Nutritional Status Trends in Brazilian Preschoolers: A Cohort Study

Abstract

Background:

Determining trends in the nutritional status of children may guide care prevention beyond this age in an effort to reduce the prevalence and incidence of overweight and/or obese children. The objective of this study is to evaluate the evolution of the nutritional status of preschool children in two moments, with an interval of 2 years.

Methods:

This is a cohort study of a random probabilistic sample of preschool children attending public schools within an urban area of high human development index city, in the hinterland of São Paulo state. In 2016, we reassessed the nutritional status of 351 preschoolers evaluated in 2014, comparing the prevalence of overweight according to BMI >1 z-score.

Results:

The prevalence of overweight was 31.05% (2014) and 31.06% (2016) and mean BMI z-scores were 0.58 and 0.57, respectively. The nutritional status classification of the preschool children showed almost no agreement between the two time points (κ = 0.053). Nevertheless, children with overweight in 2014 had a relative risk of 1.96 of being overweight or obese in 2016 (p = 0.0473).

Conclusions:

Prevalence of overweight among preschoolers was the same at 2 and 5 years of age, with no tendency to grow. Despite this, 2-year-old preschoolers with overweight present a twofold higher relative risk for excessive weight at 5 years of age. These changes of nutritional status at preschool age evince the great flexibility of their nutritional condition during this period of life.

Introduction

The high prevalence of children with overweight is currently considered a major public health problem and poses an urgent and serious challenge.¹ Overweight or obesity in childhood has immediate implications for physical and mental health and both are major risk factors for the development of cardiovascular diseases, diabetes, and death, including in adults.²

Anthropometric evaluation is recommended by the World Health Organization³ (WHO) and the Ministry of Health⁴ as an important method for assessing nutritional status because it is highly sensitive, low cost, and easy to apply. In childhood, a stage characterized by rapid development growth monitoring allows early detection of health problems and situations of nutritional risk,⁵ such as overweight and obesity.

Obesity is affecting the general population increasingly earlier, and it is, therefore, important to monitor rates and risks in the pediatric population. There are few studies in Brazil prospectively investigating overweight and obesity trends in preschool children and the components of anthropometric evaluation related to this outcome.

Thus, determining trends in the nutritional status of children of excess and adequate weight, as well as analyzing changes at two-time points during preschool age, can help guide care prevention beyond this age, in an effort to reduce the prevalence and incidence of children with overweight and/or obese.

Methods

Study Sample

This is a cohort study of a probabilistic randomized sample of infants and preschool children attending municipal day care centers in Taubaté (state of São Paulo, Brazil). The selection of the sample was carried out by clusters, in a probabilistic and random manner, having as sample unit the public day care centers. The first evaluation was conducted in 2014.⁶

The initial study included children from public day care centers in Taubaté (São Paulo, Brazil) born in 2011/2012.

All children attending day care centers were evaluated before inclusion in the sample to be carriers of diseases and/or receiving treatment that could influence their growth or confuse the nutritional assessment by anthropometry, to avoid false diagnosis and, consequently, undue conclusions from the results. About two and a half years after, in 2016, we reevaluated the nutritional status of those children who at 5 years of age were still attending municipal day care centers in the same town.

In 2016, we visited the 40 existing municipal day care centers to locate children evaluated in 2014 who were still in the municipal network to recover their anthropometry data. From the initial sample, the loss was of 55 preschoolers: 4 were not authorized to take part by parents, 5 were not found for the anthropometric reevaluation, and for 55 children, the informed consent form was not returned by parents. Therefore, the final sample of 2016 comprised 364 children.

Out of this total, 13 children were excluded a posteriori because some of their data were considered as an outlier: 10 children who had a BMI ≥4 z-scores above the group mean; 3 with a height difference (between 2014 and 2016) well below the average. This gave a final sample of 351 children, for whom complete and consistent data were available to carry out the comparative analysis.

The study project was approved by the Research Ethics Committee of the Public Health School of the University of São Paulo (COEP-FSP-USP) under permit no. 1815019.

