Association of nutrient patterns with anthropometric indices in children and adolescents: The weight disorders survey of the CASPIAN-IV study

Abstract

BACKGROUND:

Nutrient patterns play a role as an interface between food patterns and the food metabolome integrating measurements. The nutrients can improve our knowledge about the reason of some chronic diseases.

OBJECTIVE

The aim of the present study was to identify the major nutrient patterns in adolescents and to assess their relationship with obesity.

METHODS

This is a nationwide cross-sectional study. Usual dietary intakes were collected using a validated 168-item semi-quantitative food frequency questionnaire (FFQ).

RESULTS

Dietary data were analysed in 4288 subjects aged 11.43±3.23 years. Subjects in the fourth quartile of the first nutrient pattern tended to have higher weight, body mass index (BMI), waist circumference (WC) and hip circumference than those in the first quartile. Individuals in the fourth quartile of the second nutrient pattern had significantly lower means of weight, WC and hip circumference than those in the first quartile. The third nutrient pattern was not correlated with any alteration in BMI and wrist circumference in boys as well as in BMI, waist circumference and wrist circumference among girls.

CONCLUSIONS

Findings indicated that second nutrient pattern which mostly characterized by high consumption in mono-unsaturated fatty acid, poly-unsaturated fatty acid, potassium, calcium, vitamin E, biotin and vitamin K was associated with lower risk of obesity, while first nutrient pattern with high amounts of carbohydrate, thiamin, iron and manganese was correlated with higher risk of obesity.

Keywords

Anthropometry diet nutrient intake children

1 Introduction

Recently, overweight and obesity are one of the main public health concerns in children and adolescents. Its prevalence has increased in the past three decades [1]. Evidence has shown the increasing prevalence of childhood and adolescent obesity in developing countries as well as in Iran [2]. Excess body weights have a devastating effect on health and quality of life in adolescents [3]. In addition, it associates with several risk factors which lead to a cluster of chronic disease in older age [4].

Multiple aetiologies are the cause of imbalance between energy intake and expenditure [5]. Genetic and environmental factors, lifestyle preferences, and cultural environment have a major role in obesity [6]. Lifestyle is the most modifiable factor for management of obesity. The role of diet as a determinant factor of obesity has been well established [7].

An alternative method in nutritional epidemiology for examining diet-disease relationships is the dietary pattern analysis [8]. Dietary patterns provide better insight into the correlation between diet and obesity than the intake of individual nutrients or foods [9]. According to evidence, food patterns can predict the risk of chronic diseases. However, dietary patterns approach cannot explain the mechanisms of patterns that might affect the risk of diseases. Food patterns through nutrient intakes influence the risk of chronic diseases. A combination of nutrients more than one nutrient affect the risk of diseases. More information about underlying mechanisms of diseases and their aetiology may be provided by nutrient patterns. There are several studies on food patterns and chronic diseases. However, few studies have assessed the relationship between nutrient patterns and chronic diseases [10, 11]. Foods and methods of cooking differ between population because of behavioural, cultural and geographical differences. Thus, comparisons of findings between studies are difficult. Unlike food patterns, nutrients are universal and consumed in various population and cultures [12 –14]. According to finding, nutritional biomarkers, metabolites and nutrient patterns play a role as an interface between food patterns and the food metabolome integrating measurements [15]. The nutrients can improve our knowledge about the reason of some chronic diseases. In fact, nutrients clearly derived from foods. So, nutrient patterns provide information on the aetiology of chronic diseases [11].

The aim of the present study was to determine the major nutrient patterns in Iranian children and adolescents and to assess relationship of nutrient patterns with anthropometric indices.

2 Materials and methods

This nationwide cross-sectional study was performed within the framework of the Weight disorders survey of the national survey of school students’ high-risk behaviours” of the school-based surveillance system entitled Childhood and Adolescence Surveillance and PreventIon of Adult Noncommunicable disease (CASPIAN) that included a large group of Iranian adolescents from 30 provinces in Iran. Detailed information about the study design, participants and data collection methods has been published previously [16, 17]. Study protocols were reviewed and approved by Research and Ethics council of Isfahan University of Medical Sciences, Iran. Written informed consent and verbal consent were obtained from the parents and students, respectively. This manuscript was written in accordance with the STROBE guidelines checklist.

Usual dietary intakes were collected using a validated 168-item semi-quantitative food frequency questionnaire (FFQ) in 5000 sample. This questionnaire was designed and validated previously. Detailed information about the foods included, design and the face validity of this FFQ has been reported previously [18].

Participants were asked to report their dietary intakes of foods. Amount of each food item was converted into gram using household scale guide and calculated for one day. We used Nutritionist IV software (First DataBank, San Bruno, CA) for nutrient analysis.

In the current analysis, we used carbohydrate, protein, total dietary fat, cholesterol, saturated fatty acid, mono unsaturated fatty acid, poly unsaturated fatty acid, vitamin C, thiamine, niacin, vitamin B6, folate, vitamin K, vitamin E, biotin, vitamin A, vitamin D, riboflavin, vitamin B12, pantothenic acid, magnesium, iron, manganese, selenium, chromium, potassium, calcium, phosphorus, zinc, copper, total dietary fibre, caffeine and sugar to identify nutrient patterns.

Body weight and height were measured with subjects in light clothing and without shoes. Body mass index (BMI) was calculated as weight (kg) divided by the height (m²). Waist circumference (WC) was measured to the nearest 0.1 cm by a non-elastic tape between the lower border of the rib cage and the iliac crest. Hip circumference was measured at the widest part of the hip at the level of the greater trochanter to the nearest 0.1 cm. Wrist circumference was measured to the nearest 0.1 cm on the dominant arm by a tape meter. Subjects were asked to hold their arm on a flat surface as a table. The superior border of the tape measure was placed just distal to the prominences of radial and ulnar bones. Neck circumference was measured with an accuracy of 0.1 cm. The most prominent portion of the thyroid cartilage was considered as a landmark.

