Associations of maternal employment and three-generation ... - Nature

International Journal of Obesity (2011) 35, 945–952 & 2011 Macmillan Publishers Limited All rights reserved 0307-0565/11 www.nature.com/ijo

PEDIATRIC ORIGINAL ARTICLE Associations of maternal employment and three-generation families with pre-school children’s overweight and obesity in Japan E Watanabe1, JS Lee1 and K Kawakubo2 1 Department of Health Promotion Science, School of Public Health, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan and 2Department of Food Science and Nutrition, Kyoritsu Women’s University, Tokyo, Japan

Backgrounds: Maternal employment has been shown to be associated with childhood overweight and obesity (Ow/Ob), but the presence of family members who care for children in place of the mothers might influence children’s Ow/Ob and lifestyles. The influence of maternal employment on children’s Ow/Ob should be examined together with the presence of caregivers such as grandparents. Objectives: The effects of maternal employment and the presence of grandparents on lifestyles and Ow/Ob in Japanese pre-school children were investigated. Design/Subjects: Cross-sectional study on 2114 children aged 3–6 years who attended all childcare facilities in a city and primary caregivers was conducted. Measurements: Children’s weight and height, family environments (family members, maternal employment, single parent, number of siblings and parental Ow/Ob) and lifestyles (dietary, physical activity and sleeping habits) were surveyed using a selfadministered questionnaire. Ow/Ob was defined by the International Obesity Task Force cut-offs. Results: The eligible participants were 1765 children. The prevalence of Ow/Ob was 8.4% in boys and 9.9% in girls. Maternal employment was associated positively with irregular mealtimes, unfixed snacking times, bedtime after 10 p.m. and nighttime sleep duration of less than 10 h, whereas three-generation families were associated negatively with irregular mealtimes after adjustment for children’s characteristics and family environments. Irregular mealtimes (OR (95% CI); 2.03 (1.36, 3.06)) and nighttime sleep duration of less than 10 h (1.96 (1.28, 3.01)) were associated with increased risks of being Ow/Ob. Both maternal employment and three-generation families were significantly associated with children’s Ow/Ob. However, threegeneration families maintained a significant association (1.59 (1.08, 2.35)) after adjustment for maternal employment. Conclusions: These study results suggest that the grandparents who care for pre-school children in place of mothers are more likely to contribute to childhood Ow/Ob than maternal employment. The family-focused lifestyle strategies to prevent childhood Ow/Ob must include grandparents who care for children. International Journal of Obesity (2011) 35, 945–952; doi:10.1038/ijo.2011.82; published online 26 April 2011 Keywords: childhood overweight; family environment; grandparents; lifestyle; maternal employment; three-generation family

Introduction The increasing prevalence of overweight and obesity in children is a major worldwide health concern.1–3 Prevalence of overweight including obesity (Ow/Ob) among Japanese children has more than doubled in the past three decades

Correspondence: E Watanabe, Department of Health Promotion Science, School of Public Health, Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan. E-mail: [email protected] Received 6 October 2010; revised 4 February 2011; accepted 7 March 2011; published online 26 April 2011

from 1970 to 2000, with the most recent estimates indicating that about 10% of children are Ow/Ob.4 Ow/Ob children are more likely to become Ow/Ob when they grow up,5–7 and are associated with several health problems such as coronary heart disease,8 metabolic syndrome9 and premature death.10 Therefore, early prevention of Ow/Ob is emerging as an important strategy to reduce associated short- and long-term morbidity. Daily lifestyle behaviors affect Ow/Ob. Several studies have shown that skipping breakfast,11 snacking at unfixed times,12 watching television (TV) for long duration13–17 and sleeping for short time duration18,19 are associated with an increased risk of Ow/Ob in childhood and adolescence.

