Nuts consumption and hypertension risks in children: a mediating role of circulating lipid metabolites

ABSTRACT

Background

Although nuts play an important role in preventing cardiovascular disease, the metabolic cues by which nuts regulate blood pressure have not been fully understood.Aims:We conducted a nested case–control study in a prospective cohort study of Southwest China children to explore the potential lipid metabolites related to the relationship between nut dietary and blood pressure.

Methods

Forty-three hypertension cases and 53 controls serum samples were obtained for lipidomic data analysis using a liquid chromatography mass spectrometry platform.

Results

We identified four lipid metabolites that are associated with nut intake by a generalized linear model and logistic regression analysis, including phosphatidylglycerol 43:6 [PG (43:6)], phosphatidylcholine 18:0/20:3 [PC (18:0/20:3)], and two phosphatidylethanolamine (PE) compounds [PE (P-16:0/20:4) and PE (P-22:0/18:2)]. Logistic regression analysis indicated that the levels of PG (43:6) and PE (P-16:0/20:4) were negatively associated with hypertension in children, which might be useful biomarkers for predicting childhood hypertension. Further mediation analysis revealed that PG (43:6) and PC (18:0/20:3) function as mediating variables between nut intake and blood pressure levels.

Conclusion

This study provides scientific evidence that nut consumption induces some beneficial changes in lipid metabolism, which may reduce the risk of hypertension in children.

KEYWORDS:

• Nuts
• hypertension
• children
• lipid metabolite
• diet

1. Introduction

Childhood essential hypertension is less prevalent than adult hypertension, causing people to pay less attention to it. Indeed, the prevalence of hypertension among children has increased dramatically over the past 20 years (1). Meanwhile, growing evidence indicated that higher blood pressure levels in children were associated with adulthood and lifelong cardiovascular events (1–4). Lifestyle factors have been well established as playing a significant role in the development of prehypertension and hypertension in children, in addition to genetic susceptibility (5–9). A healthier diet may ameliorate these adverse events.

Nuts, a nutrient-dense food, contain unsaturated fats, soluble fiber, vitamins, minerals, antioxidants, and phytosterols (10) that may improve lipid profiles, reduce oxidative stress and inflammation, ameliorate insulin resistance, and improve vascular reactivity (11). In recent years, evidence has shown that nuts are effective food for preventing hypertension (12–15). An intervention study found that participants taking 30 g of nuts per day for 12 weeks had a statistically significant decrease in systolic blood pressure (SBP) (16). Additionally, our previous cross-sectional study (17) demonstrated that intaking nuts between 50 and 100 g per day effectively controls childhood hypertension, but the underlying mechanisms have not yet been fully understood. Several studies have examined the relationship between nut consumption and traditional lipids indexes (18–20). A systematic review of 61 randomized controlled trialsnoted that the intake of nuts or nut products could decrease total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), and triglycerides (TGs), which suggests that circulatory lipid profiling can act as an intermediate variable between nut consumption and blood pressure control (21).

In addition to traditional lipid risk factors [LDL-C and very low-density lipoprotein cholesterol (VLDL-C), TGs], alterations in the composition of other circulatory lipid classes may be associated with hypertension (22). For example, α‐linolenic acid (ALA) has been considered a reliable blood biomarker of nut consumption since walnuts have a high ALA content in comparison to other foods (23). A population-based study of 990 adolescents identified several novel glycerophosphocholine metabolites using mass spectrometry (MS) analysis that was associated with multiple cardiovascular disease (CVD) risk factors (24). Significantly, phosphatidylcholine (PC) 16:0/2:0 was negatively associated with blood pressure level. However, it is not clear whether these metabolites are beneficial for the cardiovascular system of children.

To clarify the role of the lipid metabolites on hypertension risk in children, we conducted a nested case–control study of hypertension in children and adolescents from the Chongqing Children’s Health Cohort. Based on the diet information collected previously, we performed association analyses between nut intake and differences in lipid metabolite levels. We focused specifically on the impact of nut-related lipid metabolites on blood pressure levels.

