Introduction

The metabolic syndrome (MetS), described by the clustering of cardiometabolic risk factors that include anthropometric and physiological parameters, appears to be caused through pathophysiological processes related to insulin resistance [1]. The prevalence and incidence of MetS has increased rapidly around the world [2, 3]. The prevalence of MetS in Iranian adolescents increased from 4.1% in 1990 to 7.6% in 2018 [4, 5]. In addition to higher presence of mental health disorders, non-alcoholic fatty liver, and obstructive sleep apnea in later childhood or adolescence [6], the evidence indicates that MetS and obesity in children and adolescents are strong predictors of future type 2 diabetes, insulin resistance, and cardiovascular diseases in adulthood [7, 8]. Although the etiology of the disease is not well known, dietary and lifestyle pattern are recognized as the leading modifiable risk factors that appear to have a substantial effect on its development [9]. From a nutritional point of view, various healthy dietary patterns and particularly some food groups, such as dairy products, may have a beneficial impact on MetS prevention in adults [10], but the evidence in children and adolescents is limited.

Dairy products have a desirable ratio of nutrient density to energy density [11] and supply protein and some micronutrients, especially calcium, in a regular diet. However, inclusion of dairy as part of a healthy diet and the effects of milk, yogurt, and cheese on health and longevity is controversial [12]. There are major differences in the components of individual dairy products that may alter their role in metabolic health. Diverse profiles of fatty acids and total content in each dairy food may influence the risk of MetS progression as well. Although dietary guidelines recommend to eat low or reduced fat dairy products, recent evidence casts doubt on the detrimental effect of dairy fat on metabolic health, indicating there is neutral or inverse association between full-fat dairy consumption and cardiometabolic health outcomes [13, 14]. The presence of lactic acid bacteria in fermented products such as yogurt and cheese, in addition to the content of specific saturated fatty acids (SFA), has distinct effects on cardiometabolic disease and insulin resistance [15, 16].

Contrary to the results of a cross-sectional study illustrating an inverse relationship between dairy consumption and abdominal obesity [17], Matthews et al. demonstrated that dairy consumption doubles the risk of obesity among children and adolescents [18]. A recent meta-analysis of prospective cohort studies reported a 23% reduction in the incidence of MetS with dairy product intakes in adults [19]. However, to our knowledge, no previous longitudinal study has examined the association of dairy products with the incidence of MetS in children and adolescents. Therefore, we aimed to investigate the association of total and individual dairy food consumption with the incidence of MetS among children and adolescents.

Materials and methods

Study population

We conducted the current study within the framework of the Tehran Lipid and Glucose Study (TLGS), which is an ongoing community-based prospective investigation aiming to evaluate the risk factors for non-communicable diseases. The baseline survey was a cross-sectional study (1999–2001), and participants are monitored every 3 years [20].

In the fourth examination survey of the TLGS (2009–2012) of 12,523 participants who had complete medical history and physical examination data, a representative sample of 7956 participants was randomly selected for dietary assessment. There were 829 children and adolescents aged 6–18 years with complete nutritional data.

Participants who had missing data on MetS components (n = 10) were excluded. Furthermore, participants with over- or underreporting energy intake, i.e., those in the top and bottom 1% of the energy intake to estimated energy requirement (EER) [21] ratio (n = 9), were also excluded. Depending on outcome variables under analysis, participants were also excluded if they presented that outcome at baseline: (MetS, n = 91; hypertension, n = 72; high triglycerides (TGs), n = 144; low high-density lipoprotein cholesterol (HDL-C), n = 84; high fasting plasma glucose (FPG), n = 89; or abdominal obesity, n = 189). Some participants fell into more than one category of exclusion; the final sample size varied by the following outcomes: MetS (n = 531), hypertension (n = 517), high TGs (n = 449), low HDL-C (n = 509), high FPG (n = 504), and abdominal obesity (n = 412). Data from the sixth survey of the TLGS (response rate: 82%) were used to evaluate the incidence of MetS and its components after a mean 6.6 years of follow-up.

Protocols of this study were approved by the institutional ethics committee of the Research Institute for Endocrine Sciences, affiliated to the Shahid Beheshti University of Medical Sciences, and written informed consent was obtained from participants' parents.

