| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research Communication |
1 From the Oulu Regional Institute of Occupational Health, Oulu, Finland; the Centre for Paediatric Epidemiology and Biostatistics, Institute of Child Health, London; the Department of Health and General Practice, University of Oulu, Finland; and the Department of Public Health, Imperial College School of Medicine, London.
2 Supported by the Academy of Finland, Yrjö Jahnsson Foundation. 3 Address reprint requests to J Laitinen, Oulu Regional Institute of Occupational Health, Aapistie 1, 90220 Oulu, Finland. E-mail: jaana.laitinen{at}occuphealth.fi.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objective: The objective was to evaluate the associations between BMI at 31 y of age and family social class during early childhood, maternal body mass index (BMI) before pregnancy, BMI at birth and at 1 and 14 y of age, and age at menarche.
Design: This was a longitudinal study of the northern Finland birth cohort for 1966. Subjects were measured at birth and at 1, 14, and 31 y of age. The analysis was restricted to individuals for whom BMI data were available for all measurement points (n = 2876 males and 3404 females).
Results: The mean BMI at birth was highest in offspring from the highest social classes, but BMI was inversely related to social class at 1 y. BMI, the waist-to-hip ratio, and the proportion of obese subjects were inversely related to social class at 31 y. The heavier the mother, the heavier the offspring from birth to 31 y. The paired analyses between maternal BMI and daughter's BMI at 31 y showed no significant difference in BMI after adjustment for the age difference. BMI at 14 y was the most important predictor of BMI at 31 y. Early menarche in females was associated with a higher BMI at 14 and 31 y.
Conclusions: Differences in BMI by social class are formed at least partly during early childhood. Low social class of the child's family, a high maternal BMI before pregnancy, a high BMI during adolescence, and early menarche are predictors of obesity in adulthood.
Key Words: Body mass index • BMI • obesity • social class • longitudinal studies • childhood • adolescence • adulthood • age at menarche • cohort study
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Obesity is inversely related to socioeconomic status in the developed world, especially among white women (9). Children from lower social classes are more likely to become overweight or obese than are children from higher social classes and are more likely to remain overweight or obese throughout early adulthood (10, 11). Genetic factors are also related to the development of obesity, as was observed in some studies of twins and adopted persons (12–14). The likelihood that a child will become obese in adulthood is markedly increased if both of his or her parents are obese (15, 16), if he or she is obese during childhood or adolescence (4), or if puberty occurs at an early age (2). Most obese adults, however, were not obese as children (2–4). The percentage of obese adults who were obese as children varies from 5% to 60% (4), depending on the cutoff used to define obesity.
Few longitudinal studies have examined the role of factors related to early childhood in adulthood obesity. In this longitudinal study of the northern Finland birth cohort for 1966, we hypothesized that a low family social class, a high maternal BMI, a high BMI during the subjects' childhood, and menarche occurring at an early age predict overweight and obesity at the age of 31 y. We are not aware of any previous studies that simultaneously considered these factors as predictors of obesity.
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
600 g (18). Thirteen women delivered twice. The total number of children born was 12231, of whom 12058 were born alive, representing 96% of all births in the district. At 31 y of age, 11637 subjects (5906 men and 5731 women) were alive, but a questionnaire sent in 1997–1998 was not received by 96 of them. The analysis was additionally restricted to those who responded to the questionnaire (75.3%, n = 8767), who gave their written consent for the available data to be used for research purposes (n = 8690), and whose BMI data at birth and 1, 14, and 31 y of age were available (2876 men and 3404 women). This study was approved by the Ethical Committee of the Finnish Institute of Occupational Health and by the Ethical Committee of the University of Oulu.
Data collection
In 1965–1966, midwives at all antenatal and postnatal clinics in the provinces of Oulu and Lapland used structured study forms to collect information concerning the mother and her pregnancy from the sixth month of gestation until term (17). Data on the mother's weight, height, and social class were obtained from the standard forms for the pregnancy and from the maternity cards carried by all mothers. The offspring were followed up at 1, 14, and 31 y of age. Public health nurses used the special forms of the child health cards to collect data on growth at 1 y of age. Data for 14 y were obtained from a questionnaire that had been sent to the participants in 1980. The most recent follow-up, at 31 y, took place in 1997–1998; data were obtained from a questionnaire that had been sent to the participants, a clinical examination, or both.
