INTRODUCTION

Worldwide, metabolic syndrome (MS) in the pediatric population is increasing and its prevalence reaches 6.7% in low- and middle-income countries ¹. This health issue is caused by a combination of factors such as lifestyle changes and eating habits ^{2 , 3}, obesogenic environments ⁴, among others. MS in children and adolescents increases the risk of developing MS, carotid intima-media thickening, and type II diabetes mellitus in adulthood ⁵; it has also been reported that greater severity of MS in childhood can predict cardiovascular accidents during adulthood ⁶.

Metabolic disorders are frequent in the pediatric population of high Andean areas. In indigenous children in Argentina -residents at an altitude of 3750 m - the prevalence of low HDL cholesterol levels (c-HDL) was 33% ⁷; another study reported that these children from the Argentine Andes had a higher proportion of hypertriglyceridemia compared to children at sea level (28.8% compared to 3.5%) ⁸. These conditions also occur in adolescents from Peruvian high Andean regions. In 2011, the frequency of low c-HDL and hypertriglyceridemia reached 44.8% and 19.4% among schoolchildren in Arequipa, respectively ⁹. In 2014, a study found 3.2% of MS, 46.4% of hypertriglyceridemia, and 37.0% of low c-HDL levels among 11- to 13-year-old schoolchildren in Cajamarca ¹⁰. This worrisome scenario in the high Andean regions requires the implementation of interventions that contribute to prevent the occurrence of the components of MS.

Nutritional educational interventions have shown to be beneficial in reducing abdominal obesity in children and adolescents ^{11 , 12}; in the latter, interventions that combine the nutritional aspect with physical activity reduce the body mass index (BMI) ¹³. We identified 46 systematic reviews (SR) on nutritional interventions in schools, in 39 SR the population was healthy; among these, 30 evaluated the effect on food intake, 21 on intermediate outcomes such as quality of life, anxiety-depression, academic performance, and only six evaluated the effect on outcomes such as MS, overweight-obesity, anthropometric measures, lipid profile, among others ¹⁴.

Childhood and adolescence are a window of opportunity to establish healthy lifestyles, and educational interventions are valid strategies for this purpose ^{2 , 14 - 16}. Randomized clinical trials (RCTs) are the most suitable designs for evaluating the efficacy of preventive or therapeutic interventions. Between 2006 and 2019 only ten RCTs evaluated the efficacy of educational interventions on MS in pediatric population, and only four carried out a nutritional intervention during school hours. These studies had a low risk of bias; however, they did not report their effect on MS and their results have been inconclusive ¹¹.

Schools are ideal scenarios for implementing educational interventions, due to their scope and influence on adolescents during their schooling ¹⁷. However, it is necessary to evaluate the effect in a real scenario of an educational intervention as a preventive strategy for MS. Therefore, the aim of this study was to estimate the effect of an educational intervention based on nutrition and healthy lifestyles to increase the frequency of adolescents free of some component of MS in two schools in an Andean region of Peru.

MATERIALS AND METHODS

RESULTS

Baseline characteristics of the sample

A total of 388 students were included, all of whom received the complete sessions of the intervention (Figure 1 of supplementary material). The 18 students excluded after the baseline measurement did not differ from those included according to age, sex and BMI (Table 1 of supplementary material). At baseline measurement, the mean age and standard deviation (SD) was 14.2 (1.7) years; 195 (50.3%) were male; other baseline characteristics are shown in Table 2.

Figure 1 Comparison of the pre- and post-intervention proportion (total and according to sex) of two outcomes in adolescents: absence of metabolic syndrome components (upper) and metabolic syndrome (three or more components) (lower). 

Table 2 Baseline characteristics of school adolescents who completed both measurements and stratified by sex. 

Characteristic	Total (n=388)	Male (n=195)	Female (n=193)	p-value
Abdominal circumference (cm), median (IQR)	72.00 (68.00-78.00)	73.00 (68.00-79.00)	71.00 (67.00-77.50)	0.058 a
Triglycerides (mg/dL), median (IQR)	100.00 (76.00-138.75)	96.00 (73.00-133.00)	104.00 (79.50-142.00)	0.078 a
HDL cholesterol (mg/dL), median (IQR)	37.00 (33.00-43.00)	35.00 (30.00-41.00)	38.00 (34.00-45.00)	<0.001a
LDL cholesterol (mg/dL), median (IQR)	87.20 (75.10-102.95)	83.00 (70.20-97.20)	92.00 (79.20-108.10)	<0.001a
Total cholesterol (mg/dL), median (CI)	147.00 (133.00-168.00)	143.00 (129.00-160.00)	156.00 (137.00-176.00)	<0.001a
Fasting plasma glucose (mg/dL), median (IQR)	80.00 (74.00-86.00)	81.00 (74.00-86.00)	80.00 (74.00-85.00)	0.775 a
Nutritional status (BMI), n (%) c
Normal	245 (63.3)	121 (62.4)	124 (64.2)	0.237 b
Overweight	111 (28.7)	53 (27.3)	58 (30.1)
Obese	31 (8.0)	20 (10.3)	11 (5.7)

IQR: interquartile range; BMI: body mass index.