2. Methods

2.1. Study participants and design

A case–control study nested within the Chongqing Children’s Health Cohort was used to investigate the impact of nuts on hypertension in children and adolescents. The Chongqing Children’s Health Cohort was established in 2014, and details about the study design have been published previously (17,25,26). In brief, participants were recruited from the following two populations: students without serious diseases aged 6–12 years were recruited from urban and rural areas in Chongqing, China. The recruiting criteria were as follows: (1) aged 10 and 13 years old; (2) resided in the target region for more than 6 months; and (3) did not have serious diseases (e.g., nephropathy, CVD , or cancer); and written informed consent was obtained from participants and their parents/guardians. The exclusion criteria are as follows: (1) incomplete information of age, gender, and nut intake dosage; (2) any diseases or taking any medications that may influence blood pressure or lipid metabolism. A formula of $n_A=\kappa n_B\mathrm{a}\mathrm{n}\mathrm{d}{n}_B=\left(1+\hspace{0.17em}\frac{1}{\kappa }\right){\left(\sigma \frac{{\mathcal{Z}}_{1-\alpha /2}+{\mathcal{Z}}_{1-\beta }}{{\mu }_A-{\mu }_B}\right)}^2$ was used to calculate the sample size. We assumed an α level of 0.10, under two-sided hypothesis testing, and β error of 0.20 with a difference of lipid metabolism of 12 between the case and control groups and a standard deviation of 8. To conduct this study, 38 hypertension participants and 38 control participants were needed. Following the exclusion of participants with no nut intake information, 96 participants (53 in the control group and 43 in the case group) completed a detailed questionnaire regarding their nut intake and the key variables used for this research. All the work was conducted under the ethical guidelines of the 1964 Declaration of Helsinki and its later amendments. Ethical review and approval were obtained by the Institutional Review Board of Chongqing Medical University, and informed consent was obtained from the parents or legal guardians of all the enrolled participants.

2.2. Data collection and diagnostic criteria

The demographic data consisting of age, gender, living environment (urban or rural), and location were collected by trained interviewers through a structured questionnaire. Blood pressure, height, and weight were measured by well-trained pediatric nurses, and body mass index (BMI) was calculated by using the formula BMI = weight/(height × height) (kg/m²). The protocol for these measurements has been published previously (27). In this study, hypertension was defined as three time occasions (28,29): mean SBP and/or diastolic blood pressure (DBP) ≥95th percentile based on the age-special, sex-special; and height-special percentiles. Those who met the diagnostic criteria for hypertension were diagnosed with hypertension in the case group. The control group included children with normal blood pressure. The mean arterial pressure (MAP, mmHg) was calculated as MAP = DBP (mmHg) + 1/3 (SBP (mmHg) – DBP (mmHg).

2.3. Nut intake information

Nut intake and the dietary information were collected using a quantitative food frequency questionnaire, which has been described in detail in a previous publication (30). In brief, we investigated the amount of food consumed in each category by interviewing the parents regarding the amount of food consumed and the frequency of intake during a measurable time interval. Frequency was represented by the following numbers: 0 = not eating, 1 = days, 2 = weeks, 3 = months, and 4 = years. Children’s daily, weekly, monthly, and annual consumption of nuts were indicated using the corresponding amount. To calculate the average daily intake, the total amount of food was divided by the number of days. The questionnaire also contained salt intake habit. Taste degree was considered as 5 levels (1 = very salty, 2 = salty, 3 = moderate, 4 = light, 5 = very light).

2.4. Biochemical index

Venous blood samples were taken in the morning after an overnight fast and after a supine rest of at least 15 minutes. Biochemical indexes and glycosylated hemoglobin were measured within 2 h after venous blood was drawn. A fully automatic biochemical analyzer（Mindray BS-800）was applied to test the serum biochemical parameters, including TC, LDL-C, HDL-C, TGs, fasting blood glucose (FBG), insulin, and creatinine.