Family environments and children’s overweight E Watanabe et al

946 Family environments influence children’s Ow/Ob. One study showed that single parents and the number of siblings influenced children’s physical activity and TV viewing time.20 Another study showed that caregivers’ short sleep duration was associated with children’s nighttime short sleep duration.19 Therefore, family environments as well as children’s daily lifestyles need to be considered when assessing risks of children’s Ow/Ob. Maternal employment is one of the family environments, which influences children’s lifestyles. A recent longitudinal study in UK showed that children with working mothers were more likely to be Ow/Ob than those of non-working mothers, and children’s likelihood of being Ow/Ob increased with the mother’s working time.21,22 The maternal employment might induce long hours of absence at home and affect children’s lifestyles such as dietary, physical activity and sleeping habits. Employment among women has increased, including mothers of pre-school children in Japan23 as well as in western countries.24 However, more than 20% of parents with children under the age of 6 lived with other family members such as grandparents in Japan,25 and mothers living with their grandparents had a higher employment rate.26 The presence of grandparents who care for children in place of the mothers might also influence children’s lifestyles and/or the incidence of Ow/Ob. However, there is only a limited study to examine the association between the presence of grandparents and children’s lifestyles. One qualitative study conducted in China showed that grandparents influence children’s eating behaviors,27 and another study conducted in Japan showed that living with grandparents influences children’s physical activity.28 Therefore, the influence of maternal employment and children’s lifestyles on Ow/Ob should be examined together with the presence of caregivers in place of mothers. The objectives of this study were to investigate the influence of maternal employment and the presence of grandparents on pre-school children’s Ow/Ob and their daily lifestyles in Japan. In Japan, most of the pre-school children aged 3 and older are attending childcare facilities such as nursery schools or kindergartens. Therefore, childcare facilities are a suitable target for investigating the influence of family environments and lifestyles on childhood Ow/Ob. The study was conducted on all pre-school children attending all childcare facilities in a city.

Subjects and methods Study population and study design The study population was 3–6-year-old children who attended all childcare facilities (24 nursery schools and 10 kindergartens) in a city in the Tohoku region and International Journal of Obesity

their principal caregivers. This cross-sectional study was performed in April 2003. Only those parents who agreed to participate in the study anonymously returned the questionnaire. We considered informed consent was obtained by receiving the filled-in questionnaire from the parents. Research ethical approval was received from the research ethics committee of Kyoritsu Women’s University (approval number 09011).

Measures A self-administered questionnaire was delivered by staff members of each facility to the principal caregivers of the children and was returned to each facility after completion of the questionnaire at home. The questionnaire included items about the children’s characteristics, children’s daily lifestyles and family environments. Children’s characteristics and anthropometric measurements. Children’s characteristics included age, sex and the birth weight. Children’s body weight and height were measured with standard methods at each facility in early April. Body weight was recorded in kilograms to one decimal place, and height was recorded to the nearest millimeter. The measurements were recorded in health handbooks, which were given to principal caregivers. The principal caregivers filled out the questionnaire after referring to the handbook. Body mass index (kg m2) was calculated and children were categorized as non-overweight, overweight or obesity, according to the International Obesity Task Force criteria for child body mass index.29 Overweight and obese children are combined and referred to as Ow/Ob. Children’s daily lifestyles. Children’s daily lifestyles consisted of dietary, physical activity and sleeping habits. Dietary habits included skipping breakfast, having meals at regular times and eating snacks at fixed times. Physical activity included time spent watching TV and time spent playing outside per day. Time spent watching TV involved estimating the total time spent watching TV and videos and playing electronic games on weekdays and weekends, respectively. Based on expert committee recommendations of TV viewing,31,32 time spent watching TV was dichotomized at o2 h per day or X2 h per day. Time spent playing outside was calculated by adding together the weekdays and weekends. Based on physical activity recommendations,33 time spent playing outside was dichotomized at o1 h per day or X1 h per day. Sleeping habits included wake time and bedtime on weekdays and weekends. Nighttime sleep duration per day was calculated as wake time minus bedtime, summing the total hours for the week, and dividing by 7 days. Based on sleep recommendations,34 sleep duration was dichotomized at o10 h per day or X10 h per day. Family environments. Family environments included type of family household, maternal employment status, parental


947 status, number of siblings and parental body weight status. Type of family household was dichotomized into twogeneration family or three-generation family. A two-generation family consists of only parents and their children. A three-generation family consists of children who live with parents and grandparents. Maternal employment status was made into a dichotomous variable with the type of employment: employed (full-time, part-time and self-employed) or unemployed (housewife), with regard to whether or not the mother had worked. Parental status was dichotomized to one parent or two parents. Number of siblings ranged from one to five, and was collapsed into three categories: one, two, three or more. Parental body weight status was determined based on self-reported body weight and height, and Ow/Ob were defined as body mass index of 25 kg m2 or more.30

regression models. The second model was built to adjust potential confounders to examine the association between maternal employment status or type of family household and children’s lifestyles, and the association between children’s lifestyles and children’s Ow/Ob. Further, to test the independent association between maternal employment status or type of family household and children’s Ow/Ob, the third model was built to mutually adjust the maternal employment status or type of family household. Adjusted odds ratios (OR) and 95% confidence intervals (CI) were used as a measure of effects and statistical significance. All analyses were conducted by using SAS (version 9.1, SAS Institute Inc., Cary, NC, USA). A P-value o0.05 was considered statistically significant and P-values were based on two-sided tests.

