Prevalence and secular trend of childhood overweight and obesity in a Mediterranean area of Southeast Spain

ABSTRACT

Introduction

Childhood obesity epidemic is a worldwide public health problem, but recent studies show a stabilization trend.

Objective

To study the prevalence of overweight and obesity in school-age children from a Mediterranean area in Southeast Spain from 1992 to 2011.

Methods

Cross-sectional study at two-time points (1992 and 2011) of representative samples of children aged 6-11 years (n = 737 and 620, respectively). Weight and height were measured by trained personnel. Overweight and obesity were defined according to Body Mass Index (BMI) using the International Obesity Task Force (IOTF) and the World Health Organization (WHO) criteria.

Results

The prevalence of overnutrition (overweight plus obesity) decreased significantly from 1992 to 2011, with a decrease of 45.4% to 36.0% (according to the WHO cut-offs) or 37.3% to 30.0% (according to the IOTF cut-offs). Obesity decreased significantly according to WHO criteria (19.6% to 13.5%) while overweight remained stable. When grouping by sex, overnutrition in girls remained stable with a prevalence of 39.0% (WHO) or 34.0% (IOTF) in 1992, and 35.7% (WHO) or 31.1% (IOTF) in 2011. However, in boys a significant decrease is observed, with a prevalence of 50.0% (WHO) or 38.8% (IOTF) in 1992, decreasing to 38.8% (WHO) and 29.5% (IOTF) in 2011.

Conclusions

Even though the prevalence of childhood overweight and obesity remained at high levels, according to our results in the studied period there was a decrease in the Mediterranean area of Southeast Spain. There was variability between sex, with a significant decrease only in males. Prevalence estimates varied depending on the reference values used.

KEYWORDS:

• Childhood
• obesity
• overweight
• trend
• Spain
• Mediterranean

Introduction

Childhood obesity has become a major concern in public health in the 21^st century. According to the latest data of the World Health Organization (WHO), in 2016 around 41 million children throughout the world had obesity or overweight (World Health Organization 2019). Childhood obesity increases the risk of suffering type 2 diabetes, cardiovascular diseases, musculoskeletal disorders, some types of cancer, and psychosocial disorders (World Health Organization 2019). Moreover, children who are obese prior to puberty are more likely to be obese as adults than prepubescent children with normal weight (Cole and Lobstein 2012; Sabin et al. 2015).

It is difficult to determine the exact etiology of childhood obesity, since there are environmental as well as genetic factors involved (Sabin et al. 2015). Even during early development, there may be factors influencing the appearance of childhood obesity, such as pre-pregnancy body mass index (BMI), gestational weight gain, maternal lifestyle, and child environment during the first years of life (Larqué et al. 2019). Socioeconomic status (SES), lifestyle and diet during childhood are also key factors (Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016). In the last decade several public health programs, like “Estrategia NAOS” (Ministerio de Sanidad y Consumo 2005) or “Niñ@s en movimiento” (Sandra et al. 2007), have been implemented in Spain to tackle childhood obesity.