Definition of explanatory variables and outcome
The mother's weight before pregnancy was recorded during her first visit to the antenatal clinic, which occurred on average during week 16 of gestation. Height was measured or self-reported. The mother's age was her age before pregnancy and was calculated as the date of the offspring's birth minus the gestational age of the offspring in days (or 280 d if the gestational age was missing) minus the date of the mother's birth. The age of the offspring was calculated as the date of the health examination minus the date of birth. Ages in years (ages in days/365.25) were used subsequently in the analyses.
Social classes were defined on the basis of the father's occupation in 1966 and its prestige (19, 20); data recorded in 1966 were used. Social classes I and II include occupations with the highest prestige and the longest education, eg, elementary school teachers, dentists, graduated engineers, priests, and office managers. Social class III includes skilled workers, eg, clerks and stewards, and social class IV includes unskilled workers, eg, domestic help, seamen, and cleaners. Those with no occupation or with an unknown occupation were included in social class IV. Farmers form a social class of their own. For 234 subjects with missing data, social class was based on corresponding information from the mother.
Body weight and height were measured at birth and at 1 and 31 y but were self-reported at 14 y. If weight or height was not measured at 31 y, self-reported measurements were used (30%). BMI (in kg/m2) was calculated at birth and at 1, 14, and 31 y. Adult BMI at 31 y was classified as follows: underweight (<18.5), normal weight (18.5–24.9), overweight (25.0–29.9), and obese (30.0) (1). At birth, BMI was used for comparability with other measurement points because there appear to be no simple alternative indexes for weight relative to height for infants. Adolescence (age 14 y) overweight was defined as a BMI 85th to <95th percentile and obesity as a BMI 95th percentile, both by sex (21). The waist-to-hip ratio was measured as the ratio between the circumferences of the waist (at the level midway between the lowest rib margin and the iliac crest) and the hip (at the widest trochanters). For the female cohort members, age at menarche was indicated by the age at which menarche started, as reported in the questionnaire at 31 y.
Statistical methods
Cross-tabulation and the usual summary measures (means and SDs) were used as descriptive statistics. Analysis of covariance was used to adjust for maternal age in comparisons of mean BMIs across different social classes and between the maternal BMI categories because BMI increases with age (2, 3, 11, 22). Analysis of variance was used to compare mean BMIs across different categories of age at menarche. A paired analysis was performed for the difference between maternal BMI and the daughter's BMI, in which the age difference between mother and daughter was taken into account by a linear regression model. Linear regression analysis was used to investigate jointly the effect of the following predictor variables associated with BMI at 31 y: maternal BMI and age; social class as a categorical variable, with social class III as the reference class; birth weight; BMI at 1 and 14 y; and age at menarche for the women. Mean-centered values of the predictor variables were used, except for maternal BMI, which was centered to the mode of the class of normal weight (BMI of 22.50).
Analyses were based on only those subjects whose BMI data at birth and at 1, 14, and 31 y were available (n = 6280, 55% of those eligible). Data on the family's social class and the mother's age were available for all 6280 subjects, BMI was available for 5760 mothers, and the age at menarche was available for 3352 females born in 1966. Those subjects excluded because of missing data at some measurement point had an average BMI at birth and at 1 y that was not significantly different from those of the subjects who were included. Excluded subjects had a slightly higher mean BMI at 14 and 31 y. No significant difference between the BMIs of the mothers of included and excluded subjects was observed. The proportion of subjects from social class III was somewhat higher in the excluded subjects than in the included subjects (men: 31% compared with 23%; women: 30% compared with 24%). We used SPSS software for WINDOWS (version 9.0; SPSS Inc, Chicago).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. At 31 y, 40% of the men and 20% of the women were classified as overweight and 8% of the men and 9% of the women were classified as obese (Table 2). At 31 y, the proportion of men with abdominal obesity was 5% [waist-to-hip ratio (WHR) > 1.00 (1)] and that of women was 21% [WHR > 0.85 (1)]; 12% of men and 6% of women had a waist circumference 100 cm.
|
|
). These differences in BMI remained after adjustment for maternal BMI (data not shown) and maternal age. Correspondingly, the proportion of overweight (BMI: 25.0–29.9) and obese (BMI 30.0) 31-y-olds was lowest in men in social classes I and II (Table 2
). At 14 y, BMIs did not differ significantly between the social classes. In 31-y-old men, mean waist and hip circumferences and WHRs did not differ significantly between the social classes (Table 3). In women, waist circumference and WHRs were highest in those in the farming class and from social class IV.
|
).