Statistical analysis Using logistic regression analyses, models were built in three stages. The first model, univariate logistic regression analyses to examine the significance and strength of associations of children’s characteristics and family environment variables with children’s Ow/Ob, were conducted. Children’s characteristics and family environment variables associated with children’s Ow/Ob at the Po0.1 level were considered as potential confounders and included in the further logistic

Results

Table 1

Participants At the time of the survey, 2114 children (aged 3–6 years) attended all childcare facilities in the city, and 1867 (88.3%) children returned a completed questionnaire. When their mother had missing work responses (n ¼ 91) or their father had not worked (n ¼ 11), children were excluded from the study analysis. Finally, 1765 (83.5%) children were eligible

Study participants’ characteristics and association with overweight and obesity in pre-school children

Children’s characteristics Age (years) 3-years old 4-years old 5-years old 6-years old Sex Boys Girls Birth weight (kg) Boys Girls Family environments Parental status Two parents One parent Number of siblings 1 (only child) 2 3 or more Parental Ow/Ob Both parents (BMI o25) Father (BMI X25) Mother (BMI X25) Both parents (BMI X25)

Mean±s.d.

Ow/Ob (%)

4.2±0.9

9.1 7.9 8.6 9.3 16.8

1.21

(1.00, 1.45)

(23.8) (34.1) (36.7) (5.4)

910 855 1748 899 849

(51.6) (48.4)

8.4 9.9

1 1.21 2.74

(0.88, 1.68) (1.80, 4.16)

1613 151

(91.4) (8.6)

8.8 12.6

1 1.49

(0.90, 2.49)

373 981 411

(21.1) (55.6) (23.3)

8.6 10.7 5.8

1 1.28 0.66

(0.84, 1.93) (0.38, 1.14)

1052 340 82 46

(69.2) (22.4) (5.4) (3.0)

7.3 9.4 17.1 17.4

1 1.32 2.61 2.67

(0.85, 2.03) (1.40, 4.85) (1.20, 5.92)

n

(%)

1765 420 602 648 95

Univariate OR (95% CI)

3.17±0.42 3.06±0.41

Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; Ow/Ob, overweight and obesity. Missing number of cases: birth weight (17), parental status (1) and parental Ow/Ob (245).

International Journal of Obesity


948 for the analyses. Over 98% of main respondents were children’s mothers. Study participants’ characteristics are shown in Table 1. The mean age of the children was 4.2-years old and 51.6% were boys. The prevalence of Ow/Ob was 8.4% in boys (6.1% Ow; 2.3% Ob) and 9.9% in girls (8.8% Ow; 1.1% Ob). Single parent family was 8.6%. The prevalence of one or both parents’ Ow/Ob was 26.5%. The variables associated with children’s Ow/Ob at a level of Po0.1 were children’s characteristics (age, sex and birth weight) and family environments (number of siblings and parental Ow/Ob) (Table 1), and were considered potential confounders. These children’s characteristics and family environment variables were included in the multiple logistic regression models for adjustment after here.

Association of children’s daily lifestyles with Ow/Ob Table 2 shows the association between children’s daily lifestyles and Ow/Ob. Children’s lifestyles that were significantly associated with Ow/Ob included irregular mealtimes (OR (95% CI); 1.75 (1.22, 2.50)), watching TV more than 2 h per day (1.54 (1.09, 2.19)) and nighttime sleep duration of less than 10 h (1.49 (1.05, 2.12)) in the univariate analysis. The significant association of irregular mealtimes and

Table 2

nighttime sleep duration of less than 10 h remained (2.03 (1.36, 3.06), 1.96 (1.28, 3.01), respectively) after adjustments for children’s characteristics, family environments, maternal employment and three-generation family. However, dietary habits of breakfast and snacking, physical activity habits including time for TV watching and outside playing and sleeping habits of bedtime were not significantly associated with children’s Ow/Ob.