^a Mann-Whitney U test

^b Chi-square for homogeneity between levels.

^c A case of thinness in men was not included in the analysis

Comparison of the components of metabolic syndrome according to sex and nutritional status

According to sex, the proportion of abdominal obesity decreased from 3.1% (6/195) to 2.1% (4/195) (p=0.500) in men and from 3.6% (7/193) to 2.1% (4/193) (p=0.250) in women; AHT was reduced from 8.7% (17/195) to 4.1% (8/195) (p=0.049) in men, and from 3.1% (6/193) to 1.6% (3/193) (p=0.453) in women; hypertriglyceridemia decreased from 36.9% (72/195) to 19.5% (38/195) (p<0.001) in men and from 45.1% (87/193) to 18.7% (36/193) (p<0.001) in women; the proportion of low c-HLD decreased from 70.3% (137/195) to 64.1% (125/195) (p=0.096) in men and from 61.1% (118/193) to 59.6% (115/193) (p=0.755) in women; finally, in men we found no cases of hyperglycemia at baseline measurement, but one post-intervention; among women the proportion remained at 0.5% (1/193) (p=1.00). Table 2 of the supplementary material shows the comparison of the medians of lipid profile, serum glucose, blood pressure and physical activity before and after the intervention, stratified by sex; Table 5 shows the changes in the number of MS components between baseline and post-intervention measurements. The changes in the proportion of MS and its components, according to nutritional status, are shown in Table 4.

Table 5 Comparison between the number of metabolic syndrome components at baseline and post-intervention measurement among adolescents, stratified by sex. 

n=frequency, p=proportion, McNemar-Bowker test (two-tailed test) in males (p<0.001), females (p<0.001) and total (p<0.001).

Cells shaded in green correspond to participants who decreased the number of components for the post-intervention measurement, cells in gray are those who remained with the same number of components, and cells in red are those who increased the number of components.

Note to facilitate reading of the table: Among 188 adolescents who had one component in the baseline measurement, 56 no longer had it postintervention; among the 92 males with one component, 26 presented such a change, and among the 96 females, 30 also had it. Among 103 students who had two components at baseline, 55 reduced to one component and 13 no longer had two components.

DISCUSSION

The educational intervention that was implemented in two secondary schools in a high Andean region of Peru achieved two relevant changes; first, an increase of 12 pp in the proportion of adolescents without MS components, and second, a significant reduction in the frequency of MS. After the intervention was completed, 11 of 18 adolescents who initially had MS no longer had having MS. This change could be explained mainly by the effect on hypertriglyceridemia, since 101 of 159 adolescents went from elevated triglyceride values to normal.

Unlike our research, most evaluations of educational interventions with single-group pre- and post-intervention design were conducted in adolescents or overweight children. Two studies were conducted in the educational setting, one in 98 Mexican children aged 6 to 12 years and reported a 44 to 16% reduction in MS, also reduced hypertriglyceridemia (64 to 35%), AHT (19 to 0%), and c-HDL≤ 40 mg/dL (60 to 41%) ¹⁷. The other study took place in Chile and found that 28 of 50 obese students, who completed an 8-month intervention, achieved a reduction in MS from 36 to 18%; this effect was caused by decreases in glucose, triglyceride, and c-LDL levels ²⁵. Other pre-experimental studies-but in non-school settings-show a varied effect in reducing MS. An intervention in 53 children and adolescents in Colombia, from a hospital-based obesity program, reduced MS from 37 to 32% ²⁶. A community intervention in 144 obese children and adolescents in the United States found a modest reduction of MS (30.6 to 27.1%) ²⁷. On the other hand, a study in 85 obese children from an Italian hospital found a decrease of 17.1 to 4.9% between pre- and postintervention measures ²⁸.

Our results are best compared with a study that evaluated the effect of an educational intervention on MS in adolescents from the general population. The study was conducted in Spain in 263 adolescents aged 12 to 16 years and found a significant reduction in MS (32.2 to 19.7%); in hypertriglyceridemia (4.7 to 0.8%), and the in those affected by c-HDL≤40 mg/dL (2.8 to 0.7%) ²⁹. Although we observed consistent results regarding the effect of these educational interventions on MS, the size of the effect differs between studies. This variability could be explained by the different types of target population -overweight or obese versus general population-; the implementation setting -schools, hospitals or community-; the duration and dosage of the intervention sessions; the sample size; the criteria for MS -ATP-III, Cook or IDF-, and the baseline prevalence of MS.