BMI of offspring from birth to 31 y of age by maternal BMI
Children of overweight or obese mothers had higher mean BMIs at each age point than did children born to underweight or normal-weight mothers (Table 4). At 31 y, overweight and obesity were more common in subjects whose mothers were overweight or obese before pregnancy (men: 43% overweight and 12% obese; women: 27% overweight and 14% obese) than in subjects whose mothers were underweight or of normal weight (men: 39% overweight and 7% obese; women: 18% overweight and 7% obese; P < 0.001, chi-square test). Circumferences of the waist and hip and WHRs of the offspring at 31 y increased as maternal BMI increased.
|
 | (1) |
The point estimate of the intercept parameter (0.13) and its 95% CI (-0.07, 0.33) indicates that no significant differences in BMI were observed between mothers and their daughters when the age difference was taken into account.
To what extent did BMI in childhood predict obesity in adulthood?
BMI at 14 y was a more important predictor of adult BMI than was BMI at birth or 1 y (Tables 5 and 6). Of males with a low birth weight (<3000 g), 39% were overweight and 8% were obese at 31 y (Table 5). Of males with a high birth weight (4000 g), 38% were overweight and 10% were obese at 31 y. Of females with a low birth weight, 20% were overweight and 8% were obese at 31 y. Of females with a high birth weight, 23% were overweight and 10% were obese at 31 y.
|
|
). Of the males who were obese at 14 y, 41% were overweight and 47% were obese at 31 y. Of the females who were overweight at 14 y, 42% were still overweight and 22% were obese at 31 y. Of the females who were obese at 14 y, 27% were overweight and 55% were obese at 31 y.
Linear regression analysis indicated that the most important predictor of BMI at 31 y of age was BMI at 14 y (Table 6; model 3). When BMI at 14 y was not included in the regression analysis (models 2 and 3), the most important predictors were maternal BMI and age, the subject's BMI at 1 y, and age at menarche in women.
Age at menarche and BMIs of the female cohort members
The mean (±SD) age at menarche was 12.9 ± 1.3 y (n = 3352), with menarche occurring in 12% of the females by 11 y, in 27% at 12 y, in 32% at 13 y, in 17% at 14 y, and in 12% at 15 y. Age at menarche did not differ significantly between the social classes. BMI at 14 and 31 y increased linearly as the age at menarche decreased (Figure 1), but no significant differences in BMI at birth or at 1 y were observed. At 31 y, the proportions of overweight and obese females were significantly higher in those who had matured earlier (by 
11 y: 29% overweight and 15% obese; at 15 y: 16% overweight and 4% obese).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The sample of the original northern Finland birth cohort for 1966 is representative of the original study population, although when the data were restricted to individuals for whom BMI data were available for all measurement points, the proportion of subjects included from the lowest social class was smaller than the proportion excluded. The missing data introduced a small selection bias, which diluted the differences between the social classes because obesity was found to be more common in the lowest social class. The mean BMIs of the 31-y-old men and women in the 1966 birth cohort are similar to those previously observed in 30–35- and 30–39-y-old Finns in the 1990s (23, 24) and are slightly lower than those observed in 33-y-olds from a British birth cohort for 1958 (2). The paired analyses of differences in BMIs between the mothers and their 31-y-old daughters showed that BMI did not increase markedly during the 31-y period. This finding agrees with that of another Finnish study that showed weight gain to remain unchanged in Finnish women over a 25-y period (25). This lack of increase in BMI may have been because energy and fat intakes have decreased in Finland in recent decades (26). These dietary changes have probably enhanced the promotion of weight control in women, especially because women usually consume a healthier diet than do men (27). Furthermore, women are usually more concerned about their body weight than are men (28) and may feel more societal pressure to control their body weight than do men. Unfortunately, corresponding data for fathers have not been collected; thus, we cannot compare the male offspring with their fathers. The prevalence of obesity has increased by 
10–40% in most European countries over the past 10 y (1). This increase may be at least partly related to the aging of populations and also to changes in alcohol consumption and physical activity at work and during leisure time.
Our findings on the relation between BMI and social class add to the evidence emerging from other longitudinal studies (3, 10, 29). Although many cross-sectional studies show a consistent inverse relation between socioeconomic status and obesity for adult white women only (9, 30), longitudinal studies have consistently shown a relation in both sexes (3, 10, 11). For example, Braddon et al (3) reported differences in BMI by social class of origin in the British cohort of 1946 even at the age of 36 y, whereas Power and Moynihan (10) found such differences in the younger 1958 British cohort at 23 y of age. Data presented here for the Finnish 1966 cohort show a persisting influence of social class on BMI in both sexes during infancy and adulthood. We showed that this relation remained after adjustment for maternal BMI and age. To our knowledge this is a unique finding because most previous studies of socioeconomic status failed to adjust for maternal BMI (3, 10, 11). These results strengthen the observations that parental socioeconomic status predicts the development of obesity in adulthood in whites (3, 10, 11) and that differences in the prevalence of obesity between social classes stem at least partly from the socioeconomic circumstances of the family of origin.