Associations of maternal employment status and type of family household with children’s daily lifestyles Table 3 shows the association between maternal employment status or type of family household and children’s daily lifestyles. Maternal employment was significantly associated with irregular mealtimes, snacks at unfixed times, bedtime after 10 p.m. and less than 10-h sleep duration. These associations were observed not only in univariate analysis but also after adjustment for potential confounders and three-generation family. Three-generation family was significantly associated with regular mealtimes and snacks at unfixed times. The association between three-generation family and regular mealtimes remained after adjustment for potential confounders and maternal employment.

Associations of daily lifestyles with overweight and obesity in pre-school children n

Ow/Ob %

Dietary habits Breakfast Daily eating Skipping Having meals at regular times Regularly Irregularly Snacking pattern Fixed times Unfixed times Physical activity habits Time spent watching TV a day Less than 2 h More than 2 h Time spent playing outside a day More than 1 h Less than 1 h Sleeping habits Bedtime in the evening 10 p.m. or earlier 10 p.m. or later Nighttime sleep duration a day More than 10 h Less than 10 h

Adjusted ORa (95%CI)

Univariate OR (95%CI)

1588 174

9.2 8.6

1 0.93

(0.53, 1.63)

1 0.65

(0.31, 1.37)

964 555

7.0 11.7

1 1.75

(1.22, 2.50)

1 2.03

(1.36, 3.06)

809 908

8.6 9.7

1 1.13

(0.82, 1.58)

1 1.03

(0.70, 1.51)

721 907

7.3 10.9

1 1.54

(1.09, 2.19)

1 1.43

(0.97, 2.12)

1122 307

8.8 9.8

1 1.12

(0.73, 1.72)

1 1.06

(0.64, 1.74)

1316 434

8.9 9.9

1 1.13

(0.78, 1.63)

1 1.33

(0.88, 2.03)

667 1082

7.2 10.4

1 1.49

(1.05, 2.12)

1 1.96

(1.28, 3.01)

a

Abbreviations: CI, confidence interval; OR, odds ratio; Ow/Ob, overweight and obesity. Adjusted for children’s characteristics (age, sex and birth weight), family environments (number of siblings and parental Ow/Ob), maternal employment status and type of family household.



949 Table 3

Associations of maternal employment status and type of family household with daily lifestyles in pre-school children Maternal employment status

Type of family household

Employmenta

Three-generation familyb

Adjusted OR (95% CI)c

Univariate OR (95%CI) Dietary habit Breakfast Daily eating 1 Skipping 1.71 Having meals at regular times Regularly 1 Irregularly 1.41 Snacking pattern Fixed times 1 Unfixed times 2.07

Adjusted OR (95% CI)c

Univariate OR (95%CI)

(1.13, 2.58)

1 1.47

(0.94, 2.31)

1 1.12

(0.82, 1.53)

1 0.99

(0.69, 1.43)

(1.10, 1.81)

1 1.56

(1.19, 2.05)

1 0.78

(0.63, 0.96)

1 0.74

(0.59, 0.94)

(1.65, 2.59)

1 1.93

(1.51, 2.47)

1 1.41

(1.16, 1.70)

1 1.24

(0.99, 1.53)

Physical activity habits Time spent watching TV a day Less than 2 h 1 More than 2 h 0.90 Time spent playing outside a day More than 1 h 1 Less than 1 h 1.18

(0.72, 1.13)

1 0.90

(0.70, 1.16)

1 0.95

(0.78, 1.16)

1 0.93

(0.75, 1.16)

(0.88, 1.58)

1 1.25

(0.91, 1.73)

1 0.89

(0.69, 1.14)

1 0.93

(0.70, 1.23)

Sleeping habits Bedtime in the evening 10 p.m. or earlier 1 10 p.m. or later 2.38 Nighttime sleep duration a day More than 10 h 1 Less than 10 h 3.34

(1.77, 3.20)

1 2.50

(1.81, 3.45)

1 0.86

(0.70, 1.07)

1 0.78

(0.61, 1.01)

(2.66, 4.18)

1 3.50

(2.73, 4.48)

1 1.19

(0.98, 1.45)

1 0.97

(0.78, 1.22)

a

b

c

Abbreviations: CI, confidence interval; OR, odds ratio. Reference category is ‘housewife’. Reference category is ‘two-generation family’. Adjusted for children’s characteristics (age, sex and birth weight), family environments (number of siblings and parental Ow/Ob) and maternal employment status or type of family household.