2.5. Lipomics measurement and analysis

The sample preparation and the metabolites extraction for measurement were followed as reported previously (31). Liquid chromatography–tandem MS (LC–MS/MS) was performed using an UHPLC system (1290, Agilent Technologies), equipped with a Kinetex C18 column (2.1 × 100 mm, 1.7 μm, Phenomen) coupled to a quadruple time-of-flight mass spectrometer (Triple TOF 6600, AB SCIEX). After LC–MS/MS acquisition, raw spectra were converted to mzXML files using MSconvert (Proteo Wizard). Raw data obtained were pre-processed using the XCMS peak finding algorithm. The CentWave algorithm was used for peak detection with the MS/MS spectrum, and lipid identification was achieved through a spectral match using LipidBlast library (32). Absolute quantification was calculated by normalizing the peak area of each metabolite with respect to the area of the internal standards (IS) and by using standard curves. The data were quality-checked, filtered, and denoised. Missing values were substituted using a simulation-based statistical technique.

2.6. Statistical analysis

Following implementation of all the above quality control steps, we extracted 822 peaks and 709 metabolites from serum. Multivariate statistics including normalizing the peak area were performed using the SIMCA software package (version 16.0.2, Umetrics, Umea, Sweden). Principle component analysis (PCA) was performed to display the distribution of samples. The outliers were determined using the PCA plot (at 95% confidence interval (CI)) and removed from further analyses. The orthogonal projection of latent structure-discriminant analysis (OPLS-DA) was used to screen the significance metabolites. Then a sevenfold cross-validation was performed to calculate the values of R² and Q², where R² represents the goodness of fit and Q² represents the goodness of prediction. The identification of differentially expressed metabolites was performed by the variable importance in projection (VIP) values (VIP >1) of OPLS-DA combined with the Student’s t-test (t-test) (P ≤ .05). Q value was used for data filtering to measure significant differences between groups.

Normally distributed continuous variables were expressed as mean (SD) and compared using t-test. For data not meeting the normal distribution, equivalent non-parametric tests were applied. Categorical variables are presented as percentages and compared by the chi-squared test. Linear regression was used to examine the associations between lipid metabolites and nut intake with the adjustment of age, sex, height, and salt intake habit. Regression coefficient (β) and standard error (SE) were calculated. The logistic regression models were used to calculate the coefficient, Odds ratios (OR) and 95% CIs for the potential influencing factors of hypertension, including lipid metabolites. In addition, mediation models were performed to examine the mediation effect of lipid metabolites on the association of nut intake with blood pressure levels. The bootstrapping method with 5000 iterations was used to examine the significance of the direct effect, indirect effect, and total effect. The total effect was calculated as the indirect effect (β_Ind) plus the direct effect. The bias-corrected 95% CIs were used to evaluate the effects. Supposing the CIs didn’t include the value 0, the indirect effect indicated significant existence.

All analyses were performed using SAS version 9.4 (SAS Institute Inc), and statistical significance was defined as a 2-sided P value < 0.1.

3. Results

3.1. Basic characteristics of the studied sample

This study included 96 participants who filled out the dietary information questionnaires. The demographic and clinical characteristics of cases and controls are listed in Table 1. The case group consisted of 43 children [16 males and 27 females, mean age 11.71 ± 0.77 years old (mean ± SD)] and 20 participants in the case group were from urban areas. A statistically significant difference between the two groups was found in height, SBP, DBP, and MAP (all P < .001). Analyses of salt intake habit, BMI, and weight did not show differences between groups. The mean TC, LDL-C, HDL-C, TG, FBG, insulin, and creatinine levels were not significantly different between the normal BP group and the hypertension group.

