The association between sugars sweetened beverages and glycemic profile among children and youth: A systematic review and dose-response meta-analysis

Abstract

Background

This study aimed to assess the effects of Sugar-Sweetened Beverages (SSBs) consumption on fasting serum insulin (FSI), fasting plasma glucose (FPG), and HOMA-IR levels among children.

Methods

Databases including PubMed/MEDLINE, Scopus, EMBASE, Cochrane Library, and Web of Science were searched up to March 2025. Observational studies reporting the connection of SSBs consumption with FPG, FSI, and HOMA-IR levels were included. STATA version 15 was used to analyze the data.

Results

11 studies with 22,713 subjects were included in this meta-analysis. Greater intake of SSBs was not significantly linked to higher fasting plasma glucose (WMD: 0.01; CI -0.04 –0.07; P = 0.63) and fasting serum insulin levels (WMD: 0.54; 95 % CI, −0.4, 1.49; P = 0.26). However, high SSBs consumption was significantly associated with a 0.21 increase in HOMA-IR in adolescents and children (WMD: 0.21; CI, 0.03–0.37; P = 0.02). In dose-response meta-analysis, no departure from linearity was detected between SSBs intake and changes in FPG, FSI, and HOMA-IR levels.

Conclusions

High SSBs intake was linked to increased HOMA-IR levels among adolescents and children. Further extensive prospective long-term interventions are suggested to confirm the detected associations.

Highlights

• •Greater intake of SSBs was not significantly linked to higher fasting plasma glucose.
• •Greater intake of SSBs was not significantly linked to higher fasting serum insulin levels.
• •High SSBs consumption was significantly associated with increase in HOMA-IR in adolescents and children.
• •No departure from linearity was detected between SSBs intake and changes in FPG, FSI, and HOMA-IR levels.

1. Introduction

Prevalence of cardiometabolic abnormalities, obesity, hyperglycemia, elevated blood pressure, and dyslipidemia is rising in adolescents and children [[1], [2], [3]]. The role of sugars as an essential agent in developing cardiometabolic illnesses is actively debated [4,5]. A review summarizing 17 prospective studies stated that too much intake of sugar-sweetened beverages (SSBs) is associated with elevated risks of cardiometabolic abnormalities [6]. Sugar, especially when in the form of SSBs or free sugar, can result in heavy energy intake and obesity [7]. Consuming too much sugar can also be used to indicate a low-quality diet [8]. Numerous articles have stated a direct relationship between intake of SSBs and metabolic abnormalities in children and adults, whereas other studies have not [[9], [10], [11], [12]]. Excessive glucose intake as SSBs or other sources is recognized to affect developing glucose metabolism negatively [13,14]. A cross-sectional study stated that the HOMA-IR levels were 0.52 higher among adolescents consuming >350 mL/day of SSBs compared to those not consuming SSBs [15]. Fructose ingestion increases visceral fat accumulation by inducing hepatic de novo lipogenesis [16], which is associated with insulin resistance [17]. Insulin resistance is an impairment in insulin activity; however, its association with SSB intake in adolescents remains unclarified [13].

A precise approach about the association between children's consumption of sugar-sweetened beverages (SSBs) and the risk of glucose impairment is necessary for policymakers and physicians.

Due to the contradictory data currently available, the lack of a systematic review and dose-response meta-analysis in this field, this study was conducted to evaluate the association between SSB intake and the risk of glucose impairment in children.

2. Material and methods

This study looked at the dose-response association between children's use of sugar-sweetened beverages (SSBs) and their risk of glucose impairment.

The primary outcome of the current research was increased levels of fasting plasma glucose (mmol/L), fasting serum insulin (mU/L), and HOMA-IR in children. The reporting procedures were in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) (Table S1) [18].

2.1. Data source and search strategy

A methodical search was conducted using Scopus, Cochrane Library, PubMed, Embase, and Web of Science until to March 2025. There was no date constraint applied to the search. The reference sections of the articles were examined in order to improve the reliability of identifying the eligible research.

The Health Care Management Information Consortium (HMIC) and the European Association for Gray Literature Exploitation (EAGLE) databases were searched for Gray literature. A literature search was conducted by two reviewers (L.N. and M.A.F.), while two more reviewers (L.J. and Z.N.) looked at papers that were considered unclear.