To collect information about age, gender, education (Primary school, Middle school, High school), physical activity, living area and socioeconomic status, we used a validated questionnaire that used in our previous studies. Physical activity pattern was investigated through the questionnaires of the surveillance program [14].

To identify major nutrient patterns, factor analysis was used with orthogonal transformation (varimax procedure). We retained factors with eigenvalues≥1.5 as this cutoff could result in more interpretable nutrient patterns. Eigenvalues less than 1.5 did not explain sufficient amounts of overall variation. The number of factors retained was determined by eigenvalue (≥1.5), scree plot, factor interpretability, and the variance explained by each factor. Based on the analysis, three factors were selected [19]. Firstly, participants were classified into the three nutrient patterns and then subsequently into quartiles based on the patterns. General characteristics across quartiles were compared using one-way analysis for continuous variables. The distribution of categorical variables across quartiles was assessed by Chi-square tests. Analysis of covariance (ANCOVA) with Bonferroni correction was used for comparison of anthropometric measures means across quartiles of nutrient pattern scores. Means of anthropometric measures across quartiles of nutrient pattern scores were calculated in various models. Firstly age (continuous), and energy intake (continuous) were adjusted. In the second model, we further controlled for physical activity, education, family size, living area, socioeconomic status.

The first quartile of nutrient pattern scores was considered as a reference. For all analyses, SPSS 20 (SPSS, IBM, USA) was used. P < 0.05 was considered as significant.

3 Results

Dietary data were analysed in 4288 subjects. The mean age of participants was 11.43±3.23 years, and 47.4% of them were girls. General characteristics of participants are showed in Table 2.

We recognized three major nutrient patterns: The first pattern was high in carbohydrate, vitamin C, iron, thiamine, niacin, vitamin B6, folate, vitamin K, magnesium, manganese, selenium, chromium, total dietary fibre, and caffeine. The second pattern was high in total dietary fat, saturated fatty acid, mono-unsaturated fatty acid, poly-unsaturated fatty acid, potassium, calcium, vitamin E, biotin, phosphorus, and sugar. The third pattern was high in protein, cholesterol, vitamin A, vitamin D, riboflavin, vitamin B12, pantothenic acid, zinc, and copper. These three nutrient patterns described 63% of the total variance of nutrient consumption. Factor loading matrix for three nutrient patterns is shown in Table 1.

Table 1
Factor loading matrix for three nutrient patterns

Factor 1 Factor 2 Factor 3

Protein 0.48 0.50 0.75

Carbohydrate 0.71 0.45

Total fat 0.20 0.87 0.21

cholesterol 0.27 0.61

Saturated fatty acid 0.78 0.36

Monounsaturated fatty acids 0.84

polyunsaturated fatty acids 0.21 0.71

Potassium 0.43 0.63 0.34

Vitamin A 0.83

Vitamin C 0.54 0.23

Calcium 0.22 0.63 0.36

Iron 0.80 0.29 0.35

Vitamin D 0.30 0.64

Vitamin E 0.21 0.66

Thiamin 0.77 0.37 0.27

Riboflavin 0.22 0.41 0.80

Niacin 0.67 0.27 0.41

Vitamin B6 0.63 0.51 0.36

Folate 0.75 0.38 0.24

Vitamin B12 0.89

Biotin 0.35 0.65 0.37

Pantothenic 0.45 0.51 0.62

Vitamin K 0.51 0.22

Phosphorus 0.43 0.62 0.53

Magnesium 0.71 0.54 0.31

Zinc 0.39 0.38 0.59

Copper 0.40 0.72

Manganese 0.88

Selenium 0.58 0.37

Chromium 0.59

Fiber 0.73 0.28

Sugar 0.21 0.46

Caffeine 0.57

	Factor 1	Factor 2	Factor 3
Protein	0.48	0.50	0.75
Carbohydrate	0.71	0.45
Total fat	0.20	0.87	0.21
cholesterol		0.27	0.61
Saturated fatty acid		0.78	0.36
Monounsaturated fatty acids		0.84
polyunsaturated fatty acids	0.21	0.71
Potassium	0.43	0.63	0.34
Vitamin A			0.83
Vitamin C	0.54	0.23
Calcium	0.22	0.63	0.36
Iron	0.80	0.29	0.35
Vitamin D		0.30	0.64
Vitamin E	0.21	0.66
Thiamin	0.77	0.37	0.27
Riboflavin	0.22	0.41	0.80
Niacin	0.67	0.27	0.41
Vitamin B6	0.63	0.51	0.36
Folate	0.75	0.38	0.24
Vitamin B12			0.89
Biotin	0.35	0.65	0.37
Pantothenic	0.45	0.51	0.62
Vitamin K	0.51	0.22
Phosphorus	0.43	0.62	0.53
Magnesium	0.71	0.54	0.31
Zinc	0.39	0.38	0.59
Copper	0.40	0.72
Manganese	0.88
Selenium	0.58		0.37
Chromium	0.59
Fiber	0.73	0.28
Sugar	0.21	0.46
Caffeine	0.57

Table 2

General characteristics of participants across quartile of nutrient pattern scores