In the last four decades, there has been a global increase in childhood overweight and obesity. From 1975 to 2016, the prevalence of obesity in children and adolescents aged 5–19 years has risen from 0.8% to 6.7%, and the prevalence of overweight from 4.3% to 18.4% (NCD-RisC 2017). This increase is mainly due to the increase of obesity in developing countries, especially Asia, while in developed countries we start to detect stabilization. In Europe, childhood obesity and overweight have suffered a slight increase from 20.6% in 1999–2006 to 21.3% in 2011–2016 (Garrido-Miguel et al. 2019). Although the Iberian Peninsula has one of the highest prevalence in Europe, there is a decrease in these values from 30.3% in 1999–2006 to 25.9% in 2011–2016 (Garrido-Miguel et al. 2019). In Spain, according to the latest data of the ALADINO study (2015), there is also a stabilization of the overweight and obesity values, with a prevalence of overweight of 23.2% and obesity of 18.1% (Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016). The last National Health Survey (NHS) in 2017 showed that Southeast Spain was one of the areas with the highest prevalence of obesity (13.1%) and overweight (21.3%) (Ministerio de Sanidad, Consumo y Bienestar Social 2017). The results of the 2017-NHS (Ministerio de Sanidad, Consumo y Bienestar Social 2017) showed a north-south-southeast gradient, with a higher prevalence in the southern and Mediterranean Autonomous Communities, which had already been observed in other studies (Valdés-Pizarro and Royo-Bordonada 2012) like the enKid study (Serra Majem et al. 2003). This may be due to differences in SES, lifestyle, diet or weather between the different areas of Spain (Valdés-Pizarro and Royo-Bordonada 2012). Southeast Spain is located in the Mediterranean coast, including regions like Murcia and Alicante. Although there are some studies carried out in these regions (Ruíz Pérez et al. 2008; Espín et al. 2013), there are no studies combining the data of the child population of the whole area of Southeast Spain.

The aim of this work was to determine the prevalence of overweight and obesity in school-age children in the Southeast Mediterranean area of Spain, and to study its variation within twenty years according to the IOTF and WHO cut-off points.

Material and methods

Study design and participants

An observational study was carried out in children from 6 to 11 years old in the Mediterranean area of Southeast Spain. Two cross-sectional cut points were done in 1992 and 2011. At each cut-point, with the support and assistance of the Ministry of Education and Science, lists of schools were obtained from regional school authorities. A cluster sample design was used and four schools were selected randomly in each cut-point. All children from 6 to 11 years of the selected schools were invited to enroll in the study. Children with congenital or acquired diseases affecting the nutritional state were excluded. Their parents or legal tutors received an information sheet and signed the informed consent.

The minimum sample size calculated was 388 subjects at each cut-point, based on prevalence rates from previous similar studies (Pérez-Farinós et al. 2013), assuming a power of 80%, an alpha value of 5% and 10% of losses. A total of 1357 children participated in the study, 737 in 1992 and 620 in 2011.

Measurements of both weight and height were done by trained observers using a clinical scale (Mechanical scale SECA® 700) with a precision of 100gr to measure the weight and a stadiometer (Stadiometer SECA® 220) with a precision of 1 mm for the size. Measurements were taken during school hours. Children were measured without shoes and wearing only underclothes. Instruments were always calibrated before measurement.

Definitions of overweight and obesity

BMI was used to define childhood overweight and obesity, which is the measure most used in epidemiologic studies (de Onis and Lobstein 2010; Ogden and Flegal 2010). To compare the variables, z-scores (ZS) were calculated by using the LMS method proposed by Cole (Cole and Lobstein 2012).

Overweight and obesity were defined according to the IOTF (Cole and Lobstein 2012) and to the WHO (de Onis et al. 2007) reference values. WHO defines overweight in children over 5 years old with BMI values between +1SD and +2SD, and obesity with BMI values over +2SD. According to the IOTF, overweight and obesity correspond to a BMI greater than 25 and 30 at the age of 18 years, respectively. The term “overnutrition” is used to include individuals with both overweight and obesity.

Statistical analysis

The statistical package used was SPSS 22.0. A descriptive analysis was carried out, presenting the results of the qualitative variables with absolute frequencies and percentages with a confidence interval of 95%. To describe the quantitative variables, the standard mean and deviation was used for those with normal distribution and the interquartile mean and range for those with non-normal distribution. The χ² test was done to compare qualitative variables, the Student’s t-test to compare quantitative parametric variables and Mann-Whitney U-test to compare quantitative non-parametric variables. It was considered statistical significance when the probability (p) was below 0.05.

Results

The sample consists of 737 subjects in the first period of study and 620 in the second. BMI z-score was significantly lower in Group 2011 compared to Group 1992 (p < 0.001) (Table 1) A significant decrease of the overnutrition percentage was observed, from 45.4% to 36.0% according to WHO cut-offs (p < 0.001) and from 37.0% to 30.3% according to those of the IOTF (p = 0.009) (Table 2). If overweight and obesity were studied separately, there was a significant decrease only in the obesity group according to WHO references, with a decrease from 19.6% to 13.5% (p = 0.003).