In case of discrepancy, a discussion approach was used to reach a consensus. Only articles published in English were searched. Major medical subject heading (MESH) or title and abstract with these keywords were used to implement the search: SSB OR Sugar? Sweetened Beverage∗ AND childhood OR children OR Child OR adolescent∗ OR youth OR teenager OR pe?diatric∗ AND “fasting blood sugar” OR “fasting blood glucose” OR “fasting serum Insulin” OR “glucose impairment” OR FBS OR FSG OR Insulin OR HOMA-IR OR HbA1C OR FPG OR FG OR FSI. Table S2 displays the search technique and the number of articles classified by databases.

2.2. Inclusion and exclusion criteria

If an observational study met one of the following criteria, it was deemed appropriate: (a) it reported results across three categories of SSBs in order to conduct a dose-response meta-analysis of continuous variables; (b) it reported results for at least two categories of SSBs in order to conduct a two-class meta-analysis; (c) it used the mean and standard deviation (SD) of continuous variables to report the outcome of interest; or (d) it included children aged two to eighteen. Studies with participants of different ages were not included. Furthermore, research with unsuitable designs, unrelated subjects, and studies that did not assess the correlation between the parameters under investigation as well as reviews, meta-analyses, systematic reviews, or conference papers were eliminated.

2.3. Study selection, data extraction strategy, and quality assessment

Two reviewers (M.A.F. and L.N.) extracted the data separately. Articles that satisfied the inclusion/exclusion criteria were found by screening the titles and abstracts. When there was disagreement, a consensus was reached through debate. Reviewers obtained fulltext publications and evaluated each record for relevancy on their own. The initial author's name, the year of publication, the nation, the methodological features (study design), the average age and/or age range of the children, the sex, the kind of SSBs and the primary outcomes were among the data that were gathered from the studies. The literature review's flow chart is displayed in Fig. 1.

Fig. 1.

Search and inclusion process flow diagram.

For both the cross-sectional and cohort studies, the Joanna Briggs Institute Critical Appraisal Checklist was used to assess the quality. L.N. and M.A.F., evaluated the studies' quality. Disputes were settled through debate, and a third researcher (Z.N.) would be consulted if the results were inconclusive.

2.4. Statistical analysis

The data were analyzed using STATA version 15 (STATA Corp, College Station, TX, USA), and statistical significance was determined by considering P values less than 0.05.

2.5. Two-class meta-analysis of continuous variables

For the synthesis (the two-class meta-analysis), the studies reporting more than or equal to two categories of SSBs with low and high intake levels were selected. The case and control groups were thought to have the lowest and maximum intakes of sugar-sweetened beverages, respectively. The unstandardized mean differences were calculated using the mean and standard deviation of the variables. In the highest versus lowest SSBs categories, they served as the impact size determined by the pooled estimate of weighted mean difference (WMD) with 95 % confidence intervals (CIs). The random-effects model was used since the heterogeneity values were large (i.e., greater than 50 %). Heterogeneity was assessed using the Q statistic and the I² statistic. For the Cochran's Q test, a clinically relevant heterogeneity significance level was defined as I²>50 % or P < 0.10. Subgroup analysis was also carried out for the SSBs measurement, divided by the study design, country, gender, and age in order to pinpoint potential sources of heterogeneity Publication bias was assessed quantitatively using an adjusted rank correlation based on the Begg, Mazumdar, and Egger test, and graphically represented using a funnel plot. All units were converted so that all measures would be the same.

The standard deviation was computed using the following formula when CIs were reported: SD = √n (upper limit-lower limit)/3.92. We assumed that the same number of subjects were assigned to each group in situations where the paper did not provide the number of participants in each SSB category. As separate investigations, we examined the Zheng et al. [19] study using two distinct sets of data from various years.

2.6. Dose-response meta-analysis of continuous variables

For the dose-response meta-analysis of continuous variables, only research that reported FPG, FSI, and HOMA-IR in at least three SSBs categories were taken into account. Every study's reference dose was considered as the lowest SSB concentration category. To find the possible nonlinear effects of study-specific factors and SSBs dosage (ml), a dose-response meta-analysis of outcome variables was conducted using fractional polynomial modeling [20]. The mean and SD of glucose, insulin, and HOMA-IR were reported in at least three SSBs categories in the studies that qualified for this analysis.