It is of interest that the association between the social class at birth and adult BMI (at 31 y) was preceded by a similar trend at 1 y, despite the reverse trend observed at birth. Differences in BMI between social classes seem to be formed, at least partly, during early childhood and may affect health in the long term because unhealthy habits and ill health are more common among low social classes (31–33). The nutritional and health status of mothers from low social classes are likely to be less than optimal for fetal growth, and thus the mean BMI at birth is lower in babies from families of low than from high social classes. In this northern Finland birth cohort for 1966, Rantakallio (34) found that babies with a low birth weight were often from low social classes. Poor nutrition during the intrauterine period or other reasons leading to low birth weight, such as the mother's smoking (20), may be associated with persistent disturbances of the regulation of hormones related to growth. These irreversible disturbances in hormonal mechanisms may increase the biological vulnerability of children with low birth weight to the effects of unhealthy habits and may increase their risk of developing obesity in adulthood. However, the relation appears to be more equivocal during childhood and adolescence than during adulthood (10). In the Finnish cohort for 1966, no relation was evident at 14 y, as was similarly shown in the British longitudinal studies, at least in boys at ages 7, 11, and 16 y (29).
Mothers and their offspring from higher social classes had normal body weight more often than did those from lower social classes. During childhood and adolescence, the manner of living and concerns of suitable body image are learned in a societal context (9, 35). These features are typical for each social class and the risk of behavior damaging health is higher in lower socioeconomic groups (32). Negative attitudes toward obesity and negatively correlated perceived weight and body satisfaction are observed among girls and women but less so or not at all among boys and men (9). Societal pressure for thinness is higher among higher social classes. Low-birth-weight children from low social classes are thus predisposed by biological and environmental factors to the development of overweight and obesity.
Early maturation predicted overweight and obesity in adulthood in the present study, as in earlier studies (2, 36). However, no differences in BMI at birth or at 1 y were observed in relation to the age at menarche; thus, BMI during infancy was not associated with the age at menarche. The lack of an association between the age at menarche and BMI at birth and at 1 y suggests that factors occurring at a later stage of childhood may affect the age at menarche. Cooper et al (37) observed that girls who reached menarche at the youngest age had low birth weight but grew rapidly in childhood (during the first 7 y). The association between early maturation and adulthood obesity is probably multifactorial, partly determined by endocrine factors that promote the accumulation of body fat (38) and partly by behavioral factors. Late maturers have been described to be physically more active than early maturers (39), and the differences in energy intake and expenditure between early and late maturers during adolescence and adulthood are important. Early puberty may cause problems with identity; those undergoing puberty early may also psychologically mature earlier than girls of the same age without menarche. Among friends of the same age, the stress due to differences in appearance and in behavior between early and late maturers may lead early maturers to assume eating habits that predispose them to obesity in adulthood. This remains to be assessed.
Our study provided further evidence of maternal-offspring relations concerning body size (15, 16, 21). It also verified the increased risk of adult obesity among those who are fatter in childhood and adolescence (2–4, 11, 21, 40, 41). The mother's weight before pregnancy predicted the BMI of her offspring in all social groups. The heavier the mother, the heavier the offspring from birth to 31 y. This replicates the earlier findings of other longitudinal studies, which reported that children of obese parents are at increased risk of obesity throughout childhood and early adult life (15, 16, 42). With regard to child-adult continuities in BMI, BMI at 14 y predicted obesity in adulthood better than did birth weight or BMI at 1 y. This observation was similar to the earlier findings reviewed by Serdula et al (4), whereby the risk of becoming an obese adult was found to be greater among children who were obese at or around puberty. Thus, children who are obese near or at puberty form a special high-risk group that should be targeted for obesity prevention.
