Maternal Excess Weight and Nutritional Status of Preschool Children

Method: Using a cross-sectional study, the nutritional status of 224 children aged two-six from seven private schools in São Paulo city was classified into two categories: a low or appropriate body mass index (BMI) group called ‘normal weight’ and an overweight or obese group, called ‘excess weight’ according to the World Health Organization. The independent variables were: maternal and paternal nutritional status, sex of the child, family income, maternal education, duration of breastfeeding, age of mother and father and weight and length of the child at birth. A multiple generalised linear regression examined the relation between the independent variables and child excess weight.


Introduction
Over the years, obesity has mistakenly been associated with well-being and considered by many people to be synonymous with health, wealth and power.However, today it is considered a public health problem as its presence implies a decline in the quality and life expectancy of the individual [1].World Health Organization (WHO) data indicate a sharp increase in the prevalence of obesity in children in all regions of the world, including Brazil [2].
In the 1990s, researchers [3] were already warning about the increase of obesity, which has now become the principle nutritional problem among children.Gigante et al. (2003) [4] showed that the prevalence of overweight children was two times higher than the prevalence of short stature.More recent studies found a prevalence up to 35% of excess weight (EW) in children of preschool age [2,5].
According to the WHO (1990), obesity is the result of the interaction between diet, environmental factors and genetic predisposition.The environmental factors account for the majority of cases of obesity, while only a few cases are primarily attributable to genetic factors [6], as has been discussed on most recents studies on nutrigenomics [7,8].
Analysing the psychological, socioeconomic and cultural aspects, a preschool child is dependent on the environment in which he/she lives, which predominantly consists of his/her family.The child's behaviour is a reflection of this environment and as a consequence may have several features in common with the family, as they share the same socioeconomic and cultural conditions [9,10].
In relation to food choices, as one of the determinants of nutritional status, significant correlations have been found between food preferences of children and their families [10].Additionally if a mother did not like a food it would never be offered to her child, supporting the role of the family environment in determining food habits.
As already mentioned, food is one of the determinant factors of nutritional status of individuals, particularly with regard to EW, thus, the objective of this study was to evaluate the association between EW of parents and the nutritional status of their preschool aged children.

Method
This was a cross-sectional study with children aged two-six from seven private schools in the São Paulo city.A convenience sample consisting of 224 study units (trinomial: mother, father and preschool child) met the following inclusion criteria: aged two-six complete years, child not adopted, born in a single delivery without malformations and/or special needs or other health problems, whose mothers were not pregnant at the time of the evaluation.Of the 224 preschool children in the sample, 117 (52.2%) were male.
Data were collected with a standardised questionnaire, previously tested, delivered through the school, completed by the parents at home and returned to the researchers by the preschool children.
By this way, data collected included birth date of the father, mother and child; weight and height of the parents; weight and length of the child at birth; duration of breastfeeding; educational level of the mother and family income.
Subsequently, anthropometric measurements were collected (weight and height) by the researcher at the preschools of the children.A TANITA ® solar scale with a capacity of 150 kg and precision of 200 g was used to measure weight and a SECA ® stadiometer with a capacity of two metres to measure height.The anthropometric measurements followed the techniques proposed by Lohman et al [11].
To classify the child's nutritional status, Body Mass Index (BMI) was used, obtained from the ratio between weight (kg) and height (m²) [12], according to the cut-off points proposed by the WHO.Those classified as overweight or obese by the WHO [13,14] presented BMI z scores corresponding to ≥ +1 and <+2 and ≥ +2 respectively, were allocated to a group called EW.As a result, for statistical analysis, the dependent variable was operationalised in a dichotomous way, namely with or without EW.
The maternal and paternal nutritional status, independent variables, were obtained through the BMI and classified according to the WHO (1995) [15].For statistical analysis the variable was categorised as either 'no EW' (BMI ≤ 24.9 kg/m²) or 'with EW' (BMI> 24.9 kg/m²).
The other independent variables were maternal educational level (classified as: ≤ 8 years, ≥ 9 and <11 years, and ≥ 11 years), household income (classified as <5 times the minimum wage, ≥ 5 and <10 times the minimum wage, and ≥ 10 times the minimum wage) and duration of breastfeeding (classified as: <6 months; ≥ 6 to <12 months, and ≥ 12 months).
For the description of the sample, proportions and prevalence ratios (PR) were calculated for the categorical variables.A multiple generalised linear model (GLM) was used with the binomial family and logarithmic link, allowing us to directly obtain the PR to investigate the relationship between nutritional status of preschool children and the independent variables.
Variables with p≤0.20 were selected from the bivariate analysis to enter into the multivariate analysis.In the multiple regression analysis (GLM), the value of p≤0.10 was considered for the variable to remain in the final regression model.The 90% confidence interval was calculated for each variable associated with EW in children.Microsoft Office Excel 2003 software was used to prepare the database and Stata software, version 9.2, for the statistical analysis [16].
Those responsible for each preschool child signed a