Parameters for Anthropometric Evaluation

The anthropometric measurements were taken at preschool centers by professionals previously trained to carry out the research. Weight readings were taken using portable digital scales (Wiso^®) with a capacity of 200 kg accurate to the nearest 50 g. Height was determined using a wall-mounted portable stadiometer (WCS^®), measured in centimeters and millimeters. Measurements were taken according to the recommendations of the WHO,⁷ in triplicate, where the value used for analysis was the average of the three measurements obtained from each preschooler.

For the characterization of nutritional status, z-score of BMI (zBMI) according to the WHO (2006) was used, classified based on the cutoff points proposed by the Ministry of Health³ and the WHO.⁸ For the statistical analysis, the term overweight was used for children with a zBMI > +1, which included the group of children at risk of overweight (zBMI between +1 and +2), overweight (zBMI between +2 and +3), and obesity (zBMI > +3).

Statistical Analysis

The keying in of the variables and their consistency were checked by dispersion plots and frequency tables using the software MedCalc^® version 18.6.4.0. Data items whose values were highly discrepant were reviewed by checking the original datasheets.

Continuous variables were expressed as mean, minimum and maximum values, standard deviation (SD), and 95% confidence interval (95% CI), and by distribution frequency after categorization according to zBMI. The κ coefficient was calculated to determine concordance between nutritional status in 2014 and 2016.⁹ The association for categorical variables was assessed by Fisher's exact test. Relative risk was estimated with the respective 95% CI. For all analyses, the significance level was defined as p = 0.05.

All calculations were performed using SPSS (Statistical Package for the Social Sciences) version 21.0 and MedCalc version 18.4.

Results

Of the 351 children included in the survey, 165 (47%) were boys and 186 (53%) were girls. Mean age in 2014 was 2.4 years with SD of 0.3 years and that in 2016 was 4.8 years with SD of 0.3 years. The time elapsed between the two evaluations was a mean of 2.4 years, with SD of 0.1 year.

The values for weight and height measured, as well as the BMI calculated, for each of the two evaluations are given in Table 1, showing that the mean BMI was very similar on both evaluations.

Table 1.

Age, Weight, Height, and BMI of Preschoolers in 2014 and 2016

Variables	2014					2016
Variables	Mean	SD	Minimum	Maximum	95% CI	Mean	SD	Minimum	Maximum	95% CI
Age (years)	2.4	0.3	1.7	3.1	2.4–2.6	4.8	0.3	4.1	5.4	4.7–4.8
Weight (kg)	13.5	1.9	8.9	23.0	13.3–13.7	19.4	3.3	12.9	37.5	19.0–19.7
Height (cm)	90.3	4.3	77.1	104.5	89.8–90.7	108.9	4.9	93.3	128.8	108.4–109.4
BMI (kg/m²)	16.5	1.5	13.2	24.3	16.4–16.7	16.2	1.9	12.6	26.9	16.0–16.4

CI, confidence interval; SD, standard deviation.

The difference of the coefficients of variation (CV) of weight in 2014 (13.9%) and 2016 (17.2%) was statistically significant (p < 0.0001), as well as the BMI (2014: CV = 8.8%; 2016: CV = 11.7%, p < 0.0001), which means that the variability of the parameters increased intragroup in the studied period. In the same analysis for height values, the difference of the CV was not statistically significant (2014: CV = 4.8%, 2016: CV = 4.5%, p = 0.2288).

Approximately 70% of the children were of normal weight in 2014, where this proportion remained practically unchanged 2 years later. The proportion of children with overweight (zBMI > +1) was also the same for the 2 years, 2014 (31.05%) and 2016 (31.06%). There was a higher frequency of children with risk of overweight in the first evaluation, but the proportion of children with overweight or obesity was higher in 2016 (Table 2).

Table 2.