Table 4

Associations of maternal employment status and type of family household with overweight and obesity in pre-school children

Maternal employment status Housewife Employment Type of family household Two-generation family Three-generation family

Univariate OR (95% CI)

Adjusted OR (95% CI)a

Adjusted OR (95% CI)a,b

n

Ow/Ob %

433 1332

6.2 10.1

1 1.68

(1.10, 2.58)

1 1.60

(1.01, 2.53)

1 1.42

(0.89, 2.28)

847 918

7.2 10.9

1 1.58

(1.13, 2.20)

1 1.69

(1.16, 2.48)

1 1.59

(1.08, 2.35)

a

Abbreviations: CI, confidence interval; OR, odds ratio; Ow/Ob, overweight and obesity. Adjusted for children’s characteristics (age, sex and birth weight) and family environments (number of siblings and parental Ow/Ob). bAdjusted for maternal employment status or type of family household.

Associations of maternal employment status and type of family household with children’s Ow/Ob The association of maternal employment status and type of family household with children’s Ow/Ob is shown in Table 4. Among the participants, 75.5% of the children’s mothers were employed and 52.0% of the children lived in threegeneration families. Both maternal employment and threegeneration family were significantly associated with childhood Ow/Ob in the univariate and adjusted analysis. However, in a mutually adjusted further analysis, maternal employment was not significantly associated with childhood

Ow/Ob after adjustment for three-generation family, whereas three-generation family maintained a significant association after adjustment for maternal employment.

Discussion The study purpose was to clarify the influence of maternal employment and three-generation family on pre-school children’s Ow/Ob. The study results revealed that both International Journal of Obesity


950 maternal employment and a three-generation family were significantly associated with children’s Ow/Ob. However, a three-generation family, that is grandparents who care for children in place of mothers, had around 1.6 times higher risk to be Ow/Ob independent of maternal employment status and other potential confounding factors, such as children’s characteristics and family environments. A three-generation family was significantly associated with a lower risk of irregular mealtimes, whereas maternal employment was associated with a higher risk of irregular mealtimes and snacking at unfixed times. In a UK study, rapid weight gain occurred in children cared for by grandparents than in children cared for by their parents.35 A qualitative study in Chinese three-generation families has reported that grandparents used food as an educational and emotional tool, which shaped the behavior of their grandchildren.27 Our current study did not measure quantitative and/or qualitative contents of children’s meals and snacks. It might mean that grandparents regularly offer unlimited quantity of meals and/or sugar-sweetened snacks to their grandchildren, and this might be one of the reasons for the increased risk of Ow/Ob in children cared for by their grandparents in Japan, too. In the present study, maternal employment was not significantly associated with childhood Ow/Ob after adjustment for three-generation family. Children are largely cared for by mothers in Japan, and employed mothers tend to be more reliant on grandparental care than housewives. Among the study participants, 58.0% of employed mothers were living in three-generation families as compared with 33.7% of fulltime housewives, regardless of single or two parents. A single parent consisted of only mothers’ families in our study participants (99.3%). Thus, maternal employment and a three-generation family could interact with the children’s Ow/Ob. However, there are no studies concerning the influence of maternal employment and three-generation family on children’s Ow/Ob simultaneously. This reason will be one of strengths of this study. Furthermore, a single parent has appeared to be one of the family environments related to children’s Ow/Ob.21,36 However, there was no significant association in the current study. Single parents could be more reliant on grandparents than two parents, because of living in proximity. In current study participants, single parents had higher maternal employment (91.4%) and living in three-generation families (68.2%) as compared with two parents (74.0 and 50.5%, respectively). Future studies are needed to examine whether to receive grandparents’ support, regardless of living with or without them. In the study of UK children, Hawkins et al.21 reported that children, regardless of attending childcare facilities or not, were more likely to be Ow/Ob at age 3 years if their mother held any employment since childbirth in a longitudinal study. Our study children ranged in ages 3–6, and the study design was cross-sectional. The prevalence of employed mothers typically rises with children’s age, but mothers International Journal of Obesity