Results
The preschool children sample showed a prevalence of 53.6% of EW, and of these 53.6% children, slightly less than half, 23.7%, were obese.No differences were observed regarding sex (p = 0.72).
Regarding the characteristics of the children, the birth weight average was 3122.6 (± 600.8) grams (Table 1).The median duration of breastfeeding was six months.The maximum duration of breastfeeding reported was 38 months and 3.2% of mothers reported not having breastfed their children.
In relation to the parents' BMI, for the mothers the average was 23.6 (± 3.2) kg/m 2 while the average for the fathers was 26.6 (± 3.7) kg/m 2 , the maximum values were, respectively, 33.6 kg/m 2 and 47.8 kg/m 2 (Table 1).
As seen in   maternal education, of every six mothers, five reported having studied for over 11 years.
In Table 3, the categorical characteristics associated with EW, or nearly reaching statistical significance in the bivariate analysis (p≤0.20), were maternal BMI, paternal BMI and duration of breastfeeding ≥ 12 months.Among the continuous variables, according to the same statistical criteria, only the birth weight and age of the father were statistically significant or close to significant (Table 4).
From these results, multiple regression analysis was carried out in which in the final model, considering the 90% confidence intervals, revealed the following characteristics were related to EW of preschoolers: maternal BMI and birth weight as risk factors and total duration of breastfeeding as a protective factor (Table 5).
The risk attributable to the EW of the child having a mother with EW was 31%, while birth weight, although statistically significant, was close to zero.The duration of breastfeeding for 12 months or more was a protective factor, showing a 29% reduction in overweight risk, when compared with the duration of up to six months of lactation.

Discussion
The findings of the present study demonstrated a high prevalence of EW in the children.More than half the preschool children, 53.6%, presented EW, with no difference between the sexes.Other recent Brazilian studies found a prevalence of up to 35% of EW in preschool children; some also found no differences according to the sex of the children [17,18].The high prevalence observed in the present study may be attributed to the fact that different indices and/or references were used for the classification of nutritional status or, more probably, to the fact that the study population belonged to a higher socioeconomic level.
Our study showed a positive association between maternal EW and above adequate nutritional status of the preschool children.This positive association has been shown in several locations over the past 15 year [19][20][21][22].More recently, two studies published in 2015, one from Mexico and another from Brazil also suggested that there is an important relationship between maternal overweight and the nutritional status of her preschool child [23,24].
A study conducted by Whitaker (2004) demonstrated that the risk of obesity was higher in children two-four years of age whose mothers presented obesity during pregnancy [25].
Evidence of a positive relationship between the nutritional status of Brazilian mothers and children, as well as overweight, were reported by Engstrom and Anjos (1996).Studying data from a 1989 national sample of children under ten years of age concluded that the risk of a child being overweight was 3.19 times higher when the mother was also overweight [19].
The interpretation of our results, even taking into consideration the possible limitations of this study, provides evidence of an association between maternal nutritional status, but not paternal, and EW in preschool aged children.
Contrary to our results, a study conducted in 2000 with 452 Brazilian schoolchildren, found a higher frequency of overweight and obesity in the fathers of children with EW [26].
A case-control study with three year old Japanese children, in 1989, also found an association between paternal and/or maternal obesity with obesity in their children [27].
This apparent discrepancy may be due to the methodological aspects of the present study, since the paternal nutritional status was estimated indirectly from information reported by the mothers and not from anthropometric measurements carried out by the researcher.It is worth pointing out, however, that the same did not occur regarding to the mothers' nutritional status, as it was also evaluated in the same indirect way.
Additionally, this study did not aim to evaluate the influence of both parents being overweight, but only the isolated influence of the nutritional status of each of parent, which may also have generated different results.
From these considerations, it is possible to assume that the association of EW between the mother and child is more intense and relevant than any isolated influence of the nutritional status of the father.
Overweight or obese mothers will probably have eating habits and practices that are not healthy.Taking into account that a little child's eating behaviour is consequence of the family [27] and only secondarily of other psychosocial and cultural interactions, it is very likely that the association between EW of the mother and of the child more probably occurs in the highest socioeconomic status families, as our results indicated.
In relation to birth weight and obesity/overweight in childhood, in this study, the greater the birth weight, the greater the prevalence of EW in preschool aged children, but with a very small relative risk.It was also noted, in the present study, that more than 12 months of breastfeeding was a protective factor against preschool EW.
Similar results regarding birth weight and duration of breastfeeding in association with EW in infancy have already been demonstrated in the literature [28][29][30][31].
These studies emphasise the relevance of our findings, whereby the importance of maternal overweight was a risk factor and/or marker of early development of EW among their children, regardless of birth weight and/or a longer period of breastfeeding.
As a novelty, the interpretation of the results of our study, even considering the possible limitations inherent to a convenience sample, allows us to highlight the assumption that even in higher income families, maternal nutritional status can be a good marker of the risk of children being overweight at preschool age, regardless of the coexistence of other risk factors.
This findings may also be of relevance to public health such that policies and programmes aimed at reducing the prevalence of overweight should incorporate maternal variables in relationship to the nutritional care of children.
Free and Clear Term of Consent.The Research Ethics Committee of the Faculty of Public Health of the University of São Paulo (n.1561/2006) approved the research project, according to the norms of Resolution 196/96 of the National Health Council in December 2006.

Table 2 ,
in relation to family income, just over half the families reported income of less than ten minimum wages.Concerning the level of

Table 1 .
Distribution of the study population, according to certain individual characteristics.São Paulo, SP, 2007.

Table 3 .
Distribution of the study population and results of the simple analysis (prevalence ratios and p values) according to the nutritional status of the preschool children and categorised individual characteristics.São Paulo, SP, 2007.

Table 2 .
Distribution of the study population, according to socioeconomic characteristics.São Paulo, SP, 2007.

Table 4 .
Distribution of the study population, according to means, standard deviations, minimum and maximum values of the characteristics analysed as continuous variables, as a function of the nutritional status of the preschool children.São Paulo, SP, 2007.
*Selected for GLS.

Table 5 .
The final multiple generalised linear regression model (GLM) with the characteristics associated with excess weight of preschool children.São Paulo, SP, 2007.