Variables	First nutrient pattern				P^¥	Second nutrient pattern				P	Third nutrient pattern				P
	Q1	Q2	Q3	Q4		Q1	Q2	Q3	Q4		Q1	Q2	Q3	Q4
Age	10.97±3.27 *	11.11±3.18	11.68±3.23	12.01±3.07	<0.001	11.73±3.23	11.44±3.22	11.31±3.22	11.19±3.18	0.001	11.93±3.25	11.48±3.18	11.15±3.16	11.15±3.24	<0.001
Gender % (n)
female	48.5 (560)	49.4 (539)	47.7 (509)	43.8 (450)	0.055	43.6 (497)	45.6 (490)	49.4 (517)	51.2 (554)	0.001	47.7 (506)	48.3 (525)	48.3 (533)	45.3 (494)	0.43
male	51.5 (595)	50.6 (552)	52.3 (559)	56.2 (577)		56.4 (643)	54.4 (584)	50.6 (529)	48.8 (527)		52.3 (554)	51.7 (562)	51.7 (570)	54.7 (597)
Education % (n)
Primary school	51.7 (597)	51.6 (563)	42.1 (450)	35.2 (361)	<0.001	39.7 (453)	45.6 (490)	48.2 (504)	48.5 (524)	<0.001	35.5 (376)	43.2 (470)	49 (541)	53.5 (584)	<0.001
Middle school	28.7 (331)	28.2 (308)	29.4 (314)	35.8 (368)	33.9 (386)	29.4 (316)	27.5 (288)	30.6 (331)	38.2 (405)	31.8 (346)	29.2 (322)	22.7 (248)
High school	19.7 (227)	20.2 (220)	28.5 (304)	29 (298)	26.4 (301)	25 (268)	24.3 (254)	20.9 (226)		26.3 (279)	24.9 (271)	21.8 (240)	23.7 (259)
Energy
(kcal/day)	1975.53±709	2359.95±618	2709.34±626	3170.47±623	<0.001	1798.05±580	2297.24±519	2756.82±508	3363.05±494	<0.001	2252.14±808	2302.62±718	2609.47±670	3051±647	<0.001
Fat (g/day)	84.21±40.17	91.17±35.25	100.37±34.87	106.97±33.73	<0.001	55.41±16.23	80.62±14.66	104.19±15.71	142.47±25.95	<0.001	85.80±40.52	85.14±34.01	96.68±32.82	115.10±32.36	<0.001
Protein
(g/day)	64.47±26.58	77.95±25.52	90.35±26.34	105.12±28.20	<0.001	60.32±24.69	76.39±21.46	90.83±23.39	110.34±27.83	<0.001	60.67±21.12	7289±19.97	88.40±20.34	115.92±28.03	<0.001
Carbohydrate
(g/day)	248.42±84.49	318.49±76.42	377.22±79.26	469.83±104.68	<0.001	273.71±109.55	329.67±103.93	380.81±106.13	429.78±95.00	<0.001	322.88±132.35	324.74±113.06	361.87±105.97	404.51±103.11	<0.001
Dietary fiber
(g/day)	17.59±6.61	24.64±7.31	31.34±8.80	42.50±15.60	<0.001	22.58±14.28	27.56±11.88	31.53±12.86	34.40±12.87	<0.001	26.23±14.24	27.28±13.48	29.87±12.44	32.70±13.81	<0.001

^*Mean±standard deviation (SD). ^¥Obtained from ANOVA or Chi-square test, where appropriate.

Individuals in the fourth quartile compared with those in the first quartile were older and less likely to be girls in the first nutrient pattern. Within the second nutrient pattern, participants in the fourth quartile were more girls compared with those in the first quartile. Individuals in the fourth quartile in all nutrient patterns had higher intakes of energy, fat, protein, carbohydrate and dietary fibre in comparison with those in the first quartile.

Multivariable-adjusted means of anthropometric measures across quartiles of nutrient pattern scores are summarized in Table 3 and 4. There were not any significant differences in wrist circumference across quartiles of all nutrient patterns in crude and adjusted models in both genders.

Table 3

Multivariable-adjusted means for anthropometric measures across quintiles of nutrient patterns for Boys