Table 1. Characteristics of participants with nut data.

Variables	Control	Hypertension	P
Sample size (%)	53 (57.3)	43 (42.7)
Age, years	11.71 ± 0.53	11.71 ± 0.77	0.988
Sex, male (%)	29 (54.7)	16 (37.2)	0.133
Region, urban (%)	34 (63.0)	20 (37.0)	0.127
Salt intake habit, n (%)
Very high salt intake	2 (50.0)	2 (50.0)	0.729
High salt intake	6 (46.2)	7 (53.8)
Moderately salt intake	33 (54.1)	28 (45.9)
Low salt intake	10 (71.4)	4 (28.6)
Very low salt intake	2 (50.0)	2 (50.0)
Height, cm	150.25 ± 9.33	154.48 ± 7.49	0.018
Weight, Kg	48.76 ± 11.55	51.92 ± 12.28	0.198
BMI, Kg/m^2	21.88 ± 5.65	21.69 ± 4.36	0.862
SBP, mmHg	100.79 ± 5.91	124.13 ± 6.80	<0.001
DBP, mmHg	58.32 ± 5.05	75.18 ± 5.19	<0.001
MAP, mmHg	72.47 ± 4.65	91.50 ± 3.99	<0.001
TG, mmol/L	1.10 ± 0.47	1.27 ± 0.75	0.189
TC，mmol/L	3.61 ± 0.42	3.58 ± 0.52	0.704
HDL-C, mmol/L	1.33 ± 0.30	1.40 ± 0.33	0.307
LDL-C, mmol/L	1.94 ± 0.30	1.92 ± 0.39	0.859
FBG, mmol/L	4.46 ± 0.47	4.44 ± 0.42	0.84
IR	3.05 ± 3.34	2.68 ± 2.82	0.564
Insulin, pmol/L	102.19 ± 98.07	91.53 ± 88.35	0.581
Creatinine, μmol/L	52.96 ± 11.29	52.64 ± 6.41	0.871
UA, μmol/L	320.07 ± 72.59	338.99 ± 83.28	0.238

BMI: body mass index, SBP: systolic blood pressure, DBP: diastolic blood pressure, MBP: mean arterial pressure, TG: triglyceride, TC: total cholesterol, HDL-C, high-density lipoprotein cholesterol, LDL-C: low-density lipoprotein cholesterol, FBG: fasting blood glucose, IR: insulin resistance index, UA: uric acid.

3.2. Nut intake and serum lipid alterations

Our analysis identified four lipid species that were related to nut consumption. Table 2 shows the correlation coefficient between nut intake and serum lipidomics. Specifically, phosphatidylglycerol (PG) (43:6), PC (18:0/20:3), and phosphatidylethanolamine (PE) (P-22:0/18:2) were positively correlated with nut (all P < .1) with and without adjustment for age, sex, height, and salt intake habit. PE (P-16:0/20:4) was not significantly associated with the nut in the cured model, but after adjusting additional covariates, this association was stronger (P = .092).

Table 2. Linear regression analysis of the relationship between nuts and different lipid metabolites.

Dependent Variables	β (95% CI)	SE	P
PG (43:6)
Model 1	0.048 (0.005,0.092)	0.022	0.032
Model 2	0.043 (0.000,0.087)	0.022	0.052
PE (P-16:0/20:4)
Model 1	4.613 (−1.331,10.556)	3.033	0.132
Model 2	5.178 (−0.788,11.144)	3.044	0.092
PC (18:0/20:3)
Model 1	0.120 (0.006,0.233)	2.061	0.042
Model 2	0.124 (0.008,0.240)	0.059	0.040
PE (P-22:0/18:2)
Model 1	0.072 (−0.006,0.151)	0.04	0.073
Model 2	0.082 (0.003,0.161)	0.04	0.045

β: regression coefficient, SE: standard error, CI: confidence interval, PG: phosphatidylglycero, PE: phosphatidylethanolamine, PC: phosphatidylcholine. Model 1: The crude model. Model 2: Adjusted age, sex, height, and salt intake habit.