Table 1. Description and analysis of the variables studied among 6–11 years old schoolchildren in Southeast Spain from 1992 to 2011

	Group 1992 (n = 737)	Group 2011 (n = 620)	p-value^a
Age (years) Median (IQR)	9.02 (2.97)	8.91 (2.94)	0.096
Weight (kg) Median (IQR)	31.65 (13.68)	32.15 (13.68)	0.948
Height (cm) Median (IQR)	132.90 (16.87)	136.65 (18.10)	<0.001
BMI (kg/m^2) Median (IQR)	17.78 (4.48)	17.16 (4.20)	<0.001
Z-score BMI^b Mean (SD)	0.90 (1.22)	0.57 (1.27)	<0.001
Sex % women(n)	44.8 (330)	49.2 (305)	0.061

BMI: Body Mass Index. IQR: Interquartile range. SD: standard deviation.

^aMann-Whitney U test for continuous non-parametric variables. T-student test for continuous parametric variables. Chi-squared for categorical variables.

^bCalculated by LMS method.

Table 2. Trends in the prevalence of BMI categories from 1992 to 2011 among 6–11 years old schoolchildren in Southeast Spain by sex using WHO and IOTF definitions

	Group 1992 (n = 737)		Group 2011 (n = 620)
	% (CI 95%)	n	% (CI 95%)	n	p-value
Overnutrition
WHO
Girls	39.7(34.6–45.1)	131	35.7 (29.6–40.0)	109	0.304
Boys	50.0 (45.3–54.9)	203	36.2 (32.0–42.6)	114	<0.001
Total	45.4 (41.8–48.9)	334	36.0 (32.3–39.8)	223	<0.001
IOTF
Girls	34.8 (30.0–40.1)	115	31.1 (25.4–35.4)	95	0.322
Boys	38.8 (34.2–43.6)	158	29.5 (24.8–34.8)	93	0.009
Total	37.0 (33.6–40.6)	273	30.3 (26.8–34.1)	188	0.009
Overweight
WHO
Girls	26.4 (21.9–31.4)	87	25.9 (20.6–30.1)	79	0.895
Boys	25.4 (21.3–29.8)	103	19.0 (15.1–23.8)	60	0.044
Total	25.8 (22.8–29.1)	190	22.4 (19.3–25.9)	139	0.146
IOTF
Girls	26.4 (21.9–31.4)	87	25.8 (20.6–30.1)	77	0.748
Boys	27.3 (23.2–31.8)	111	20.0 (16.0–24.8)	63	0.023
Total	26.9 (23.8–30.2)	198	22.6 (19.5–26.0)	140	0.069
Obesity
WHO
Girls	13.3 (10.1–17.4)	44	9.8 (6.8–13.3)	30	0.170
Boys	24.6 (20.6–29.0)	100	17.1 (13.4–21.7)	54	0.015
Total	19.6 (16.8–22.6)	144	13.5 (11.1–16.5)	84	0.003
IOTF
Girls	8.5 (5.9–12.0)	28	5.9 (3.6–8.9)	18	0.210
Boys	11.5 (8.8–15.0)	47	9.5 (6.8–13.6)	30	0.382
Total	10.2 (8.2–12.6)	75	7.7 (5.9–10.1)	48	0.120

CI 95%: Confidence Interval of 95%, Overnutrition: overweight plus obesity, WHO: World Health Organization, IOTF: International Obesity Task Force

Sex differences were detected. Boys presented higher overnutrition prevalence than girls according to WHO criteria (p = 0,021), but not according to IOTF criteria (p = 0,511). If we studied by years, we only observed a significant sex difference in 1992 according to WHO criteria (p = 0,005). Only boys presented a significant decrease in overnutrition prevalence, from 50% to 37.2% according to WHO and from 38.8% to 29.5% according to IOTF. There is also a significant decrease of obesity prevalence in boys when using the WHO cut-off points, from 24.6% to 17.1% (p = 0.015). Nevertheless, girls did not show a significant change in obesity and overweight prevalences.