	First nutrient pattern				P	Second nutrient pattern				P	Third nutrient pattern				P
	Q1	Q2	Q3	Q4		Q1	Q2	Q3	Q4		Q1	Q2	Q3	Q4
Weight
Crude	38.34±0.66	40.14±0.77	43.06±0.77	47.17±0.76	<0.001	43.78±0.72	43.50±0.78	40.67±0.71	40.19±0.79	<0.001	41.93±0.73	42.53±0.73	43.34±0.83	40.91±0.72	0.135
Model I	39.89±0.61	40.95±0.57	42.72±0.56	54.08±0.62	<0.001	44.25±0.65	43.58±0.56	40.82±0.59	39.36±0.74	<0.001	41.19±0.58	42.50±0.57	43.84±0.56	41.10±0.58	0.001
Model II	40.64±0.55	41.94±0.51	42.05±0.51	44.01±0.56	0.001	43.28±0.58	43.45±0.50	40.98±0.53	40.51±0.67	0.002	41.43±0.52	42.03±0.51	43.69±0.50	41.48±0.52	0.004
BMI
Crude	18.01±0.17	18.56±0.25	18.53±0.18	19.64±0.27	<0.001	18.69±0.17	19.04±0.22	18.42±0.27	18.54±0.23	0.244	18.51±0.17	18.52±0.17	18.95±0.25	18.74±0.27	0.121
Model I	18.21±0.23	18.67±0.22	18.49±0.22	19.37±0.24	0.009	18.78±0.25	19.06±0.21	19.43±0.23	18.40±0.28	0.221	18.42±0.22	18.52±0.22	19.02±0.21	18.76±0.22	0.238
Model II	18.34±0.23	18.82±0.22	18.39±0.21	19.21±0.23	0.029	18.66±0.24	19.02±0.21	18.44±0.22	18.60±0.28	0.314	18.51±0.22	18.46±0.22	18.98±0.21	18.78±0.22	0.316
Waist circumference
Crude	64.83±0.46	66.20±0.54	67.42±0.54	70.57±0.53	<0.001	68.11±0.48	68.25±0.55	66.23±0.53	66.11±0.53	0.002	66.36±0.53	67.83±0.54	68.23±0.54	66.58±0.48	0.029
Model I	65.64±0.49	66.63±0.46	67.26±0.46	69.47±0.50	<0.001	68.82±0.52	68.44±0.45	66.15±0.48	65.10±0.60	<0.001	66.02±0.47	67.86±0.46	68.50±0.45	66.60±0.46	<0.001
Model II	66.15±0.47	67.06±0.44	66.83±0.44	68.92±0.48	0.001	68.39±0.50	68.30±0.43	66.21±0.46	65.68±0.58	0.001	66.33±0.45	67.62±0.44	68.28±0.43	66.74±0.45	0.008
Hip circumference
Crude	76.72±0.51	77.84±0.56	80.02±0.56	83.18±0.54	<0.001	80.63±0.52	80.07±0.58	78.60±0.53	78.09±0.56	0.003	79.09±0.58	79.99±0.54	80.15±0.57	78.52±0.52	0.469
Model I	77.80±0.47	78.41±0.44	79.79±0.44	81.72±0.48	<0.001	81.19±0.50	80.20±0.43	78.63±0.46	77.22±0.57	<0.001	78.68±0.45	80.03±0.44	80.48±0.43	78.53±0.45	0.002
Model II	78.28±0.45	78.94±0.42	79.35±0.41	81.11±0.45	0.001	80.66±0.47	80.07±0.41	78.75±0.43	77.86±0.55	0.005	78.91±0.43	79.71±0.42	80.34±0.41	78.75±0.42	0.023
Neck circumference
Crude	30.24±0.23	30.20±0.21	30.76±0.20	31.59±0.19	<0.001	31.03±0.20	30.83±0.20	30.57±0.22	30.28±0.23	0.081	30.37±0.18	30.98±0.24	30.90±0.22	30.56±0.20	0.154
Model I	30.64±0.20	30.39±0.19	30.67±0.19	31.07±0.20	0.136	30.96±0.21	30.79±0.18	30.68±0.19	30.33±0.24	0.457	30.17±0.19	30.93±0.19	31.01±0.18	30.67±0.19	0.007
Model II	30.80±0.19	30.61±0.18	30.54±0.18	30.86±0.20	0.551	30.73±0.21	30.78±0.18	30.73±0.19	30.55±0.24	0.902	30.16±0.18	30.83±0.18	31.02±0.18	30.78±0.18	0.008
Wrist circumference
Crude	15.29±0.35	14.91±0.21	14.95±0.08	15.69±0.27	0.113	15.35±0.24	15.35±0.26	14.83±0.08	15.29±0.35	0.433	15.34±0.28	15.08±0.19	15.21±0.19	15.25±0.32	0.911
Model I	15.55±0.27	15.02±0.25	14.88±0.25	15.37±0.28	0.219	15.32±0.29	15.33±0.25	14.87±0.26	15.30±0.33	0.535	15.27±0.26	15.07±0.25	15.25±0.25	15.26±0.26	0.935
Model II	15.62±0.28	15.11±0.26	14.84±0.26	15.28±0.28	0.219	15.23±0.29	15.34±0.25	14.89±0.27	15.40±0.34	0.528	15.25±0.26	15.03±0.26	15.26±0.25	15.33±0.26	0.875

Data are mean±standard error (SE). Model I: adjusted for age and energy intake. Model II: additionally adjusted for education, family size, physical activity, living area, socioeconomic status.

Table 4

Multivariable-adjusted means for anthropometric measures across quintiles of nutrient pattern for Girls