Distribution of Nutritional Status of the Preschool Children According to z-Score of BMI for Age in 2014 and 2016

zBMI	2014	2016
zBMI	n (%)	n (%)
< −3	0 (0.00)	0 (0.00)
≥ −3 and ≤ −2	0 (0.00)	2 (0.57)
≥ −2 and ≤ +1	242 (68.95)	240 (68.38)
> +1 and ≤ +2	87 (24.79)	77 (21.94)
> +2 and ≤ +3	17 (4.84)	22 (6.27)
> +3	5 (1.42)	10 (2.85)

Classification: < −3, very underweight; ≥ −3 and ≤ −2, underweight; ≥ −2 and ≤ +1, normal weight; > +1 and ≤ +2, risk of overweight; > +2 and ≤ +3, overweight; > +3, obesity.

zBMI, z-score of BMI.

Regarding nutritional status over time, 48.7% of the children were classified as being of normal weight on both evaluations. The proportion of children initially eutrophic and who were later classified as risk of overweight, overweight, or obese (20.20%), and the proportion initially overweight who became eutrophic in the second evaluation (20.00%) proved similar (Table 3).

Table 3.

Change in Nutritional Status Classification According to BMI on First and Second Evaluations

First evaluation (2014)	Second evaluation (2016)
First evaluation (2014)	Total (%)	Normal weight (%)	Risk of overweight (%)	Overweight and obesity (%)
Normal weight	242 (68.9)	171 (48.7)	54 (15.4)	17 (4.8)
Risk of overweight	87 (24.8)	56 (16.0)	20 (5.7)	11 (3.2)
Overweight and obesity	22 (6.3)	14 (4.0)	4 (1.1)	4 (1.1)
Total	351 (100.0)	241 (68.7)	78 (22.2)	32 (9.1)

κ = 0.053; CI = −0.031 to 0.137; p = 0.216.

Bold values indicate values with no change in nutritional status classification on first and second evaluations.

The κ value of 0.053 confirms that nutritional status of the preschool children showed poor agreement⁸ between the two time points (2014 and 2016), indicating that the change in the classification of the nutritional status observed in the two evaluations, although within the statistical limit, did not occur by mere chance.

Separate analysis of the group of children with overweight in 2014 and their relationship with overweight or obesity in 2016 (Table 4) revealed a statistically significant association between being overweight in 2014 and being overweight or obese in 2016 (p = 0.0473). Overweight status in 2014 showed a relative risk of 1.96 (95% CI 1.02–3.78) of being overweight or obese in 2016.

Table 4.

Relative Risk of Change in Nutritional Status According to Presence/Absence of Overweight in 2014 and Presence/Absence of Overweight and Obesity in 2016 among Preschool Children

Overweight (2014)	Overweight and obesity (2016)			RR	95% CI
Overweight (2014)	Yes (%)	No (%)	Total (%)	RR	95% CI
Yes (%)	15 (4.3)	94 (26.8)	109 (31.1)	1.96	1.02–3.78
No (%)	17 (4.8)	225(64.1)	242 (68.9)
Total	32 (9.1)	319 (90.9)	351 (100.0)

Fisher's exact test p = 0.0473.

Discussion

Most of the preschool children studied were of normal weight and about one-third were overweight in 2014 and 2016, almost double the expected rates according to the WHO, whose data indicate that 15.9% of normal children may show a zBMI > +1³.

The proportion of preschoolers with overweight did not increase during the 2-year follow-up, although previous cross-sectional studies in the city of Taubaté showed a progressive increase in the prevalence of this nutritional status among preschool children.^6,10

Recent studies show evidence that the tendency for increased prevalence of overweight in developed countries,^11,12 and also in Brazil,^13,14 observed a few years ago, has stabilized, mostly among preschool children.¹⁵

Despite this positive evidence, there is still a substantial proportion of children with overweight who will likely suffer the consequences of this nutritional status in the short, medium, and long term.¹⁶ In Brazil, there is still a high prevalence of children with overweight,¹⁷ also among younger groups,¹⁸ and that has been observed even in children residing in rural regions.^19,20

Among children living in rural regions of Brazil, there are still undernutrition conditions, but on the other hand, the incidence of overweight is increasing.^19,20 Probably, in a country with the characteristics similar to Brazil, in rural areas the prevalence of overweight and obesity should be lower than in urban regions. The possibility of generalization of these results is, therefore, limited to children from urban areas attending public schools, since it is not possible to know whether this trend can be extrapolated to the rural area currently. However, this study may be considered, along with other findings, to support future perspectives for children in rural areas.