2.7. Search results and study characteristics

After applying the search approach to identify 1735 titles, the number of titles was decreased to 984 using the Endnote 7.2.1 literature manager program, after duplicates were eliminated. After the titles and abstracts were first evaluated, the total number of papers was reduced to 383. After a more thorough evaluation of the full-text papers using the inclusion and exclusion criteria, only 11 articles remained. The literature retrieval flow chart is displayed in Fig. 1.

2.8. Study characteristics

Table 1 summarizes the characteristics of included studies. In the present meta-analysis, 11 studies performed between 2009 and 2021, with 22,713 individuals were included. Of 11 studies reporting FPG, three reported that categories with higher SSB intake had higher FPG levels in comparison to lower intake categories [19,21,22].

Table 1.

Characteristics of included studies.

First author	Country	Journal/year	Setting	Design/gender	Num. (total-each category)	Age range (y)	Dietary assessment tool	SSB type/Type of report	Main Results	Adjustments
Bremer [23]	USA	Arch Pediatr Adolesc Med/2009	Home	Cross-sectional/ Both	6967	12–18	24-h recall was assessed on two separate days	Colas, sugar-sweetened fruit drinks or other SSBs/child reported	Each additional SSB serving equivalent consumed per day was associated with a 5 % increase in HOMA-IR.	–
Ambrosini [29]	Australia	Am J Clin Nutr/2013	Clinic	Cohort/Both	1433	14–17	FFQ	Carbonated (soft) drinks, cordials or squash (fruit drink concentrate), and fruit juice drinks (with the exclusion of 100 % juice)/child report	Serum glucose levels decreased significantly with higher SSB intakes. Insulin and HOMA-IR levels were not significantly different among SSB intake quartiles.	–
Lin [26]	Taiwan	International Journal of Obesity/2013	School	Cross-sectional/Both	2727	12–16	Food-frequency questionnaire	Soft drinks, fruit drinks/child reported	Serum glucose levels were not significantly different among SSB intake quartiles.	The study area, age, gender, physical activity, total calories, the intake of meat, seafood, fruit, fried food and a food with jelly/honey, as well as for alcohol drinking and cigarette smoking.
Zheng [27]	Denmark	European Journal of Clinical Nutrition/2014	School	Longitudinal study/Both	283	9–18	24 h recall	Regular soft drinks, fruit drinks, cordials, and sugar-sweetened sports drinks/child reported	Fasting glucose, fasting insulin, and HOMA-IR were not significantly different among SSB categories.	Adjusted for age at 15years, gender, BMI/WC/S4SF at 15years, SSB consumption at 9years, socioeconomic status, pubertal status, and physical activity at 15years. and change in energy intake from ages 9–15years
Mirmiran [24]	Iran	Nutrition & Metabolism/2015	Community	Longitudinal study/Both	424	6–18	FFQ	All kinds of sugar-sweetened carbonated soft drinks (SSSDs) and fruit juice drinks/parent-reported	Serum glucose levels increased significantly with higher SSB intakes.	Baseline age, sex, total energy intake, physical activity, and family history of diabetes Model, dietary fiber, tea and coffee, red and processed meat, fruit, and vegetable, body mass index
Loh [22]	Malaysia	Pediatr Obes/2016	School	Cross-sectional/Both	873	13	Questions	Carbonated drinks, sugar-sweetened fruit drinks, non-dairy beverages or tetra-packed drink/child reported	SSB intake was deleteriously associated with elevated FBG, insulin, and insulin resistance.	Physical activity level, body mass index, and dietary patterns
Lin [15]	Taiwan	The journal of pediatrics/2016	School	Cross-sectional/Both	1454	12–16	Food-frequency questionnaire	Soft drinks, fruit drinks/child reported	Adolescents who consumed a greater amount of SSBs were more likely to have elevated fasting serum insulin and HOMA1-IR.	Sample weight and the complex study design.
He [25]	China	J Atheroscler Thromb/2018	School	Cross-sectional/Both	2032	7–18	Questions	Carbonated drinks, juices, and sports and sweet tea beverages/child report	Fasting glucose levels were not significantly different among SSB categories.	Age, gender, physical activities, sleeping duration, sedentary behavior, and dietary information
Li [21]	China	Public Health Nutrition/ 2020	School	Cross-sectional/Both	5258	7–18	Valid questions	Energy drinks, milk-containing drinks, soda, fruit drinks with added sugar, and other sugar-added beverages/self-report and parent report	Fasting glucose levels were not significantly different among SSB categories.	Age, gender, dietary intake (intake and frequency of fruits, vegetables, meat, dairy products, high-energy foods, and fried foods) and physical activity (frequency of intensive physical activity, moderate physical activity and walking), as well as the family environment factors (parental educational level, their attitudes towards SSB and their SSB intake)
Zhu [28]	China	Pediatric Obesity/2020	School	Cross-sectional/Both	3958	6–17	24-h dietary record and FFQ	Sugar-sweetened sodas, juice beverages, lactobacillus beverages, milk beverages, tea beverages, coffee drink, and typical Southeast Asian milky tea/professional interviewer with child	Fasting glucose levels were not significantly different among SSB categories.	–
Wu [13]	Taiwan	International Journal of Obesity/ 2021	Community	Cross-sectional/Both	1454	12–16	Semi-quantitative food-frequency questionnaire	Soft drinks, fruit drinks/child reported	Higher SSB intake was associated with higher levels of HOMA1-IR. Fasting plasma glucose was not significantly different between SSB drinkers and nondrinkers.	Sample weight and complex study design.