	First nutrient pattern				P	Second nutrient pattern				P	Third nutrient pattern				P
	Q1	Q2	Q3	Q4		Q1	Q2	Q3	Q4		Q1	Q2	Q3	Q4
weight
Crude	39.21±0.66	41.44±0.71	44.66±0.72	44.28±0.76	<0.001	43.18±0.74	42.57±0.71	42.48±0.70	40.91±0.71	0.132	43.62±0.75	42.38±0.71	41.96±0.72	41.02±0.70	0.089
Model I	38.55±0.73	41.28±0.70	44.86±0.72	45.05±0.82	<0.001	45.27±0.92	43.30±0.75	41.97±0.72	38.84±0.88	<0.001	43.77±0.74	42.61±0.72	41.90±0.70	40.66±0.77	0.044
Model II	40.34±0.55	42.43±0.52	43.42±0.54	43.24±0.62	0.001	43.40±0.68	42.51±0.56	42.46±0.54	40.88±0.66	0.131	41.41±0.56	42.21±0.53	43.12±0.52	42.34±0.58	0.179
BMI
Crude	18.72±0.23	19.06±0.23	19.92±0.24	20.33±0.74	0.013	20.15±0.68	19.40±0.22	19.51±0.24	18.85±0.23	0.124	19.29±0.23	19.81±0.64	19.37±0.25	19.35±0.25	0.765
Model I	18.41±0.40	18.97±0.38	20.03±0.39	20.71±0.45	0.002	20.83±0.50	19.64±0.41	19.35±0.39	18.18±0.48	0.014	19.25±40	19.81±0.39	19.37±0.38	19.40±0.42	0.754
Model II	18.81±0.39	19.20±0.37	19.74±0.39	20.35±0.44	0.098	20.47±0.49	19.45±0.40	19.45±0.39	18.61±0.47	0.133	18.78±0.40	19.74±0.38	19.62±0.37	19.76±0.42	0.266
Waist circumference
Crude	63.45±0.46	65.40±0.50	67.40±0.52	67.58±0.53	<0.001	66.32±0.50	65.85±0.49	66.58±0.54	64.71±0.49	0.040	66.03±0.53	65.95±0.50	65.80±0.47	65.58±0.53	0.934
Model I	62.94±0.52	65.27±0.49	67.57±0.51	68.18±0.58	<0.001	67.81±0.65	66.36±0.53	66.24±0.51	63.25±0.62	<0.001	66.12±0.52	66.06±0.51	65.77±0.50	65.39±0.55	0.791
Model II	63.85±0.46	65.74±0.44	66.87±0.46	67.30±0.52	<0.001	66.82±0.57	66.02±0.47	66.40±0.45	64.27±0.56	0.013	65.03±0.47	65.89±0.45	66.32±0.44	66.10±0.49	0.250
Hip circumference
Crude	79.08±0.57	81.14±0.59	83.81±0.66	84.80±0.63	<0.001	82.53±0.62	82.25±0.66	82.81±0.62	80.68±0.58	0.062	82.16±0.69	82.79±0.60	82.03±0.59	81.12±0.60	0.306
Model I	78.48±0.63	80.99±0.60	84.00±0.62	85.51±0.71	<0.001	84.64±0.79	82.97±0.65	82.32±0.62	78.62±0.76	<0.001	82.36±0.64	83.03±0.62	81.98±0.61	80.72±0.67	0.103
Model II	79.80±0.50	81.83±0.48	82.86±0.50	84.25±0.57	<0.001	83.10±0.63	82.45±0.52	82.66±0.49	80.19±0.61	0.007	80.51±0.51	82.69±0.49	82.96±0.48	81.97±0.53	0.002
Neck circumference
Crude	29.63±0.19	29.74±0.28	30.38±0.23	30.27±0.23	0.062	30.55±0.27	29.96±0.23	29.76±0.21	29.69±0.23	0.051	30.51±0.25	30.09±0.26	29.75±0.20	29.59±0.23	0.035
Model I	29.37±0.24	29.67±0.23	30.47±0.24	30.58±0.27	0.003	30.72±0.30	30.03±0.25	29.72±0.24	29.53±0.29	0.077	30.47±0.24	30.08±0.24	29.74±0.23	29.64±0.26	0.096
Model II	29.64±0.23	29.88±0.21	30.19±0.22	30.34±0.26	0.242	30.39±0.28	29.96±0.23	29.76±0.22	29.87±0.27	0.424	30.03±0.23	30.02±0.22	29.97±0.22	29.95±0.24	0.995
Wrist circumference
Crude	14.78±0.27	15.15±0.38	15.00±0.16	15.12±0.14	0.778	15.60±0.41	14.87±0.28	14.84±0.22	14.75±0.19	0.129	15.60±0.38	15.10±0.31	14.60±0.09	14.73±0.28	0.061
Model I	14.58±0.29	15.10±0.28	15.06±0.29	15.35±0.33	0.406	15.71±0.36	14.90±0.30	14.81±0.28	14.64±0.35	0.217	15.56±0.29	15.08±0.28	14.60±0.28	14.78±0.30	0.114
Model II	14.74±0.30	15.15±0.28	14.97±0.29	15.21±0.33	0.704	15.59±0.37	14.81±0.30	14.84±0.29	14.80±0.36	0.331	15.47±0.30	15.07±0.28	14.63±0.28	14.85±0.31	0.252

Data are mean±standard error (SE). Model I: adjusted for age and energy intake. Model II: additionally adjusted for education, family size, physical activity, living area, socioeconomic status.

In crude and adjusted models, both genders in the fourth quartile of the first nutrient pattern tended to have higher weight, BMI, waist circumference and hip circumference than those in the first quartile. No significant difference in neck circumference was shown across the first nutrient pattern quartiles in boys after adjustment for confounding factors. First nutrient pattern was correlated with increased neck circumference in crude model in boy and in model I in girls.

The second nutrient pattern was not correlated with any changes in BMI, neck circumference and wrist circumference in boys and in neck circumference and wrist circumference among girls. Boys in the fourth quartile of the second nutrient pattern had significantly lower means of weight, waist circumference, and hip circumference than those in the first quartile. Girls in the fourth quartile of the second nutrient pattern had significantly lower means of weight and BMI than those in the first quartile only in model 1. When the more potential confounders were considered, these relationships became non-significant. The mean of waist circumference and hip circumference in the fourth quartile of the second nutrient pattern had significantly lower than the first quartile in model 1 and 2 in girls.

The third nutrient pattern was associated with changes of weight, hip and neck circumference only in adjusted models in boys. The third nutrient pattern was not correlated with any changes in BMI and wrist circumference in boys and in BMI, waist circumference and wrist circumference among girls. Girls in the fourth quartile of the third nutrient pattern had significantly lower means of weight than those in the first quartile only in model 1. Third nutrient pattern associated with hip circumference in adjusted model II in girls. Changes in neck circumference associated with third nutrient pattern in crude model in girls.

4 Discussion

To the best of our knowledge, this is the first study that assesses the relationship between nutrient patterns and weight disorders in Iranian adolescents and children. In this cross-sectional study, we found significantly protective association between second nutrient pattern and some anthropometric indices in both genders. In contrast, the first nutrient pattern characterized by carbohydrate, thiamin, iron, vitamin B6, folate, and fibre was positively related to increase weight, BMI, waist circumference and hip circumference. However, after full adjustment including age, energy, physical activity, education and family size, a completely different result emerged in some of the variables. Previous studies on nutrient patterns have shown a significant relationship between nutrient patterns and obesity in adult [20, 21] and adolescence [22].