3.3. The relationship between lipid metabolites levels and hypertension in children

To determine whether these lipid species had effect on hypertension, we performed a logistic regression analysis (Table 3). The results showed that two lipid species were associated with hypertension. PG (43:6) was a protective factor against children’s hypertension in the adjusted model (OR 0.945; 95% CI: 0.903–0.989, P = .016), as did PE (P-16:0/20:4) (OR 0.966; 95% CI:0.934–0.994, P = .039). However, no significant correlation was observed between hypertension and the other two lipids.

Table 3. Logistic regression analysis of the association between lipids and hypertension.

Variables	β	SE	OR (95% CI)	P
PG (43:6)
Model 1	−0.063	0.022	0.939 (0.900,0.980)	0.004
Model 2	−0.056	0.023	0.945 (0.903,0.989)	0.016
PE (P-16:0/20:4)
Model 1	0.000	0.000	1.000 (0.999,1.000)	0.018
Model 2	0.000	0.000	0.966 (0.934,0.994)	0.039
PC (18:0/20:3)
Model 1	−0.009	0.007	0.991 (0.977,1.005)	0.193
Model 2	−0.012	0.008	0.988 (0.973,1.003)	0.112
PE (P-22:0/18:2)
Model 1	−0.011	0.010	0.989 (0.969,1.009)	0.277
Model 2	−0.013	0.011	0.987 (0.965,1.009)	0.229

β: regression coefficient, SE: standard error, CI: confidence interval, PG: phosphatidylglycero, PE: phosphatidylethanolamine, PC: phosphatidylcholine. Model 1: The crude model. Model 2: Adjusted age, sex, height, and salt intake habit.

3.4. The mediation effects of lipid metabolites

Mediation analysis indicated that PG (43:6) had significant mediation effects on the association between nut intake and blood pressure, as shown in Table 4. It indicated that the major effect of PG (43:6) was negatively regulating blood pressure levels. Additionally, an increase in mediator PC (18:0/20:3) might have a negative effect on SBP (β = −0.011; 95% CI: −0.035 to −0.001). Also, other lipids were investigated as possible mediators. It was found that neither PE (P-16:0/20:4) nor PE (P-22:0/18:2) had significant mediation effect on the association between nuts and lower blood pressure levels. Pathway analysis was performed using only PG (43:6) and PC (18:0/20:3). The direct, indirect, and total effects of mediation model were presented in Table 5. Consumption of nuts might have a total impact on SBP (β = −0.047; 95% CI: −0.097 to 0.002; P = .064) but no direct effect. These results indicated that the regulation of blood pressure would require a mediator. The mediation effect of PG (43:6) was still significant when PC (18:0/20:3) was simultaneously entered into the path as a mediator but no significant indirect effect through PC (18:0/20:3) emerged. A path model of mediation effects of PG (43:6) and PC (18:0/20:3) was displayed in Figure 1(Figure 1a and Figure 1b).

Figure 1. Mediation effect of PG (43:6) and PC (18:0/20:3) on the association between nut intake and blood pressure in crude model. (a) Mediation effect of PG (43:6) and PC (18:0/20:3) on the association between nut intake and systolic blood pressure (SBP). (b) Mediation effect of PG (43:6) on the association between nut intake and diastolic blood pressure (DBP) and mean arterial pressure (MAP).

Table 4. Mediation effect of lipid metabolites on nuts intake and blood pressure.