FSI was reported in four studies. In one study, higher fasting serum insulin levels were reported in higher SSBs intake categories compared to lower intake categories [22]. In four of eight studies, HOMA-IR was higher among high SSBs consumers [13,[21], [22], [23]].

Several types of sugar-sweetened beverages, such as carbonated soft drinks, juices or cordials, energy drinks, sugar-sweetened sodas, milk beverages, lactobacillus beverages, fruit drinks with added sugar, tea beverages, coffee drinks, yogurt drinks, chocolate milk, and other sugar-added drinks, were asked about in the SSBs intake questionnaires. Participants ranged in age from 6 to 18 years old. All of the questions came from kids and teenagers who appeared to be in good health. Numerous settings, such as communities [13,24], schools [15,21,22,[25], [26], [27], [28]], and clinics [29] were used for the studies.

The places of the inquiries were China [21,25,28], Iran [24], Australia [29], Malaysia [22], Taiwan [15,26,28], and USA [23]. Most of the articles were cross-sectional [13,15,21,22,[24], [25], [26], [27], [28]]. One study had cohort [29] and two studies had longitudinal [23,24] design in which the cross-sectional baseline data was used.

2.9. Findings from the two-class meta-analysis of the comparison of FPG, FSI, and HOMA-IR between different categories of SSBs

2.9.1. Mean FPG

As presented in Fig. 2, high SSBs intake was associated with a 0.01 increase in FPG levels in adolescents and children (WMD: 0.01; CI -0.04 –0.07; P = 0.63), which was not significant. Studies that reported mean serum glucose had significant between-study heterogeneity (I² = 100.00 %; P < 0.001). We performed subgroup analysis to find the heterogeneity source, and the results are shown in Table S3. Subgrouping according to setting and assessment tools reduced the heterogeneity.

Fig. 2.

The forest plot of the weighted mean difference (WMD) of the effect of SSBs intake on FPG.

2.9.2. Mean FSI

Four studies were included in the meta-analysis of the relationship between serum insulin and SSBs consumption (Fig. 3). The results indicated that high SSBs intake was not significantly linked to higher serum insulin (WMD: 0.54; 95 % CI, −0.4, 1.49; P = 0.26) with significant heterogeneity among studies (I² = 74.80 %, P < 0.001). Subgroup analysis was also performed in order to determine the origins of heterogeneity (Table S4). Subgroup analysis revealed sources of variation related to age, study design, and evaluation instruments.

Fig. 3.

The forest plot of the weighted mean difference (WMD) of the effect of SSBs intake on FSI.

Subgroup analyses show that in studies conducted in schools (WMD (CI): 0.999 (0.288, 1.710), in cross-sectional studies (WMD (CI): 0.999 (0.288, 1.710), and in studies that used some questions (WMD (CI): 1.270 (1.127, 1.413)) for SSBs assessment, a higher SSB intake is associated with higher serum insulin levels.

2.9.3. Mean HOMA-IR

As depicted in Fig. 4, high SSBs intake was associated with a 0.21 increase in HOMA-IR in adolescents and children (WMD: 0.21; CI, 0.03–0.37; P = 0.02) with significant heterogeneity among studies (I² = 98.07 %, P < 0.001).