Iran as developing countries has experienced an accelerating nutrition transition due to economic and social conditions development in the past decades that lead to significant changes in the dietary habits of the population [23, 24].

Overweight and obesity have the adverse effects on metabolic changes. The adipose cell is known as the endocrine cell. The excess dietary calories lead to increase of fat cells, metabolic changes, and aberrations [25, 26].

Diet consists of a mixture of nutrients have a synergistic influence on health outcomes. Investigations of nutrient patterns in comparison with food patterns have several advantages. Consist of nutrients are global and their function are not exchangeable. Thus, behaviours and habits do not affect nutrient structures and the evaluation of nutrient patterns in different populations is effective for showing the association between nutrient patterns and health outcomes in various geographical regions [27].

Assessment the combination of nutrients and nutrient patterns are an alternative approach to investigate the influence of food intake on diseases. Nutrient patterns are an easier way compared with dietary patterns to compare between populations. The same component nutrients in foods are consumed across populations despite greater differences in food patterns [28]. In addition, nutrients demonstrate a limited number of non-consumers, unlike foods. The nutrient pattern show better association between a combination of bioactive nutrients in complex biological mechanisms and diseases compare with food patterns [15].

According to studies, the main source of energy intake among Iranian population is plant foods. The intake of grains, especially refined grains is the main food item of the traditional Iranian diet [29]. The physiological reason for weight gain is the high energy intake from foods and drinks. Satiety responses or appetite control involve in the maintenance of energy balance that depends on the type of foods intake [30]. Potential mechanisms of our findings that first nutrient pattern correlate with weight gain may be related to the carbohydrate intake. Also, nutrients including thiamine [31] and niacin [32] in the first nutrient pattern have been positively correlated with obesity. According to previous findings, appetite may be stimulated by B vitamins. So, intake in long-term may lead to more energy consumption and finally weight gain [33]. In contrast, dietary intakes of folate [34], iron [35] and dietary fibre [36] have been reversely correlated with obesity in previous studies. According to evidence, fibre has desirable effects on satiety and appetite control [30]. The combination of nutrients with inducing or protective effects on body fat accumulation in nutrient patterns lead to difficult interpretation. However, our findings support the nutrient pattern approach for investigation the relation between diet and disease [37, 38]. In addition, according to our findings, the interaction between nutrients including obesity-inducing or obesity protective may affect diet– disease relations and require much further research.

The second nutrient pattern was high in total dietary fat, saturated fatty acid, mono-unsaturated fatty acid, poly-unsaturated fatty acid, potassium, calcium, vitamin E, biotin, phosphorus, and sugar. We showed an inverse association between second nutrient pattern and obesity. However, it contains both useful and harmful nutrients. Saturated fatty acid and sugar intake increase the risk of cardiovascular disease, diabetes, inflammation and etc. However, unsaturated fatty acids have beneficial effect on health [39]. According to our findings, combination of all type of fatty acids can determine their influence on health. Useful and harmful nutrients consumption at the same time can modify the function of nutrients.

These findings suggest more studies for investigation the effects of supplementation with nutrients on obesity in further studies. The mechanism of protective effect of the second pattern may be correlated with calcium, mono, and polyunsaturated fatty acid. It was indicated that dairy products and high dietary calcium have an important effect on prevention and treatment of obesity and the energy metabolism regulation. Diets contain high calcium reduce weight gain in the overconsumption of high calorie diet, increase lipolysis and keep thermogenesis in energy restriction. Thus, calcium can speed up weight loss [40].

According to evidence, increasing the intake of polyunsaturated fatty acid (PUFA) especially long-chain n-3 may improve body composition. Potential mechanisms may be linked to the alteration the gene expression that enhance fat oxidation in some tissue including intestinal, adipose, cardiac, liver and skeletal muscle and decrease fat deposition in adipose tissue [41]. Findings demonstrated that n-3 PUFA intake may reduce postprandial sensations related to hunger [42, 43], enhances the lean body mass, increases metabolic rate and finally helps reduction of body fat [44, 45].

Findings showed monounsaturated fatty acid (MUFA) negatively related to the obesity prevalence. Some mechanisms relate to this negative association are diet thermogenesis increases by MUFA consumption that stimulates the sympathetic nervous system. β-adrenoreceptors have more density and sensitivity in abdominally obese subjects. Thus, these subjects respond more to stimulation of the sympathetic nervous system. In addition, MUFA consumption may stimulate utilisation of fat by activation of the nuclear receptor, PPAR-α [46].

In the present study, failure to observe meaningful relationships between nutrient patterns and some anthropometric indices may be due to unexpected confounder that we were unable to recognize them or interaction between nutrients.

A major strength of the present study is the large population which makes our finding more reliable. We used several variables including neck, waist and wrist circumstance instead of BMI and weight alone to the comprehensive measure of variables related to overweight and obesity in adolescents. Also, maximum number of nutrients related to obesity and bioactive compounds were considered in the present study.

There are some limitations. First, because of the cross-sectional design of the study, the relationship between the three nutrient patterns and obesity among adolescences cannot infer causality. Second, due to the observational nature of the data, we could not consider all residual confounding factors. Third, self-reported dietary data lead to measurement error. It is one of the limitations in all epidemiological studies. Energy-adjusted nutrient intake was used for minimizing the bias. We adjusted age however the effects of pubertal stage were not considered. In addition, participants with implausible diet reporting were excluded.