Measurements

Dietary measurements

A valid and reliable semi-quantitative food frequency questionnaire (FFQ) was applied to assess the habitual dietary intake of children and adolescents at the baseline [22,23,24]. The relative validity of the FFQ for measuring dairy intake has been previously reported, and it shows a reasonable correlation with estimates from dietary records. Trained dietitians asked participants during a face-to-face interview to determine their consumption frequency for each food item consumed during the previous year on a daily, weekly, or monthly basis. Mothers were asked about the type and quantity of meals and snacks when children had difficulties or were unable to recall. Portion sizes of consumed foods that were reported in household measures were then converted to grams. To assist with determining portion size, interviewers provided photographs of household portions of each food item.

The questionnaire items for dairy products included low-fat milk, whole milk, low-fat yogurt, high-fat yogurt, regular cheese, cream cheese, and ice cream. The standard serving sizes were 240 mL for low-fat milk and whole milk; 120 mL for yogurt and ice cream; 220 mL for low- or high-fat yogurt, and 30 mL for cream cheese and regular cheese. Furthermore, we summed low-fat milk (< 2% of total fat content) with low-fat yogurt (< 2% of total fat content) and labeled as low-fat dairy, and high-fat yogurt (≥ 3% of total fat content), high-fat milk (≥ 3% of total fat content), ice cream, regular cheese, and cream cheese labeled as high-fat dairy.

Measurements of covariates

Physical activity (PA) was estimated by the Modifiable Activity Questionnaire (MAQ), and metabolic equivalent task (MET) minutes per week was calculated. The Persian-translated MAQ in adolescents showed high reliability (97%) and moderate validity (49%) [25]. A low level of PA was considered as MET < 600 min/week.

Weight was recorded in light clothing to the nearest 0.1 kg on a SECA digital weighing scale (Seca 707; Seca Corporation, Hanover, Maryland; range 0.1–150 kg), and height was measured without shoes to the nearest 0.1 cm. Waist circumference (WC) was measured to the nearest 0.1 cm using a non-elastic tape over light clothing, midway between the iliac crest and the lowest rib in a standing position. Arterial blood pressure (BP) was measured with a suitable cuff size for each participant, after a 15-min rest in a supine position. Systolic- (SBP) and diastolic blood pressure (DBP) were measured, and the average of two measurements was recorded as the participants' BP.

A blood sample was drawn after 12–14-h overnight fast. Fasting plasma glucose was measured on the day of blood collection by the enzymatic colorimetric method, using glucose oxidase. Serum TGs concentration was measured by commercially available enzymatic reagents (Pars Azmoon, Tehran, Iran), adapted to a Selectra auto analyzer. HDL-C was measured after precipitation of the apolipoprotein B-containing lipoproteins with phosphotungstic acid and evaluated using Pars Azmoon kits. Inter- and intra-assay coefficients of variations at baseline were 2.2% for FPG, 2.0, and 0.5% for HDL-C and 1.6 and 0.6% for TGs, respectively.

Definitions

In children and adolescents, MetS was defined according to the definition proposed by Cook et al. [26] as three or more of the following: WC ≥ 90th percentile for age and sex, according to national reference curves [27]; SBP and DBP ≥ 90th percentile for sex, age, and height (the National Heart, Lung, and Blood Institute's recommended cut-points) [28], and FPG ≥ 100 mg/dl (American Diabetes Association) [29]. Fasting TGs ≥ 110 mg/dl; HDL-C < 40 mg/dl; In subjects aged > 18 years after follow-up, MetS was defined according to the joint interim statement [30] as the presence of any 3 of 5 risk factors following: (i) Enlarged WC (≥ 95 cm according to the newly introduced cut-off points for Iranian adults for both genders) [31]; (ii) FPG ≥ 100 mg/dl or use of hyperglycemic medication; (iii) Fasting TGs ≥ 150 mg/dl or use of antihypertriglyceridemia medication; (iv) Fasting HDL-C < 50 mg/dl for women and < 40 mg/dl for men or drug treatment; and (v) High BP was defined as SBP ≥ 130 mm Hg, DBP ≥ 85 mm Hg, or antihypertensive drug treatment.

Statistical analysis

All dietary exposures were adjusted for total energy using the residual method [32]. The Kolmogorov–Smirnov test and histogram were applied to determine the normality of the quantitative variables. Continuous variables were presented as the mean (standard deviation (SD)) or median (inter-quartile 25th and 75th), whereas categorical data were presented as percentages. Energy-adjusted dairy intake and its subgroups were categorized into tertiles. Linear regression and chi-square were used to test the trend of continuous and categorical variables, respectively, across the tertile of total dairy intakes.