Fig. 4.

The forest plot of the weighted mean difference (WMD) of the effect of SSBs intake on HOMA-IR.

The subgroup analysis revealed sources of variability related to age, study design, setting, and evaluation instruments (Table S5).

Subgroup analyses show that studies conducted in school [WMD (CI): 2.748 (2.090, 3.406)], clinic [WMD (CI): 0.210 (0.178, 0.241)], and community [WMD (CI): 0.400 (0.390, 0.409)] settings, as well as cross-sectional [WMD (CI): 0.212 (0.181, 0.243)] and longitudinal studies [WMD (CI): 0.400 (0.391, 0.409)], and in studies that used some questions [WMD (CI): 0.210 (0.178, 0.242)] and 24-h recall [WMD (CI): 0.241 (0.007, 0.475)] for SSB assessment, a higher SSB intake is associated with higher HOMA-IR.

Results of the dose-response meta-analysis examining the association between the dose of SSBs and serum insulin, glucose, and HOMA-IR levelsThe details of the dose-response meta-analysis are displayed in Table 2. Fig. 5, Fig. 6, Fig. 7 give the findings for the FPG, FSI, and HOMA-IR levels, respectively.

Table 2.

Details of non-linear association between SSB intake, FPG, FSI, and HOMA-IR.

Mean difference	Coefficient 95 %	Standard error	T	P > |t|	Conf. Interval
FPG Mean difference
Dose_1	0.0022903	0.0024994	0.92	0.386	−0.0034733 0.0080538
Dose_2	−0.0014356	0.0012285	−1.17	0.276	−0.0042685 0.0013973
_cons	0.0660706	0.0371993	1.78	0.114	−0.019711 0.1518523
FSI Mean difference
Dose_1	0.0861684	0.1188476	0.73	0.601	−1.423933 1.59627
Dose_2	−0.0531232	0.0669516	−0.79	0.573	−0.9038242 0.7975778
_cons	1.511305	1.094724	1.38	0.399	−12.39848 15.42109
HOMA-IR Mean difference
Dose_1	−0.1768215	0.1687411	−1.05	0.354	−0.6453218 0.2916787
Dose_2	−0.095058	0.090997	−1.04	0.355	−0.3477061 0.1575901
_cons	0.234701	0.1270637	1.85	0.138	−0.1180845 0.5874865

Fig. 5.

Dose– response association between the SSBs dosage and mean difference in FPG with the study outcomes (Linear relation (solid line) and 95 % CI (gray area) of mean difference in study outcomes by 1 g/d increment in SSB dosage.

Fig. 6.

Dose– response association between the SSBs dosage and mean difference in FSI with the study outcomes (Linear relation (solid line) and 95 % CI (gray area) of mean difference in study outcomes by 1 g/d increment in SSB dosage.

Fig. 7.

Dose– response association between the SSBs dosage and mean difference in HOMA-IR with the study outcomes (Linear relation (solid line) and 95 % CI (gray area) of mean difference in study outcomes by 1 g/d increment in SSB dosage.

Given the results of the dose-response meta-analysis, no departure from linearity was detected between SSBs consumption and changes in serum glucose (P-nonlinearity = 0.27), insulin (P-nonlinearity = 0.57), and HOMA-IR (P-nonlinearity = 0.35) levels.

2.10. Publication bias

Figs. S1a, b, c display the obtained funnel plots. According to Begg's and Egger's meta-bias tests, no evidence of publication bias was found in the meta-analysis of the serum glucose, insulin, and HOMA-IR levels in highest versus lowest categories of SSBs [serum glucose: Begg's test (P = 0.27) and Egger's test (P = 0.46)]; [serum insulin: Begg's test (P = 0.31) and Egger's test (P = 0.37)], and [HOMA-IR: Begg's test (P = 0.49) and Egger's test (P = 0.56)].

2.11. Quality assessment

The Joanna Briggs Institute Critical Appraisal Checklist for the analytical cross-sectional and cohort studies was used to assess the quality of studies. Quality assessment of the included studies is presented in Table S7. All articles possessed more low-risk domains than high-risk ones and were of good quality. In two studies, SSB intake was not assessed with a valid tool [22,25].