	SBP		DBP		MBP
Mediators	βInd (95% CI)	P	βInd (95% CI)	P	βInd (95% CI)	P
PG (43:6)	−0.023 (−0.055,−0.006)	0.049	−0.014 (−0.037,−0.004)	0.070	−0.014 (−0.043,−0.005)	0.057
PE (P-16:0/20:4)	−0.011 (−0.143,0.020)	0.402	−0.006 (−0.129,0.016)	0.440	−0.008 (−0.138,0.018)	0.416
PC (18:0/20:3)	−0.011 (−0.035,−0.001)	0.245	−0.007 (−0.030,0.001)	0.353	−0.008 (−0.031,0.002)	0.301
PE (P-22:0/18:2)	−0.004 (−0.021,0.003)	0.451	−0.004 (−0.017,0.003)	0.436	−0.004 (−0.018,0.002)	0.423

SBP: systolic blood pressure, DBP: diastolic blood pressure, MBP: mean arterial pressure, β_Ind: the indirect effect, CI = confidence interval, PG: phosphatidylglycero, PE: phosphatidylethanolamine, PC: phosphatidylcholine.

Table 5. Indirect and direct effects of nut intake on blood pressure levels.

Path	Indirect effect			Direct effect			Total effect
	β (95% CI)	Boot SE	P	β (95% CI)	Boot SE	P	β (95% CI)	Boot SE	P
Nut intake → PG (43:6) → SBP	−0.021(−0.053,−0.007)	0.011	0.050	−0.020(−0.067,0.027)	0.024	0.303	-0.047(−0.097,0.002)	0.025	0.064
Nut intake → PC(18:0/20:3) → SBP	−0.006(−0.027,0.003)	0.007	0.388	−0.020(−0.067,0.027)	0.024	0.303	-0.047(−0.097,0.002)	0.025	0.064
Nut intake → PG (43:6) → DBP	−0.014(−0.037,-0.004)	0.008	0.070	0.002(−0.034,0.039)	0.019	0.898	−0.011(−0.049,0.026)	0.019	0.554
Nut intake → PG (43:6) → MAP	−0.017(−0.042,-0.005)	0.009	0.057	−0.007(−0.044,0.031)	0.019	0.732	−0.023(−0.063,0.016)	0.020	0.250

β: regression coefficients; Bootstrap sample size = 5000, SE: standard error. PG: phosphatidylglycero, PC: phosphatidylcholine, SBP: systolic blood pressure,DBP: diastolic blood pressure, MBP: mean arterial pressure.

4. Discussion

In the present study, we investigated the effects of nut-related lipids on blood pressure levels. A novel lipid species PG (43:6) was identified, which has never been studied previously and may serve as a mediator between nuts intake and blood pressure levels. Meanwhile, multivariable analysis confirmed that PG (43:6) and PE (P-16:0/20:4) played protective roles in childhood hypertension. These results could be helpful in uncovering the biological mechanisms of nut’s protective effect on CVD.

Previous findings (33–36) have also reported that lipids might play essential roles in the association between dietary nut intake and CVD risk. For example, recent evidence indicated that 19 walnut-related metabolite profiles, including lipids, purines, acylcarnitines, and amino acids, were inversely associated with T2D and CVD risks (23). The results are generally consistent with our study, which indicate nuts are sources of several types of beneficial lipids and may reduce CVD risk. However, all the evidence was from adulthood, with limited control for potential confounders of lipid metabolism. Our study added evidence from children and adolescents. In certain instances, children’s lipid levels and blood pressure are influenced to a greater extent by non-medication-related factors, such as lifestyle or dietary aspects. Moreover, numerous trials have been conducted to elucidate that daily nut consumption may reduce TC, apolipoprotein B, LDL-cholesterol, and TGs, which are the key lipid factors in the present cardiovascular risk management guideline (37). However, unexpectedly, when nuts were included as part of cardio-protective dietary patterns, we found some lipids more relevant to childhood hypertension, such as PG (43:6) and PE (P-16:0/20:4). Future studies should incorporate a more granular measurement of blood lipids to estimate CVD risks.

