyle, could also overwhelm or dilute the influence of BW on adiposity gain.

Greater awareness against excess adiposity in women from the adolescent cohort could also explain the lack of association between BW and adiposity gain. In this sense, healthy lifestyle behaviors might attenuate the effect of the adverse effects of a lower BW. Although the current literature on this topic is scarce, some studies support that physical activity might be beneficial in elderly people (19) and adults (20) with higher risk for diseases due to their low BW. Likewise, the obesity epidemic in Sweden is still rise, but slowing has been shown among adult women from urbanized areas with high education level. This tendency was explained by the increase in physically active behaviors observed in this group (21).

Although both girls and boys are susceptible to programming stimuli during pregnancy, previous studies have provided important new insights identifying marked sex differences. Sex-specific programming has also been reported for body composition (9), blood pressure (22), lipid levels (5), or insulin action (23), but the reasons for the differences between male and female subjects remain unexplained.

There are few studies examining the relationship between BW and adiposity gain in youths. Kelly et al. (24) found no relationship between BW and subsequent adiposity gain in a well-characterized sample of overweight adolescents. However, differences in the study design makes comparisons between studies difficult. Likewise, they performed the study with overweight participants while overweight prevalence in our sample was low, they did not examine the possible sex-related differences in the programming effect of adiposity gain; and their participants had initiated pubertal maturation at baseline (age range from 11 to 17 years). To our knowledge, ours is the first longitudinal study to examine the relationship between BW and subsequent body composition changes across different critical periods of obesity development and examining the sex-related differences in the programming effect.

In our study, BW was not associated with changes in overweight-obesity status either in the children or in the adolescent's cohort. However, it is difficult to examine the effect of BW on overweight development in our study due to the relatively small sample size and, hence, the low number of overweight (N = 38 and N = 35 in the children cohort at baseline and at follow-up, respectively, and N = 16 and N = 31 in the adolescent cohort at baseline and at follow-up, respectively) in our study, which reflects the low prevalence of overweight people in Sweden compared with other developed countries (25). Previous studies showed that the overweight prevalence in Sweden according to IOTF definitions was 17% (26) in 7–9 years old children and 11.4% in 15–16 years old adolescents (27).

On the other hand, we found that BW was associated with FFM in boys from the children cohort (from childhood to adolescence) at both baseline (∼9 years) and follow-up (∼15 years) and in girls at baseline (∼9 years), regardless of height. Moreover, higher BW was associated with higher increase on FFM in boys from childhood to adolescence, but this relationship seems to be explained by baseline FFM and physical activity level. These data suggest that BW could program greater FFM in children of both sexes, but not changes in FFM in the transition from childhood to adolescence. Other studies indicated that the association between BW and later body weight could in fact be related to the programming of greater lean tissue mass (9,10,28), suggesting that a high BW programs body composition rather than simply predisposing to greater body size.

Although this study includes several strengths, as that it was conducted in two longitudinal cohorts of children (from childhood to adolescence) and adolescents (from adolescence to young adulthood) and the objectively measured physical activity data, we acknowledge the limitations. First, the use of parental records of BW may be subject of errors, although we tested the validity of data in a subset of sample. Second, skinfold thickness measurement is not the reference technique to assess body composition and it has some limitations. Nevertheless, as the reference methods are costly, for the studies that involve large numbers of subjects, skinfolds are a valid alternative for the measurement of total adiposity. Third, pubertal status was assessed only by inspection and an increased adiposity could result in an overestimation of the Tanner stage in girls. Unfortunately, there was no information about pregnancy history of the females. Fourth, we do not know whether a change of 0.5 kg/m2 in BMI over a 6-year period may increase the risk of developing cardiovascular disease in the future. Finally, replication of our findings in larger sample is essential because the number of study participants was relatively small, particularly in the adolescent cohort.

In summary, our results give further support to the concept that fetal undergrowth, as reflected by BW, may have a sex-specific programming effect of subsequent adiposity gain from childhood to adolescence. These findings may contribute to explain the relationship between low BW and later metabolic disorders, and may justify prevention strategies in mothers to optimize maternal health to increase the likelihood of healthy BW (29).

ACKNOWLEDGEMENT

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References

This study was funded by grants from Stockholm County Council and by the Swedish Council for Working Life and Social Research (FAS), and by grants from FAS, from the Swedish Heart-Lung Foundation (20090635), from the Spanish Ministry of Education (EX-2008-0641), and the Spanish Ministry of Science and Innovation (RYC-2010-05957).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. ACKNOWLEDGEMENT
  8. DISCLOSURE
  9. References