In conclusion, differences in BMI and in the prevalence of obesity in adulthood between social classes appear to be formed at least partly during early childhood. In addition to the low social class of the child's family, a high maternal BMI before pregnancy, the individual's own high BMI during adolescence, and early maturation (as measured by the age at menarche) are predictors of overweight and obesity in adulthood. Persons with the above-mentioned background factors form specific high-risk groups for obesity prevention.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. E. O'Neil and T. A. Nicklas A Review of the Relationship Between 100% Fruit Juice Consumption and Weight in Children and Adolescents American Journal of Lifestyle Medicine, July 1, 2008; 2(4): 315 - 354. [Abstract] [PDF]
|
|
|
|
 |
|
 A. Emaus, S. Espetvedt, M.B. Veierod, R. Ballard-Barbash, A.-S. Furberg, P.T. Ellison, G. Jasienska, A. Hjartaker, and I. Thune 17-{beta}-Estradiol in relation to age at menarche and adult obesity in premenopausal women Hum. Reprod., April 1, 2008; 23(4): 919 - 927. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Lawlor, G. D. Smith, M. O'Callaghan, R. Alati, A. A. Mamun, G. M. Williams, and J. M. Najman Epidemiologic Evidence for the Fetal Overnutrition Hypothesis: Findings from the Mater-University Study of Pregnancy and Its Outcomes Am. J. Epidemiol., February 15, 2007; 165(4): 418 - 424. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. O'Dea and R. Wilson Socio-cognitive and nutritional factors associated with body mass index in children and adolescents: possibilities for childhood obesity prevention Health Educ. Res., December 1, 2006; 21(6): 796 - 805. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M Kivimaki, G Davey Smith, M Juonala, J E Ferrie, L Keltikangas-Jarvinen, M Elovainio, L Pulkki-Raback, J Vahtera, M Leino, J S A Viikari, et al. Socioeconomic position in childhood and adult cardiovascular risk factors, vascular structure, and function: cardiovascular risk in young Finns study Heart, April 1, 2006; 92(4): 474 - 480. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Hinkula, A. Kauppila, S. Nayha, and E. Pukkala Cause-specific Mortality of Grand Multiparous Women in Finland Am. J. Epidemiol., February 15, 2006; 163(4): 367 - 373. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. B Pierce and D. A Leon Age at menarche and adult BMI in the Aberdeen Children of the 1950s Cohort Study Am. J. Clinical Nutrition, October 1, 2005; 82(4): 733 - 739. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E Stamatakis, P Primatesta, S Chinn, R Rona, and E Falascheti Overweight and obesity trends from 1974 to 2003 in English children: what is the role of socioeconomic factors? Arch. Dis. Child., October 1, 2005; 90(10): 999 - 1004. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Must, E. N. Naumova, S. M. Phillips, M. Blum, B. Dawson-Hughes, and W. M. Rand Childhood Overweight and Maturational Timing in the Development of Adult Overweight and Fatness: The Newton Girls Study and Its Follow-up Pediatrics, September 1, 2005; 116(3): 620 - 627. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Pouta, M.-R. Jarvelin, E. Hemminki, U. Sovio, and A.-L. Hartikainen Mothers and daughters: intergenerational patterns of reproduction Eur J Public Health, April 1, 2005; 15(2): 195 - 199. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M.-R. Jarvelin, U. Sovio, V. King, L. Lauren, B. Xu, M. I. McCarthy, A.-L. Hartikainen, J. Laitinen, P. Zitting, P. Rantakallio, et al. Early Life Factors and Blood Pressure at Age 31 Years in the 1966 Northern Finland Birth Cohort Hypertension, December 1, 2004; 44(6): 838 - 846. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Silventoinen, S. Sarlio-Lahteenkorva, M. Koskenvuo, E. Lahelma, and J. Kaprio Effect of environmental and genetic factors on education-associated disparities in weight and weight gain: a study of Finnish adult twins Am. J. Clinical Nutrition, October 1, 2004; 80(4): 815 - 822. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Krieger and G. Davey Smith "Bodies Count," and Body Counts: Social Epidemiology and Embodying Inequality Epidemiol. Rev., July 1, 2004; 26(1): 92 - 103. [Full Text] [PDF]
|
|
|
|
|
|
R. C. Whitaker Predicting Preschooler Obesity at Birth: The Role of Maternal Obesity in Early Pregnancy Pediatrics, July 1, 2004; 114(1): e29 - e36. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. J. Jones, L. Jahns, B. A. Laraia, and B. Haughton Lower Risk of Overweight in School-aged Food Insecure Girls Who Participate in Food Assistance: Results From the Panel Study of Income Dynamics Child Development Supplement Arch Pediatr Adolesc Med, August 1, 2003; 157(8): 780 - 784. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Langnase, M. Mast, S. Danielzik, C. Spethmann, and M. J. Muller Socioeconomic Gradients in Body Weight of German Children Reverse Direction between the Ages of 2 and 6 Years J. Nutr., March 1, 2003; 133(3): 789 - 796. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. E. Saelens, E. Jelalian, and D. M. Kukene Physician Weight Counseling for Adolescents Clinical Pediatrics, October 1, 2002; 41(8): 575 - 585. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