Discussion

Even though overweight and obesity rates are still high in the Mediterranean area of Southeast Spain, our results showed that there has also been a decrease of childhood overnutrition prevalence in the study period, in agreement with results from some other European countries that participated in the Childhood Obesity Surveillance Initiative (COSI) (Wijnhoven et al. 2014; World Health Organization 2018a, 2018b) or national results from ALADINO study (Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016).

The age range chosen for our sample belongs to a crucial period of child development since it is the previous phase of the important changes that take place in puberty and may predict the condition in adult age. Furthermore, at 6 years old, the “adipose rebound” takes place, being the second period of a fast increase in body fat. This age range is really interesting because it includes children in primary education and most health programs are aimed to this group (World Health Organization 2018a; Robinson et al. 2019).

The variable used to study obesity and overweight was the BMI, the most frequent and consistently registered childhood growth measure and also the most chosen in epidemiological studies which allow comparing with other international works (de Onis and Lobstein 2010; Ogden and Flegal 2010). Nevertheless, we must take into account the limits of this measurement due to the heterogeneity in body composition. In clinical practice, other measurements such as fat adiposity or abdominal perimeter should be observed to value obesity (Espín et al. 2013; Robinson et al. 2019). One of the strong points in our study is using the LMS method proposed by Cole (Cole and Lobstein 2012) to calculate the z-score of the BMI.

The IOTF and the WHO cut-off points were used because they are based on data from different countries and are considered the choice values to study childhood obesity in most epidemiological studies, which allows comparing with other national and international works (de Onis and Lobstein 2010; Ogden and Flegal 2010; Rolland-Cachera 2011; Espín et al. 2013; Wijnhoven et al. 2014). However, overweight and obesity prevalence varies depending on the reference values used. Overweight and obesity rates are higher when we use the WHO cut-offs compared to those of the IOTF, which could lead to an overestimation of overnutrition also observed in other studies (Espín et al. 2013; Kêkê et al. 2015; Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016; Valerio et al. 2017; Lauria et al. 2019). In one hand, some studies suggest that IOTF cut-offs are more appropriated for population studies and should be used by international researchers and policy markers (Kêkê et al. 2015; Gandhari and Taghi 2016; Valerio et al. 2017; Lauria et al. 2019). On the other hand, the WHO charts are more suitable for clinical use in monitoring children’s growth (Gandhari and Taghi 2016). The IOTF references shall be preferable for the identification of overweight and obesity both at individual and population levels because they are associated with ill health later in life (Monasta et al. 2011). Nevertheless, there is yet no consensus in which reference values should be used to define obesity and overweight in childhood (Rolland-Cachera 2011; Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016). The latest recommendation of the European Childhood Obesity Group (ECOG) is to use both IOTF and WHO cut-offs (Rolland-Cachera 2011).