Although there was no agreement between the nutritional status of the children in 2014 and 2016, separate analysis of the group who were overweight at 2.5 years of age showed a relative risk of almost twice as high of becoming overweight or obese by age 5 compared with those who were of normal weight at the first assessment.

These results corroborate findings of a systematic review of the literature, concluding that nutritional status above normal in childhood was associated with a higher risk of subsequent obesity both in childhood/adolescence and in adulthood.²¹ Paradoxically, and positively, the “perfect agreement” between children with overweight in 2014 and with overweight in 2016 seems to indicate responsiveness to timely nutritional interventions in this age group. Factors potentially explaining the intragroup shift of children include flexibility in individual nutritional status of these children, as well as the presence of potentially obesogenic social determinants outside the school environment, which may have acted to keep the high rates of overweight and obesity stable throughout the study.

Previous cohort studies investigating BMI in preschool children concluded that change in nutritional status toward overweight can be identified at 3.5 years of age.^22,23 Analyses of BMI trajectory showed a peak in BMI at between 7 and 9 months of age^24,25 and found a positive association between the magnitude of this peak and zBMI later.²⁶

These data, taken together with the results of this study, emphasize that interventions addressing the different factors promoting excess weight should be implemented preventively before children reach excess weight. Further studies aimed at identifying possible risk markers for the development of overweight in childhood are needed to both facilitate screening of these children and to inform policies and guidelines for intervention in the course of excess weight and its consequences.

The main strength of this investigation was that it constituted a cohort study with evaluations defined by protocol carried out by the same researcher, using the same techniques, instruments, and criteria, thereby guaranteeing quality and reliability of the data. In addition, the study involved a representative sample of preschool children from the Municipal Child Education Centers, with coverage of ∼70% of the preschool population of the city of Taubaté. This fact may influence the generalization of the conclusions and should be taken into account in future applications or comparisons of the study results.

There is clearly a high prevalence of overweight among children of public preschools, but this pattern appears to be stabilizing. In parallel, although there was no concordance in the classification of individual nutritional status between 2.5 and 5 years of age, overweight children aged 2–3 years had a twofold higher risk of being overweight or obese by the age of 5 years.

Therefore, it is evident that the monitoring of the anthropometric growth parameters of preschool children should also be performed regularly after the age of 2 years to both assess the adequacy of growth and enable early identification of those children at nutritional risk. These measures may allow early intervention to prevent overweight, a condition that, once established, tends to progress to overweight or obesity in subsequent years.

Footnotes

Acknowledgment

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior–Brasil (CAPES)–Finance Code 001” is thanked for financing, in part, this study.

Author Disclosure Statement

No competing financial interests exist.

References

World Health Organization. Report of the commission on ending childhood obesity. Geneva: WHO, 2016.

World Health Organization. Guideline: Assessing and managing children at primary health-care facilities to prevent overweight and obesity in the context of the double burden of malnutrition. Updates for the Integrated Management of Childhood Illness (IMCI). Geneva, Switzerland, 2017.

World Health Organization. Child growth standards based on length/height, weight and age. Acta Paediatr, 2006; 450:S76–S85.

Ministério da Saúde. Orientações para a coleta e análise de dados antropométricos em serviços de saúde: Norma Técnica do Sistema de Vigilância Alimentar e Nutricional–SISVAN. Brasília, 2011.

Sperandio

, Sant'ana

RFL

, Franceschini

SCC

, Priore

. Comparação do estado nutricional infantil com utilização de diferentes curvas de crescimento. Rev Nutr, 2011; 24:566–574.

Lopes

, Rocha

EMB

, Silva

JPC

, et al. Aleitamento materno, introdução da alimentação complementar e excesso de peso em pré-escolares. Arch Latinoam Nutr, 2016; 66:195–200.