who were classified into employment in our study might have included mothers who had begun to work as well as those who continued to work. In the adjusted model, children’s age was included as confounding factors, but these differences of classification might attenuate the association between maternal employment and children’s Ow/Ob in our study. Further study that includes duration of maternal employment will be needed. Among the children’s lifestyles, sleep duration for less than 10 h at night and irregular mealtimes were significantly associated with children’s Ow/Ob after adjustment for children’s characteristics, family environments, maternal employment and three-generation family. The association of short nighttime sleep duration with Ow/Ob was consistent with other studies in children,18,19,37 although definitions of short sleep duration varied based on the age of the children being studied. A recent systematic review also suggested that short sleep duration might be an independent risk factor for weight gain and Ow/Ob, particularly in younger populations.38 Efforts to secure sufficient sleeping time at night as part of the family environments to prevent children’s Ow/Ob will be required. Our finding suggests that irregular mealtimes were associated with childhood Ow/Ob. Few studies have examined the influence of irregular mealtimes on Ow/Ob, although some studies have addressed this as one potential factor.21,35 Farshchi et al.36 have examined the effects of regular meal frequency on energy metabolism in healthy women. The results showed that regular eating had a beneficial effect on dietary thermogenesis. Skipping breakfast was not significantly associated with childhood Ow/Ob in the present study, although a few studies found significant association.11 Irregular mealtimes include skipping meals, thus regular eating habits could be an important element for the prevention of childhood Ow/Ob with controlled quantity and quality of meals. Both indices of physical activity, watching TV more than 2 h and daily time spent playing outside, were not associated with childhood Ow/Ob after an adjusted analysis in the present study. Many other studies showed consistent positive association between longer TV watching time and childhood Ow/Ob.13,16,18 Other recent findings of Steele et al.39 reported time spent in vigorous-intensity physical activity appeared to be more strongly and negatively associated with adiposity than with sedentary time spent with electronic games, computer, Internet and TV use in UK children. Our study participants were the children who attended childcare facilities. Childcare facilities in Japan provide extended care if necessary for the children with working mothers. Time spent at home is different for each child, thus time spent watching TV at home or time spent playing outside which represent moderate intensity of physical activity in our study may not fully reflect physical activity levels. Future studies of the association between physical activity levels and childhood Ow/Ob in pre-school children are needed.


951 Finally, for the examination of the independent influence of maternal employment or three-generation family on children’s Ow/Ob apart from the children’s daily lifestyles, which is associated with Ow/Ob, we added lifestyle variables to the adjusted model. The lifestyle variables were regular mealtimes and nighttime sleep duration. The results showed neither three-generation family nor maternal employment were significant (1.42 (0.93, 2.16), 1.10 (0.66, 1.84), respectively) (data not shown). It could be interpreted that the effects of the presence of grandparents on childhood Ow/Ob was mediated through children’s lifestyle behaviors. Children spent much more time with family members such as their grandparents in three-generation families than with their two-generation family. Therefore, family-focused education, involving not only parents but also grandparents who care for children, is a useful framework for developing a healthy lifestyle behavior in young children. The present study had several limitations. Firstly, social economic status, such as parents’ educational level and/or household economical level, might affect the children’s Ow/Ob, but our study could not include these kinds of parameters. The mean of household income of the study area was almost the same with the average household income of the Japan (the monthly income per household in the year 2000 was 596.4 thousand yen, and the national average was 566.6 thousand yen:40 approximately 59 640 and 56 660 Euro in the year 2000, respectively), and the range of household income was relatively smaller than other developed countries.41 Secondly, measurements were based on the parents’ reports. In particular, parents’ body weight and height might be under-reported, although children’s weight and height were measured at each childcare facility on the eve of the study. Thirdly, this study did not include the contents of meals and intensity of physical activity. Thus, these might cause underestimation of our study results than real associations. Further studies concerning detailed and objective measurements of not only children’s lifestyles but also family environments are essential. However, the current study surveyed all children attending all childcare facilities in the study area and covered 93.3% of the children living in that area, and yielded a relatively high response rate (83.5%). Among the study participants, more than half (52%) of the study participants were threegeneration families, and more than 75% of mothers were working at the time of the survey. Thus, our study could cover a wide range of family environmental characteristics with a high proportion of three-generation families in Japan. In conclusion, the study results suggest that the grandparents who care for pre-school children in place of mothers are more likely to contribute to childhood Ow/Ob than maternal employment. Irregular mealtimes and short sleep duration at night were contributing factors of children’s Ow/Ob. These findings indicate that family-focused behavioral strategies to prevent childhood Ow/Ob must include grandparents who care for children, because lifestyle is established early in life.