Our findings indicated that a nutrient pattern which mostly characterised by high consumption in monounsaturated fatty acid, polyunsaturated fatty acid, potassium, calcium, vitamin E and biotin was associated with lower obesity, while a pattern of nutrient intake with high amounts of carbohydrate, thiamin, iron, and manganese was associated with greater obesity. Prospective studies are needed to assess any causal correlation between nutrient patterns and obesity and confirm the present findings. The findings of these studies can help to choose the best food items that prevent weight gain based on their nutrient content.

Funding

The authors report no funding.

Conflict of interest

The authors declare that they have no competing interests.

Footnotes

Acknowledgments

The authors have no acknowledgments.

References

Frederick

, Snellman

, Putnam

. Increasing socioeconomic disparities in adolescent obesity. Proc Natl Acad Sci USA. 2014;111:1338–42.

Kelishadi

, Haghdoost

, Sadeghirad

, Khajehkazemi

. Trend in the prevalence of obesity and overweight among Iranian children and adolescents: A systematic review and meta-analysis. Nutrition. 2014;30:393–400.

Swallen

, Reither

, Haas

, Meier

. Overweight, obesity, and health-related quality of life among adolescents: The National Longitudinal Study of Adolescent Health. Pediatrics. 2005;115:340–7.

Murakami

, Livingstone

. Associations between meal and snack frequency and overweight and abdominal obesity in US children~and adolescents from National Health and Nutrition Examination Survey (NHANES) 2003-2012. Br J Nutr. 2016;115:1819–29.

Gungor

. Overweight and obesity in children and adolescents. J Clin Res Pediatr Endocrinol. 2014;6:129–43.

Dehghan

, Akhtar-Danesh

, Merchant

. Childhood obesity, prevalence and prevention. Nutr J. 2005;4:24.

Yuan

, Li

, Meng

, You

, Wu

, Li

, Chen

. Gender Difference on the Association between Dietary Patterns and Obesity in Chinese Middle-Aged and Elderly Populations. Nutrients. 2016 23;8. pii: E448.

Khayyatzadeh

, Moohebati

, Mazidi

, Avan

, Tayefi

, Parizadeh

, Ebrahimi

, Heidari-Bakavoli

, Azarpazhooh

, Esmaily

, et al. Nutrient patterns and their relationship to metabolic syndrome in Iranian adults. Eur J Clin Invest. 2016;46:840–52.

Medina-RemOn

, Kirwan

, Lamuela-Raventos

, Estruch

. Dietary Patterns and the Risk of Obesity, Type 2 Diabetes Mellitus, Cardiovascular Diseases, Asthma, and Mental Health Problems. Crit Rev Food Sci Nutr. 2018;58:262–96.

10.

Gnagnarella

, Maisonneuve

, Bellomi

, Rampinelli

, Bertolotti

, Spaggiari

, Palli

, Veronesi

. Nutrient intake and nutrient patterns and risk of lung cancer among heavy smokers: Results from the COSMOS screening study with annual low-dose CT. Eur J Epidemiol. 2013;28:503–11.

11.

De Stefani

, Boffetta

, Ronco

, Deneo-Pellegrini

, Acosta

, Gutiérrez

, Mendilaharsu

. Nutrient patterns and risk of lung cancer: A factor analysis in Uruguayan men. Lung cancer. 2008;61:283–91.

12.

Azadbakht

, Esmaillzadeh

. Dietary diversity score is related to obesity and abdominal adiposity among Iranian female youth. Public Health Nutr. 2011;14:62–9.

13.

Rezazadeh

, Rashidkhani

. The association of general and central obesity with major dietary patterns in adult women living in tehran, iran. ARYA Atheroscler. 2010;6:23–30.

14.

Kelishadi

, Ardalan

, Gheiratmand

, Gouya

, Razaghi

, Delavari

, Majdzadeh

, Heshmat

, Motaghian

, Barekati

, et al. Association of physical activity and dietary behaviours in relation to the body mass index in a national sample of Iranian children and adolescents: CASPIAN Study. Bull World Health Organ. 2007;85:19–26.

15.

Moskal

, Pisa

, Ferrari

, Byrnes

, Freisling

, Boutron-Ruault

, Cadeau

, Nailler

, Wendt

, Kühn

, et al. Nutrient patterns and their food sources in an International Study Setting: Report from the EPIC study. PloS one. 2014; 9:e98647.

16.

Kelishadi

, Heshmat

, Motlagh

, Majdzadeh

, Keramatian

, Qorbani

, Taslimi

, Aminaee

, Ardalan

, Poursafa

, et al. Methodology and Early Findings of the Third Survey of CASPIAN Study: A National School-based Surveillance of Students’ High Risk Behaviors. Int J Prev Med. 2012;3:394–401.

17.

Kelishadi

, Motlagh

, Bahreynian

, Gharavi

, Kabir

, Ardalan

, Safari

, Qorbani

. Methodology and Early Findings of the Assessment of Determinants of Weight Disorders among Iranian Children and Adolescents: The Childhood and Adolescence Surveillance and PreventIon of Adult Noncommunicable Disease-IV Study. Int J Prev Med. 2015;6:77.

18.

Kelishadi

, Majdzadeh

, Motlagh

, Heshmat

, Aminaee

, Ardalan

, Esmaillzadeh

, Azadbakht

, Poursafa

, Movahedian

, et al. Development and Evaluation of a Questionnaire for Assessment of Determinants of Weight Disorders among Children and Adolescents: The Caspian-IV Study. Int J Prev Med. 2012;3:699–705.

19.

Kim

, Mueller

. Factor analysis: Statistical methods and practical issues. Sage, Newbury Park. 1978.