According to our results, from 1992 to 2011 there was a significant decrease in childhood overnutrition using both IOTF and WHO cut-off points. However, according to the NCD Risk Factor Collaboration (NCD-RisC), the obesity prevalence from 5–19 years all around the world has increased from 1992 to 2011 (9,8% vs 21,4%) and this upward trend continued until 2016 (NCD-RisC 2017). Despite some international and national works support this increase (Ruíz Pérez et al. 2008; Salcedo et al. 2010; Ng et al. 2014), other authors noted a stabilization (Ogden et al. 2016; Ramiro-González et al. 2017; Hardy et al. 2017; De Ruiter et al. 2017; Aranceta-Bartrina and Pérez-Rodrigo 2018) and even in some cases a decrease (Schmidt Morgen et al. 2013; Skinner and Skelton 2014; Koebnick et al. 2015; Lazzeri et al. 2015; Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016; Miqueleiz et al. 2016; Bygdell et al. 2017; Sánchez-Cruz et al. 2018; Herter-Aeberli et al. 2019). In Australia (Hardy et al. 2017), USA (Skinner and Skelton 2014; Ogden et al. 2016) and some European countries (Schmidt Morgen et al. 2013; Lazzeri et al. 2015; Herter-Aeberli et al. 2019) such a stabilization and downward trend have been reported. In Spain, the figures of the NHS showed a stabilization (35,7% vs 34,5%) from 2001 to 2011 (Miqueleiz et al. 2016), and the Aladino study (Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016) reported actually a decrease (44,5% vs 41,3%), especially in overweight and with barely any changes in the obesity group. In some Spanish regional studies, a decrease was observed in Andalusia (Sánchez-Cruz et al. 2018) and Catalonia (de Bont et al. 2020) like in our area, but in some other Spanish regions like Cuenca (Martínez-Vizcaíno et al. 2012) this downward change was not observed. There are studies carried out in some areas in Southeast Spain, like Murcia or Alicante. The study carried out in Murcia (Espín et al. 2013) showed stabilization of obesity and overweight from 2005 to 2011 according to IOTF and WHO definitions. However, the study carried out in Alicante (Ruíz Pérez et al. 2008) reported an increase in childhood obesity and overweight from 1993 to 2008, but we should compare our results with theirs carefully because they did not use international references like WHO or IOTF to define BMI categories. In our study population, there was a significant decrease in obesity according to the WHO cut-off points, although not according to those of the IOTF. However, overweight remained stable. Several studies agree with our results in a tendency to stabilization in different grades of childhood overnutrition, also showing some of them a decrease in the obesity group (Schmidt Morgen et al. 2013). However, other studies showed a decrease in the overweight group, while the percentage of children with obesity remains stable (Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016; Miqueleiz et al. 2016) or even increases (Pérez-Farinós et al. 2013; Wijnhoven et al. 2014).

When studying the prevalence of overweight and obesity by sex, we observed a higher number of boys with overnutrition than girls according to WHO criteria, which is also observed in other studies (Wijnhoven et al. 2014; Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016; NCD-RisC 2017), Ng et al. (2014), Miqueleiz et al. 2016)^. This could be due, among other reasons, to a higher percentage of muscle mass in boys or a different perception of weight, which usually concerns girls more (Salcedo et al. 2010; Ramiro-González et al. 2017). However, other works showed higher percentages in the group of girls (Pérez-Farinós et al. 2013; Garrido-Miguel et al. 2019). While there was a significant decrease in overweight and obesity in our boys, there was no such a decrease in girls who showed a stabilized profile. This significant decrease observed only in the male population has been reported in other studies (Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016; Miqueleiz et al. 2016; Bygdell et al. 2017). These findings could be explained by the different growth patterns depending on the gender and the hormonal differences (Ministerio de Sanidad y Consumo 2005). Higher overnutrition prevalence in the male group (Pérez-Farinós et al. 2013; Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016), as well as the higher percentage of boys practicing sports than girls, who have a higher percentage of sedentarism (Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016), could also explain these results.