Conflict of interest The authors declare no conflict of interest.

Acknowledgements This study was supported by a grant from the Japan Ministry of Education, Culture, Sports, Science and Technology (ID 20240063). The authors express their appreciation to the participants for their cooperation, as well as to the staff members of childcare facilities and the city officers in Tsuruoka city.

References 1 Ebbeling CB, Pawlak DB, Ludwig DS. Childhood obesity: publichealth crisis, common sense cure. Lancet 2002; 360: 473–482. 2 Haslam DW, James WP. Obesity. Lancet 2005; 366: 1197–1209. 3 Han JC, Lawlor DA, Kimm SY. Childhood obesity. Lancet 2010; 375: 1737–1748. 4 Ministry of Education, Culture, Sports, Science and Technology. School Health Examination Survey. National Printing Bureau: Tokyo, Japan, 2009. 5 Whitaker RC, Wright JA, Pepe MS, Seidel KD, Dietz WH. Predicting obesity in young adulthood from childhood and parental obesity. N Engl J Med 1997; 337: 869–873. 6 Baird J, Fisher D, Lucas P, Kleijnen J, Roberts H, Law C. Being big or growing fast: systematic review of size and growth in infancy and later obesity. BMJ 2005; 331: 929. 7 Singh AS, Mulder C, Twisk JW, van Mechelen W, Chinapaw MJ. Tracking of childhood overweight into adulthood: a systematic review of the literature. Obes Rev 2008; 9: 474–488. 8 Baker JL, Olsen LW, Sorensen TI. Childhood body-mass index and the risk of coronary heart disease in adulthood. N Engl J Med 2007; 357: 2329–2337. 9 Sun SS, Liang R, Huang TT, Daniels SR, Arslanian S, Liu K et al. Childhood obesity predicts adult metabolic syndrome: the Fels Longitudinal Study. J Pediatr 2008; 152: 191–200. 10 Franks PW, Hanson RL, Knowler WC, Sievers ML, Bennett PH, Looker HC. Childhood obesity, other cardiovascular risk factors, and premature death. N Engl J Med 2010; 362: 485–493. 11 Utter J, Scragg R, Mhurchu CN, Schaaf D. At-home breakfast consumption among New Zealand children: associations with body mass index and related nutrition behaviors. J Am Diet Assoc 2007; 107: 570–576. 12 Ishihara T, Takeda Y, Mizutani T, Okamoto M, Koga M, Tamura U et al. Relationships between infant lifestyle and adolescent obesity: the Enzan maternal-and-child health longitudinal study]. Jpn J Public Health 2003; 50: 106–117. 13 Saelens BE, Sallis JF, Nader PR, Broyles SL, Berry CC, Taras HL. Home environmental influences on children’s television watching from early to middle childhood. J Dev Behav Pediatr 2002; 23: 127–132. 14 Sugimori H, Yoshida K, Izuno T, Miyakawa M, Suka M, Sekine M et al. Analysis of factors that influence body mass index from ages 3 to 6 years: A study based on the Toyama cohort study. Pediatr Int 2004; 46: 302–310. 15 Marshall SJ, Biddle SJ, Gorely T, Cameron N, Murdey I. Relationships between media use, body fatness and physical activity in children and youth: a meta-analysis. Int J Obes Relat Metab Disord 2004; 28: 1238–1246. 16 Hancox RJ, Poulton R. Watching television is associated with childhood obesity: but is it clinically important? Int J Obes 2006; 30: 171–175.