Interventions implemented in schools setting differ in duration; in our study, nine group sessions were conducted every 15 days, and each session lasted 45 min. Elizondo et al. ¹⁷ evaluated an intervention for overweight children by means of thirteen individual (not group) sessions every three weeks, during a school year; each session lasted 30 min. This study found that the intervention also reduced BMI and increased physical activity. Campos et al. ²⁹ designed an intervention for high school students characterized by group sessions every 15 days during the school year; the study did not specify the number or duration of sessions. This intervention also reduced daily calorie intake in males and females. On the other hand, Bustos et al. ²⁵ evaluated an intervention of 12 nutritional sessions provided during 8 months for obese schoolchildren, with one 40-minute session per week; physical activity sessions of 50 minutes were also carried out twice a week. This study also found an effect on BMI and the percentage of total fat in the trunk and the four extremities.

During the stratified analysis by sex, we found, a significant increase in the proportion of adolescents free of MS components in both males and females; on the other hand, the reduction in the proportion of MS -in both males and females- was not significant. A consistent reduction in hypertriglyceridemia was found in both sexes; this trend was not observed in the study by Campos et al. who, among Spanish schoolchildren, found a significant difference in mean triglycerides before and after the intervention only in boys ²⁹.

The intervention showed a differentiated effect according to the nutritional status of the adolescent, since no significant effect was found among obese adolescents. This low intensity of response to interventions in obese children and adolescents has been previously reported. This finding could suggest the need for specific interventions to reduce MS in this subgroup ^{27 , 29}. Interventions that incorporate physical exercise programs, as well as monitoring of physical activity at home are among the suggested strategies; and these should be carried out in addition to the intervention ^{25 , 29}. This study and most research have designed the intervention to be delivered in one school year ^{17 , 25 , 27 , 29}, so to improve outcomes in obese individuals, such interventions should be evaluated for more than 12 months.

This intervention produced a significant reduction in the proportion of hypertriglyceridemia, both in the overall sample and in strata defined by sex and nutritional status. A meta-analysis of RCTs to treat overweight and obese children and adolescents found that dietary interventions reduced triglyceride concentration more than 13.3 mg/dL on average, compared to the group that received diet plus exercise program ¹². A pre-experimental study in Mexican children reduced hypertriglyceridemia (≥110 mg/dL) from 64 to 35% ¹⁷, whereas in Spanish adolescents the reduction in participants with triglycerides >150 mg/dL was 4.7 to 0.8% ²⁹. Both triglycerides and c-LDL are highly correlated with food intake ³⁰; therefore, a direct effect of nutritional educational interventions on the dietary habits of adolescents is plausible, generating not only a reduction in triglyceride levels but also in c-LDL. Although overweight and obesity in Latino adolescents is associated with dyslipidemias ^{31 , 32}, in this study the prevalence of obesity was only 8%, while that of hypertriglyceridemia was 41.0%; therefore, it is unlikely that the effect on triglycerides is explained by the reduction of excess weight.

The 3.8 pp reduction in the proportion of students with low c-HDL levels was not significant; this finding differs from a study, not restricted to overweight adolescents, in Spanish schoolchildren aged 12 to 16 years in which a significant increase in the mean c-HDL level was found between pre- and post-intervention measurements ²⁹. The favorable effect of educational interventions on the increase in c-HDL levels has also been found in pre-experimental studies in overweight adolescents ^{17 , 27 , 28}. Similar to our study, Bustos et al. reported a decrease in mean c-HDL levels in Chilean children and adolescents ²⁵. Our finding could be due to the fact that 35 students with c-HDL values>40 mg/dL at baseline measurement increased to levels ≤40 md/dL postintervention, this phenomenon would correspond to a potential maturation bias, as it has been seen in adolescents aged 10 to 17 years in Brazil, in whom c-HDL levels tend to decrease approaching a mean of 40 mg/dL at older age ³³.

The limitations of the study are related to the absence of a control group; however, we implemented the intervention for 6 months and the baseline and postintervention measurements were performed at that time, thus reducing the biological maturation effect ³⁴ and the loss of participants. During both measurements we did not perform repeated tests for glucose, triglycerides, and c-HDL levels, which could affect the precision of these measurements and thus generate a potential regression bias to the mean, affecting the estimation of the effect of the intervention. To reduce this risk, tests were made on an empty stomach, with adequate pre-analytical treatment of the sample and using standardized and calibrated laboratory procedures, thus reducing instrumentation bias. The intervention was carried out in two schools, one exclusively for females and the other for males; therefore, some differences between the two settings, which were not measured in the study, could confound the effect of the intervention. Intermediate outcomes such as changes in food consumption or level of knowledge about healthy eating were not assessed. Finally, it was not possible to exclude potential cases of primary dyslipidemias, which, depending on their prevalence, could cause selection bias.

In conclusion, the educational intervention on nutrition and healthy lifestyles implemented in two schools in an Andean area of Peru increased the number of adolescents free of MS components, it also caused a significant reduction in the proportion of adolescents with MS, the latter effect being explained mainly by the change in triglyceride levels. We recommend evaluating the effect of the intervention in adolescents from schools in the coast and jungle regions, and it should be adapted to the availability of food in the region, as well as the cultural context.