It is important to highlight that, even if we observe a decrease in the childhood obesity prevalence in our area, rates are very high if we compare with the prevalence worldwide or in other developed countries (NCD-RisC 2017; Garrido-Miguel et al. 2019; Serra Majem et al. 2003; Espín et al. 2013). Southeast Spain is a special area of the Mediterranean coast, that has one of the highest prevalence of childhood obesity and overweight, behind Andalusia, Canary Islands and Ceuta-Melilla (Serra Majem et al. 2003; Valdés-Pizarro and Royo-Bordonada 2012). According to the latest numbers of the NHS , obesity and overweight prevalence was about 35%, higher than the national prevalence that is around 29% (Ministerio de Sanidad, Consumo y Bienestar Social 2017). These high numbers of obesity and overweight may be due to the particular characteristics of this area. The results of the latest NHS 2017 showed a north-south gradient, with higher prevalence in south Autonomous Communities, that had already been observed in the enKid study (Serra Majem et al. 2003; Valdés-Pizarro and Royo-Bordonada 2012; Ministerio de Sanidad, Consumo y Bienestar Social 2017). This gradient was also observed in other international studies (Garrido-Miguel et al. 2019). This is probably due to the socio-economic and lifestyle differences between north and south regions. According to the enKid study (Serra Majem et al. 2003) and Aladino study (Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016), childhood obesity is more prevalent in families with lower SES and/or lower level of education. The latest results of the National Statistics Institute of Spain (NSI) (Instituto Nacional de Estadística 2018) showed a north-south gradient in SES, where south regions have a lower SES than north regions. According to the latest numbers of NSI, Southeast Spain has a higher AROPE (At Risk of Poverty and/or Exclusion) rate than the national rate (EAPN-ES 2019).

Lifestyle, which includes diet and activity exercise, has proved to be a great influence on childhood obesity. A diet based on high frequencies of consumption of energy-dense micronutrient-poor products and low frequencies of consumption of fruit and vegetable, is related to childhood obesity (Pérez-Farinós et al. 2013). High screen time, short sleep duration and low activity exercise were associated with increased BMI (Serra Majem et al. 2003; Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016). Although Southeast Spain is a region in the Mediterranean area, the rate of fruit and vegetable ingested is not too high and it is similar to the country rate (Serra Majem et al. 2003; Ministerio de Sanidad, Consumo y Bienestar Social 2017), even though a study carried out in Alicante showed that this rate was lower compared with the results in the enKID study (Ruíz Pérez et al. 2008). There is also increased consumption of energy-dense micronutrient-poor products. Attending to the results of the study carried out in Alicante, a huge percentage of children in our area do moderate exercise compares with the results of enKID study (Ruíz Pérez et al. 2008). Nevertheless, Southeast Spain is one of the areas with the highest percentage of a sedentary lifestyle in free time (Ministerio de Sanidad, Consumo y Bienestar Social 2017). This area also has an elevated rate of screen time, but similar to national rates (Ministerio de Sanidad, Consumo y Bienestar Social 2017).

There are different causes that could explain this new stabilization situation. One of the most important is the implementation of public measures to prevent childhood obesity. In the COSI project, initiated by the WHO, a decrease of the BMI z-score was observed in those countries with the highest prevalence in the first round and where public health measures were taken to solve the situation (Wijnhoven et al. 2014). In Spain, the Aladino 2015 study (Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016) showed a decrease in the childhood overnutrition prevalence compared to the results of 2011 and authors believe it is related to the implementation of public health measures of the NAOS strategy. Besides the public health measures, there are other possible explanations for this stabilization such as having arrived at saturation levels and reaching balance (Bygdell et al. 2017). However, this stabilization phase could be transitory and there could be a new increase in the future, considering that in history there has been observed a tendency alternating phases of decrease or stabilization with phases of increase (Salcedo et al. 2010; Schmidt Morgen et al. 2013). Another reason would be the improvement of the SES since it has been observed that in recession times and in families with worse SES, there is a higher overnutrition prevalence (Schmidt Morgen et al. 2013; Lazzeri et al. 2015; Agencia Española de Consumo, Seguridad Alimentaria y Nutrición 2016; Herter-Aeberli et al. 2019). However, there is a limitation in our work because we could not study the involved factors in this decrease in our area and they should be studied in future works.

Conclusions

Even though there are still very high prevalence rates, our results showed a decrease in childhood overnutrition in the period from 1992 to 2011 in a Mediterranean area of Southeast Spain. Obesity decreased significantly according to WHO criteria while overweight remained stable. There was variability between sex, with a significant decrease only in males. The reference values should be selected with care, since the overnutrition prevalence varies depending on the criteria used. Serial studies should be carried out to confirm this trend.

Disclosure statement

The authors report no conflict of interest