20.

Mazidi

, Kengne

. Nutrient patterns and their relationship with general and central obesity in US adults. Eur J Clin Invest. 2017. [Epub ahead of print]

21.

Salehi-Abargouei

, Esmaillzadeh

, Azadbakht

, Keshteli

, Feizi

, Feinle-Bisset

, Adibi

. Nutrient patterns and their relation to general and abdominal obesity in Iranian adults: Findings from the SEPAHAN study. Eur J Nutr. 2016;55:505–18.

22.

Pisa

, Pedro

, Kahn

, Tollman

, Pettifor

, Norris

. Nutrient patterns and their association with socio-demographic, lifestyle factors and obesity risk in rural South African adolescents. Nutrients. 2015;7:3464–82.

23.

Ghassemi

, Harrison

, Mohammad

. An accelerated nutrition transition in Iran. Public Health Nutr. 2002;5:149–55.

24.

Rashidi

, Mohammadpour-Ahranjani

, Vafa

, Karandish

. Prevalence of obesity in Iran. Obes Rev. 2005;6:191–2.

25.

Bray

. Medical consequences of obesity. J Clin Endocrinol Metab. 2004;89:2583–9.

26.

Tonstad

, Nathan

, Oda

, Fraser

. Prevalence of hyperthyroidism according to type of vegetarian diet. Public Health Nutr. 2015;18:1482–7.

27.

Pisa

, Pedro

, Kahn

, Tollman

, Pettifor

, Norris

. Nutrient Patterns and Their Association with Socio-Demographic, Lifestyle Factors and Obesity Risk in Rural South African Adolescents. Nutrients. 2015;7(5).

28.

Freisling

, Fahey

, Moskal

, Ocke

, Ferrari

, Jenab

, Norat

, Naska

, Welch

, Navarro

, et al. Region-specific nutrient intake patterns exhibit a geographical gradient within and between European countries. J Nutr. 2010;140:1280–6.

29.

Bahreynian

, Esmaillzadeh

. Quantity and quality of carbohydrate intake in Iran: A target for nutritional intervention. Archives of Iranian Medicine. 2012;15:648–9.

30.

Wiseman

. The second World Cancer Research Fund/American Institute for Cancer Research expert report. Food, nutrition, physical activity, and the prevention of cancer: A global perspective. Proc Nutr Soc. 2008;67:253–6.

31.

Zhou

, Li

, Zhou

, Sun

, Liu

. B-vitamin consumption and the prevalence of diabetes and obesity among the US adults: Population based ecological study. BMC Public Health. 2010;10:746.

32.

, Sun

, Zhou

, Liu

, Zhou

, Luo

, Bian

, Zhao

, Guo

. Chronic niacin overload may be involved in the increased prevalence of obesity in US children. World J Gastroenterol. 2010;16:2378–87.

33.

Institute of Medicine Standing Committee on the Scientific Evaluation of Dietary Reference I. Dietary reference intakes for thiamin, riboflavin, niacin, vitamin B6, folate, vitamin B12, pantothenic acid, biotin, and choline: National Academies Press (US); 1998.

34.

Mojtabai

. Body mass index and serum folate in childbearing age women. Eur J Epidemiol. 2004;19:1029–36.

35.

McClung

, Karl

. Iron deficiency and obesity: The contribution of inflammation and diminished iron absorption. Nutr Rev. 2009;67:100–4.

36.

Slavin

. Dietary fiber and body weight. Nutrition. 2005;21:411–8.

37.

Rezazadeh

, Rashidkhani

. The association of general and central obesity with major dietary patterns of adult women living in Tehran, Iran. J Nutr Sci Vitaminol (Tokyo). 2010;56:132–8.

38.

Paradis

, Godin

, Perusse

, Vohl

. Associations between dietary patterns and obesity phenotypes. Int J Obes (Lond). 2009;33:1419–26.

39.

Spiteller

. Peroxyl radicals: Inductors of neurodegenerative and other inflammatory diseases. Their origin and how they transform cholesterol, phospholipids, plasmalogens, polyunsaturated fatty acids, sugars, and proteins into deleterious products. Free Radical Biology & Medicine. 2006;41(3):362–87.

40.

Zemel

. Role of calcium and dairy products in energy partitioning and weight management. Am J Clin Nutr. 2004;79:907s–12s.

41.

Buckley

, Howe

. Long-chain omega-3 polyunsaturated fatty acids may be beneficial for reducing obesity-a review. Nutrients. 2010;2:1212–30.

42.

Parra

, Ramel

, Bandarra

, Kiely

, Martinez

, Thorsdottir

. A diet rich in long chain omega-3 fatty acids modulates satiety in overweight and obese volunteers during weight loss. Appetite. 2008;51:676–80.

43.

Huang

, Xin

, McLennan

, Storlien

. Role of fat amount and type in ameliorating diet-induced obesity: Insights at the level of hypothalamic arcuate nucleus leptin receptor, neuropeptide Y and pro-opiomelanocortin mRNA expression. Diabetes Obes Metab. 2004;6:35–44.

44.

Whitehouse

, Tisdale

. Downregulation of ubiquitin-dependent proteolysis by eicosapentaenoic acid in acute starvation. Biochem Biophys Res Commun. 2001;285:598–602.

45.

Wing

, Goldberg

. Glucocorticoids activate the ATP-ubiquitin-dependent proteolytic system in skeletal muscle during fasting. Am J Physiol. 1993;264:E668–76.

46.

Moussavi

, Gavino

, Receveur

. Is obesity related to the type of dietary fatty acids? An ecological study. Public Health Nutr. 2008;11:1149–55.