Several mechanisms have been proposed for explaining why nuts affect the lipid profile positively, including reducing ApoB and increasing ApoA1; increasing the activity of enzymes involved in reducing lipid oxidation; and enhancing plasma adiponectin (a hormone that promotes fatty acid metabolism and reduces oxidative stress) (38). Toledo et al. (39) found that triacylglycerols, PCs, and lysophosphatidylethanolamines were shown to be differently associated with CVD, depending on the structure of their acyl chains, since the shorter and more saturated the acyl chains, the greater the risk. Consistent with the previous reports, a higher number of double bonds in lipids (i.e., long-chain PUFA (polyunsaturated fatty acids)s), especially −3 PUFAs, was associated with improved endothelial function and decreased myocardial oxygen requirement (40,41). Previous findings also suggested that nut intake may improve BP by phospholipids (PLs) supplementation (42). Dietary PLs from nuts contribute to cellular functions, including signaling and transport, as well as the activity of membrane bound enzymes (43). However, evidence from a prospective cohort study on the relationships between nut intake and lipid metabolites with childhood BP were quite limited. For the first time, our study found that the effect of nuts intake on blood pressure is beginning at childhood.

There are two strengths in our study. The first strength lies in its longitudinal design. We used a nested case–control study to minimize the potential effect of these confounding factors and obtain reliable results. In addition, we performed measurements and analyses of lipid metabolites to examine the effect that nuts may induce the change of lipids, and the lipids would impact blood pressure levels in children, which contributed to elucidate the underlying mechanisms linking nuts to lipid metabolism and hypertension. A further strength of the study was the fact that on three occasions blood pressure measurements were made to identify the hypertensive cases, which may reduce the number of false positive cases. Our study also has some limitations. First, as our sample size is relatively small, the power to detect the associations of nuts, lipids, and BP is limited after adjustment for multiple testing. Second, our findings were based on Chinese primary students without an available replication cohort to validate our lipidomic findings. Although, to our knowledge, these results are the first to link PG (43:6) to nut protective effect on BP, specific metabolic pathways involved in such a process, which have yet to be elucidated. It is essential to address these interactions and its regulatory mechanism in future study through a well-designed large sample size study.

5. Conclusion

This study provides scientific evidence that nut consumption induces some beneficial changes in lipid metabolism, which may reduce the risk of hypertension in children. Moreover, our results suggest that moderate nut intake can protect blood pressure levels in children and adolescents. Further study is needed to validate our results and to elucidate the pathways associated with nut-related lipids that regulate blood pressure levels, especially to find which kinds of nuts and which compounds in nuts regulate blood pressure levels through regulating lipid levels.

Author contributions

QL: Conceptualization, Methodology, Writing – Review & Editing; YL, LC and YF: Investigation, Supervision; SQ, PZ and YF: Validation, Resources; c, QL and YL: Writing – Original Draft, Writing – Review & Editing; XH: Writing – Review & Editing.

Acknowledgments

The authors would like to acknowledge all of the children and the staffs of the six elementary schools in the two counties.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data are available from the corresponding author Xiaohua Liang (Email: [email protected], or [email protected]).

Additional information

Funding

This work was supported by the Program for Youth Innovation in Future Medicine from Chongqing Medical University (No.W0088), General Project of Clinical Medical Research from National Clinical Research Center for Child Health and Disorders (No. NCRCCHD-2022-GP-01), the Young and Middle-aged Medical Outstanding Expert Project of Chongqing Municipal Health Commission (No. 78), Major Health Project of Chongqing Science and Technology Bureau (No. CSTC2021jscx-gksb-N0001), National key research and development project (No.2017YFC0211705), Intelligent Medicine Project (No.ZHYX202109), the Basic Research Project of Key Laboratory of Ministry of Education of China in 2021 (GBRP-202106), the Natural Science Foundation of Youth Project (No.81502826). The funders had no role in the whole study research process, including study design, the data collection and analysis, the decision to publish, or the preparation of the manuscript.