We applied multivariable regression models to estimate the odds ratio (OR) and 95% confidence interval (CIs) across categories of total dairy consumption with the risk of incident MetS, abdominal obesity, high FPG, high TGs, low HDL-C, and hypertension. For all outcomes, the basic model was adjusted for age in years. Consequently, we further adjusted for family history of diabetes, PA, total energy intake, and sex. We additionally adjusted for some dietary covariates, including percent of energy from protein and carbohydrate along with whole-grain consumption. We modeled dairy exposures as categorical variables and tested for a trend by modeling the median value within each category as a continuous variable. The interaction terms between age (grouped as from 6 to 12 years and from 13 to 18 years), sex (girl, boy), and PA (low, moderate, and vigorous), and total dairy consumption on the incidence of MetS were examined. No interaction was found between considered covariates and total dairy intake on the outcome in the final models.

We estimated the effect on MetS risk of increasing the intake of one serving per day of nuts, legumes, whole grain, poultry and fish, egg, or red and processed meat while concomitantly decreasing the dairy consumption, using substitution modeling. Thus, the difference in regression coefficients was used to estimate ORs for the substitution and covariance to estimate the 95% CIs. Statistical significance was set at P < 0.05. All statistical analyses were performed using the SPSS software (version 15.0; SPSS Inc, Chicago, IL).

Results

The average age of the population was 12.3 (3.1) years; 49.2% were girls, and 81% had low physical activity. The average dairy consumption was 2.8 serving/day. Among the 531 participants without MetS at baseline, there were 52 (9.8%) cases of incident MetS over an average 6.6-year follow-up. Baseline clinical and dietary characteristics of participants are presented across tertile categories of energy-adjusted dairy consumption in Table 1. In trend analyses from the lowest to the highest tertile category of dairy consumption, and those in the highest category were more likely to be younger. With increasing total dairy intake, intake of whole grains and percent of energy from carbohydrate tended to be lower. In contrast, intake of percent of energy from fatty acids, SFA, protein as well as consumption of low- and high-fat dairy tended to be higher.

Table 1 Baseline characteristics of children and adolescents according to tertiles of total dairy consumption: Tehran Lipid and Glucose Study
Full size table

Risk of incident MetS in those with normal status at baseline, according to total dairy consumption, is presented in Table 2. In the basic model, adjusted for age, increasing total dairy consumption was associated with a 55% lower risk of incident MetS (lowest (reference) compared with the highest intake, OR: 0.45, 95% CI: 0.22–0.94; P-trend = 0.029). This association remained after adjusting for sex, total energy intake, physical activity, and family history of diabetes (OR: 0.48, 95% CI = 0.23–1.00, P-trend = 0.042), and even after additionally adjusting for dietary factors (OR: 0.41, 95% CI = 0.19–0.88, P-trend = 0.018). In the final model, a 1 serving/day increase in dairy consumption was associated with lower MetS risk (OR: 0.68, 95% CI = 0.51 –0.92). In the fully adjusted model, low-fat dairy intake was inversely associated with incident MetS (highest compared with the lowest tertile, OR: 0.44, 95% CI: 0.21–0.92; P-trend = 0.023). Furthermore, one serving/day increase in low-fat dairy consumption in a diet was marginally associated with decreased MetS risk (OR: 0.74, 95% CI = 0.54–1.02).

Table 2 Odds ratio (95% confidence intervals) of incident MetS according to tertiles of dairy consumption among children and adolescents
Full size table

We further evaluated associations of types of dairy products with risk of incident MetS (Table 3). After controlling for considered confounding factors, higher low-fat milk consumption was associated with a 54% lower risk of incident MetS (highest compared with lowest consumption, OR: 0.46, 95% CI: 0.22–0.98; P-trend = 0.050); and the higher low-fat yogurt intake was associated with a 65% lower risk of incident MetS (OR: 0.45, 95% CI: 0.21–0.97; P-trend = 0.031). Moderate but not the highest intake of regular cheese consumption had a decreased risk of MetS by 67% (second compared with first tertile consumption, OR: 0.43, 95% CI: 0.19–0.97). High-fat milk, high-fat yogurt, and ice cream intake were not associated with the risk of incident MetS.

Table 3 Multivariable-adjusted odds ratio (95% confidence intervals) of incident MetS risk according to the consumptions of dairy food subtypes among children and adolescents
Full size table

After adjusting for confounding factors, higher dairy consumption was associated with 51%, 63%, and 57% lower risk of incident abdominal obesity (OR: 0.49, 95% CI = 0.26–0.94, P-trend = 0.032), high BP (OR: 0.37, 95% CI = 0.17–0.81, P-trend = 0.008), high TGs (OR: 0.43, 95% CI = 0.20–0.96, P-trend = 0.039) as presented in Supplementary Table 1.