952 17 Jouret B, Ahluwalia N, Cristini C, Dupuy M, Negre-Pages L, Grandjean H et al. Factors associated with overweight in preschool-age children in southwestern France. Am J Clin Nutr 2007; 85: 1643–1649. 18 Sekine M, Yamagami T, Handa K, Saito T, Nanri S, Kawaminami K et al. A dose-response relationship between short sleeping hours and childhood obesity: results of the Toyama Birth Cohort Study. Child Care Health Dev 2002; 28: 163–170. 19 Jiang F, Zhu S, Yan C, Jin X, Bandla H, Shen X. Sleep and obesity in preschool children. J Pediatr 2009; 154: 814–818. 20 Bagley S, Salmon J, Crawford D. Family structure and children’s television viewing and physical activity. Med Sci Sports Exerc 2006; 38: 910–918. 21 Hawkins SS, Cole TJ, Law C. Maternal employment and early childhood overweight: findings from the UK Millennium Cohort Study. Int J Obes 2008; 32: 30–38. 22 Hawkins SS, Cole TJ, Law C. An ecological systems approach to examining risk factors for early childhood overweight: findings from the UK Millennium Cohort Study. J Epidemiol Community Health 2009; 63: 147–155. 23 Statistics Bureau, Ministry of Internal Affairs and Communications. The 2007 Employment Status Survey. Japan Statistical Association: Tokyo, JPN, 2008. 24 OECD. Society at a Glance: OECD Social Indicators. OECD: Paris, 2005, pp 40–41, doi:10.1787/soc_glance-2005-en, http://www.oecd-ilibrary. org/docserver/download/fulltext/8105031e.pdf?expires=1302261513 &id=0000&accname=guest&checksum=A23043105AA560FE027ADA 78677EEAC6. 25 Cabinet office, Government of Japan. White paper on BirthrateDeclining Society. Gyosei: Tokyo, Japan, 2004. 26 Statistics and Information Department, Minister’s Secretariat, Ministry of Health and Welfare. The 7th Longitudinal Survey of Babies in 21st Century. Japan Statistical Association: Tokyo, Japan, 2010. 27 Jiang J, Rosenqvist U, Wang H, Greiner T, Lian G, Sarkadi A. Influence of grandparents on eating behaviors of young children in Chinese three-generation families. Appetite 2007; 48: 377–383. 28 Sekine M, Kanayama H, Kagamimori S. Associations of social and family characteristics with physical inactivity in a Japanese birth cohort. Research-aid report 2007; 22: 62–69.


29 Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000; 320: 1240–1243. 30 WHO expert consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 2004; 363: 157–163. 31 American Academy of Pediatrics, Committee on Public Education. Children, adolescents, and television. Pediatrics 2001; 107: 423–426. 32 Committee on media use, Japan Pediatric Association. Policy statement–Media use. J Jpn Pediatr Assoc 2004; 27: 7–10. 33 Salmon J, Shilton T. Endorsement of physical activity recommendations for children and youth in Australia. J Sci Med Sport 2004; 7: 405–406. 34 Committee for School Health, The Japanese Society of Child Health. Proposal for health and sleep in children. J Child Health 2001; 60: 817–819. 35 Griffiths LJ, Hawkins SS, Cole TJ, Dezateux C. Risk factors for rapid weight gain in preschool children: findings from a UK-wide prospective study. Int J Obes 2010; 34: 624–632. 36 Farshchi HR, Taylor MA, Macdonald IA. Beneficial metabolic effects of regular meal frequency on dietary thermogenesis, insulin sensitivity, and fasting lipid profiles in healthy obese women. Am J Clin Nutr 2005; 81: 16–24. 37 Reilly JJ, Armstrong J, Dorosty AR, Emmett PM, Ness A, Rogers I et al. Early life risk factors for obesity in childhood: cohort study. BMJ 2005; 330: 1357. 38 Patel SR, Hu FB. Short sleep duration and weight gain: a systematic review. Obesity 2008; 16: 643–653. 39 Steele RM, van Sluijs EM, Cassidy A, Griffin SJ, Ekelund U. Targeting sedentary time or moderate- and vigorous-intensity activity: independent relations with adiposity in a populationbased sample of 10-y-old British children. Am J Clin Nutr 2009; 90: 1185–1192. 40 Statistics Bureau Management and Coordination Agency Government of Japan. Social indicators by prefecture of Japan. Japan Statistical Association: Tokyo, Japan, 2010. 41 OECD. Income distribution and poverty in OECD countries. In the Second Half of the 1990s., 2005, http://www.oecd.org/ dataoecd/48/9/34483698.pdf (accessed 29 January 2011).