World Health Organization. Physical status: The use and interpretation of anthropometry. Geneva: WHO, 1995.

Onis

, Lobstein

. Defining obesity risk status in the general childhood population: Which cut-offs should we use?. Int J Peadiatr Obes, 2010; 5:458–460.

Landis

, Koch

. The measurement of observer agreement for categorical data. Biometrics, 1977; 33:159–174.

10.

Nascimento

, Silva

JPC

, Ferreira

, et al. Aleitamento materno, introdução precoce de leite não materno e excesso de peso na idade pré-escolar. Rev Paul Pediatr, 2016; 34:454–459.

11.

Olds

, Maher

, Zumin

, et al. Evidence that the prevalence of childhood overweight is plateauing: Data from nine countries. Int J Pediatr Obes, 2011; 6:342–360.

12.

NCD Risk Factor Collaboration. Worldwide trends in body-mass index, underweight, overweight, and obesity from 1975 to 2016: A pooled analysis of 2416 population-based measurement studies in 128 · 9 million children, adolescents, and adults. Lancet, 2017; 390:2627–2642.

13.

Menezes

CFB

, Neves

, Gonsalez

, Vasconcelos

FAG

. Incidence of overweight/obesity in preschool children during a two-year follow-up. Rev Nutr, 2014; 27:269–278.

14.

Conde

, Monteiro

. Nutrition transition and double burden of undernutrition and excess of weight in Brazil. Am J Clin Nutr, 2014; 100:1617S–1622S.

15.

Wabitsch

, Moss

, Kromeyer-Hauschild

. Unexpected plateauing of childhood obesity rates in developed countries. BMC Med, 2014; 12:17.

16.

Sabin

, Kao

, Jounala

, et al. Viewpoint article: Childhood obesity–looking back over 50 years to begin to loog foward. J Paediatr Child Health, 2015; 51:82–86.

17.

Sousa

CPC

, de Olinda

, Pedraza

. Prevalence of stunting and overweight/obesity among Brazilian children according to different epidemiological scenarios: Systematic review and meta-analysis. Sao Paulo Med J, 2016; 134:251–262.

18.

Dos Anjos

, Da Silveira

WDB

. Estado nutricional dos alunos da Rede Nacional de Ensino de Educação Infantil e Fundamental do Serviço Social do Comércio (Sesc), Brasil, 2012. Ciência & Saúde Coletiva, 2017; 22:1725–1734.

19.

Lopes

, Frota

MTBA

, Leone

, Szarfarc

. Nutrition profile of children in Maranhão state. Rev Bras Epidemiol, 2019; 22:E190008.

20.

Paciorek

, Stevens

, Finucane

, Ezzati

. Children's height and weight in rural and urban populations in low-income and middle-income countries: A systematic analysis of population-representative data. Lancet, 2013; 1:e300–e309.

21.

Baird

, Fisher

, Lucas

, et al. Being big or growing fast: Systematic review of size and growth in infancy and later obesity. BMJ, 2005; 331:929.

22.

Pryor

, Tremblay

, Boivin

, et al. Developmental trajectories of body mass index in early childhood and their risk factors. Arch Pediatr Adolesc Med, 2011; 165:906–912.

23.

McCormick

, Dickinson

, Haemer

, et al. What can providers learn from childhood body mass index trajectories: A study of a large, safety-net clinical population. Acad Pediatr, 2014; 14:639–645.

24.

Sun

, Nwaru

, Hua

, et al. Infant BMI peak as a predictor of overweight and obesity at age 2 years in a Chinese community based cohort. BMJ Open, 2017; 7:e015122.

25.

Roy

, Chesi

, Xiao

FMR

, et al. Body mass index (BMI) trajectories in infancy differ by population ancestry and may presage disparities in early childhood obesity. J Clin Endocrinol Metab, 2015; 100:1551–1560.

26.

Silverwood

, De Stavola

, Cole

, et al. BMI peak in infancy as a predictor for later BMI in the Uppsala family study. Int J Obes, 2009; 33:929–937.