We estimated that substituting one serving/day of total dairy for nuts was associated with a 47% lower risk of MetS incidence (OR: 0.63, 95% CI: 0. 42–0. 95; Fig. 1). However, a substitution for red and processed meat was associated with a 54% higher risk of MetS (OR: 1.55, 95% CI: 1.21–1.97) (Fig. 1).

Fig. 1
figure 1

Odds ratios and 95% confidence intervals from substitution models for incidence of metabolic syndrome associated with increasing consumption of a food group and concomitantly decreasing consumption of total dairy by one serving per day. The model was adjusted for baseline age, sex, total energy intake, physical activity, family history of diabetes, dietary intakes of protein, carbohydrate, and whole grain.

Full size image

Discussion

In this community-based prospective study of children and adolescents with a 6.6-year follow-up period, we observed that consumption of total dairy products was related to lower incident MetS and some of its components, including abdominal obesity, high BP, and high TGs after adjustment for several diet and lifestyle covariates. Furthermore, specific dairy products, including low-fat dairy, low-fat milk, low-fat yogurt, and regular cheese, were associated with a lower risk of incident MetS. Decreasing intake of total dairy and simultaneously increasing nut intake by one serving/day was associated with a 47% lower risk of MetS, whereas substituting one serving/d of red and processed meat for total dairy consumption was associated with a 57% higher risk of MetS.

To our knowledge, the present study is the first to examine the association between total and individual dairy consumption and MetS risk in a longitudinal analysis in children and adolescents. We showed that a higher intake of total dairy decreased the odds of incident MetS in children and adolescents, whereas a cross-sectional study of Iranian adolescents found no association of dairy consumption with the risk of MetS and its component [33]. It should be noted that our results are in agreement with those of other prospective and cross-sectional studies indicating an inverse association of dairy consumption with the MetS risk in adults [14, 34, 35]. A meta-analysis among studies of adults reported a 15% (RR: 0.85; 95% CI: 0.73, 0.98) lower risk of MetS incidence for the highest versus lowest category of total dairy consumption [36]. We observed that by increasing total dairy consumption in a regular diet both low- and high-fat dairy consumption was concurrently increased. Accordingly, the percentage of energy from total fatty acids and SFA was also elevated. Notably, it remains controversial whether the intake of dairy fat is harmful to cardiometabolic health, and whether the intake of low-fat dairy provides advantages with regard to MetS risk relative high-fat dairy [37, 38]. Furthermore, we observed that participants who consumed more dairy had lower intakes of whole grain and percent of energy from carbohydrate, compared to those who consumed less. Thus, the association between dairy consumption and metabolic syndrome could be the relative reduction in carbohydrates rather than (or in combination with) the relative increase in dairy consumption; however, the association still remains significant when whole grain and percent of energy from carbohydrate were adjusted in the final model to attenuate its confounding effect.

In the present study, we determined that low-fat dairy consumption was inversely associated with the risk of the MetS, unlike high-fat dairy. This finding was further supported by the lower risk of MetS among those with dairy products consumption with low-fat content, like yogurt and milk [39, 40]. Although the beneficial effect of yogurt consumption on metabolic parameters has been reported in some studies, this finding is inconsistent [33, 41, 42]. One explanation could be the inability of studies to separate the low-fat yogurt from high-fat ones. In the current study, we observed that a higher intake of low-fat yogurt was associated with a 45% decreased risk of MetS; however, high-fat yogurt consumption had no significant association. The favorable association of low-fat dairy and MetS may be attributed to the higher amounts of protective constituents, such as calcium and whey protein instead of SFA content, in low-fat dairy [43].

Nevertheless, data on the risk–benefit of cheese intake is highly controversial. A positive association between cheese consumption and MetS prevalence was illustrated in one study [44], while higher consumption of cheese had an inverse association with the incidence of MetS in another [45]. In our study, moderate consumption of cheese was associated with a lower risk of MetS compared to lower consumption. Moderate intake of regular cheese may provide some advantages of dairy products family; however, high content of fat and salt in cheese may confound its beneficial effect on the metabolism [46]. Moreover, we did not have enough detail about the types of cheeses consumed in this cohort to assay MetS risk in relation to various cheese types. More research is needed to determine the underlying mechanism that may mediate the relation between cheese consumption and MetS risk. With regard to ice cream consumption, we did not observe any association with MetS risk among children and adolescents. These findings may be explained by the fact that ice cream contains high amounts of fat and sugar, which can eliminate the favorable effects of milk component [47]. Overall, our findings suggested that low-fat dairy products, like yogurt and milk, are associated with a lower risk of MetS. It should also be noted that consuming high-fat dairy did not raise risk of MetS.

Data from previous studies revealed the desirable association of dairy consumption with cardiometabolic risk factors such as overweight/obesity, high BP, and abdominal obesity among children and adolescents [17, 48, 49] and in adults [35]. In our longitudinal study, we found that participants in the highest tertiles of dairy consumption had a lower risk for the incidence of abdominal obesity, high BP, and high TGs. In a prospective study conducted in Australian children, as compared with those who were in the lowest category of dairy consumption, children in the highest dairy consumption had lower incident SBP and DBP risk [49]. Abreu et al. found that higher consumption of dairy was associated with a 78% reduction in abdominal obesity in boys after adjustment for confounders [17]. Likewise, a recent meta-analysis of prospective cohort comprising 46,011 children and adolescents showed that those in the highest category of dairy consumption had a 38% lower risk of overweight/obesity compared with subjects in the lowest consumption during a median follow-up of 3-years [48].

Several potential mechanisms for the beneficial effect of dairy consumption on MetS have been proposed. The nutrients found in dairy products, particularly calcium and protein, may exert a synergistic and protective role against metabolic dysregulation. Calcium inhibits fat absorption and increases fecal excretion in the intestine through binding to bile acids and fatty acids, especially SFA [50]. As a result, calcium intake can improve the HDL-C to LDL-C ratio by converting cholesterol to bile acids in the liver and decrease TGs levels [51, 52]. Moreover, dietary calcium, as shown in a meta-analysis of intervention studies [53], decreases BP through repression of 1,25-dihydroxy vitamin D3, resulting in normalization of intercellular calcium in vascular smooth muscle cells [54]. Peptide fractions of dairy protein may play a major role in preventing the MetS through regulating insulinemia, dyslipidemia, BP, and fat accumulation [55]. Dairy is a substantial source of two major forms of protein, casein and whey. These proteins inhibit the function of angiotensin-converting enzyme (ACE), which is known to increase BP and adipocyte lipogenesis by upregulated of fatty acid synthesis expression. Hence, the inhibition of ACE and consequently angiotensin II production by dairy-derived proteins can decrease BP and endogenous fat deposition [56, 57].

We performed substitution analyses to predict the impacts of replacing dairy products with other foods on the risk of MetS. We found that replacing dairy with red and processed meat was associated with a higher risk of MetS. Red and processed meat consumption may lead to an elevated risk of developing MetS by several pathways such as increased insulin resistance and the loss of β-cell function, obesity, hyperinsulinemia and hyperglycemia. These comorbidities are stemmed from the high content of fat and SFA, especially long-chain SFA compared to short- and medium-chain SFA in dairy, high levels of inflammatory mediators or through elevated free fatty acid and nitrosamines production [58,59,60]. A meta-analysis indicated that total, red, and processed meat consumption was associated with higher MetS risk [61]. In contrast to red meat, replacing total dairy with nuts was associated with a lower risk of MetS. These health effects may be explained by improving oxidative stress, inflammation, and endothelial function, which can reduce the risk of insulin resistance and diabetes, BP, abdominal obesity, and dyslipidemia as components of MetS [62]. These results suggest that replacing nuts for dairy products in a regular diet may provide further advantages to metabolic health than dairy consumption.

Our study has several strengths, including a longitudinal follow-up design and a high follow-up rate, which helps to evaluate the incidence of the MetS. We have also used a validated FFQ and national cutoff points for defining abdominal obesity, WC, and BP in children and adolescents.

The study also has limitations. Firstly, the information on the puberty status of our participants was not recorded, and we were not able to control the puberty stage in our analysis. Secondly, our cohort was made up of Tehranian urban participants; thus, our findings cannot be generalized to other populations. Thirdly, although habitual dietary intake of participants was assessed by a validated FFQ, potential measurement errors are unavoidable.

Conclusion

In conclusion, the results of the present prospective study suggest that higher consumption of dairy, particularly low-fat milk and yogurt, is related to a lower risk of incident MetS among children and adolescents. Furthermore, replacing total dairy with red and processed meat is associated with a higher MetS risk, whereas replacement by nuts is associated with a lower risk. Thus, it is imperative to invest in a strategy to improve the dietary pattern and consume healthy food groups, particularly low-fat dairy, to prevent cardio-metabolic risk factors in children and adolescents.