3. Discussion

Higher sugar consumption is linked to increased risk of pediatric insulin resistance. Studies in the adult population have revealed that too much intake of SSBs can be result in weight gain and lead to a higher risk of cardiovascular illnesses, such as diabetes, hypertension, and MetS [[30], [31], [32]]. However, there is limited evidence in this field, particularly among adolescents and children. In this study, we evaluated the connection between children's consumption of SSBs and their risk of developing insulin resistance. According to the findings, children and adolescents who consumed more sugar-sweetened beverages had greater HOMA-IR.

Moreover, no evidence of departure from linearity was found for the relationship of SSBs intake dose with mean changes in FPG, FSI, and HOMA-IR levels. To the best of the authors' knowledge, no systematic review and dose-response meta-analysis has been done in this field.

There is ongoing debate on the link between insulin resistance and beverages with added sugar. Studies that show no correlation are difficult to interpret because of small sample sizes, problems with the methodology, short follow-up periods, and other uncontrolled confounding variables [13,28,33,34]. Additionally, there is discrepancies between studies because of different measures of weight, using different questionnaires to evaluate SSB intake, applying different statistical models to estimate the effect sizes, and different interviewees. Furthermore, the disparities in obesity risk observed in ethnically varied communities have been attributed to cultural difference [35,36]. Food preferences and likes appear to be influenced by cultural influences, which in turn impact eating habits and overall health [35].

This meta-analysis showed that a high intake of SSBs was associated with a 0.21 increase in HOMA-IR in adolescents and children. Previous studies have shown that too much SSBs intake can significantly increase HOMA-IR [13,15,22,37]. Both indicators of HOMA-IR were linked to all components of pediatric cardiometabolic diseases and MetS risk scores. As insulin resistance stimulates the development of cardiovascular disease [38], screening, prevention, and designing specific interventions for children with high HOMA-IR levels should be a priority. The association between SSB consumption and HOMA-IR levels may not directly indicate causation but could instead reflect their shared relationship with obesity. Wu et al. showed that when considering covariates and waist circumference (an indicator of central obesity), adolescents who had heavy and light-to-half consumption of SSBs had a 0.38 and a 0.27 increase in HOMA1-IR concentrations and a 0.22 and a 0.16 elevation in HOMA2-IR concentrations (all P ≤ 0.042), respectively [13]. This finding indicates that the associations between the two HOMA-IR markers and SSB intake were not confounded by adiposity [13].

The results revealed that high SSBs intake was not significantly connected to higher FPG and FSI levels. This finding may be due to the typical postpubertal drops in insulin levels and insulin resistance reported in the previous studies [39]. The lack of information about the puberty stage was one of the main limitations of the studies, so subgrouping based on the pubertal stage was impossible.

SSBs intake physiological effect on metabolic disorders might be due to heavy dietary sugar load resulting in a postprandial upsurge in blood sugar and insulin levels, which might lead to insulin resistance and hyperinsulinemia over time [40]. Furthermore, increased levels of inflammatory biomarkers linked to insulin resistance such as interleukin-6 were associated with high blood sugar [41]. High-fructose diet caused insulin resistance in adipose tissues and the liver, consequently leading to increased energy intake, lipogenesis, intra-abdominal fat storage, leptin resistance, blood pressure, hunger rating, CRP concentration, and impaired fasting glucose [42,43].

Using subgroup analysis, we were able to identify likely causes of heterogeneity; subgrouping based on assessment tools and setting decreased the heterogeneity.

The majority of settings for studies were schools. Since children spend the majority of their time in schools, schools and other educational settings are the main sites for nutrition teaching. As schools serve at least one-third of what kids eat on a daily basis, children's eating habits are formed there as well as and during childhood [[44], [45], [46], [47], [48], [49]]. The high efficacy of school-based interventions may be attributed to both the intervention type and the greater control over student dietary habits within school settings. [50]. Furthermore, in line with WHO recommendations, schools should be the first to reduce children's consumption of sugar-sweetened beverages (SSBs) by enacting laws that forbid the sale of the majority of SSBs on campus, remove soft drink vending machines, use water coolers in school cafeterias, and provide students with free access to clean water [44]. According to a Chinese study, the consumption of SSBs by Chinese students was significantly correlated with the views of teachers and school administrators. This might be because friends and teachers have such a big impact on the children. As a result, educators and schools can advise students on how to lead healthy lives and make dietary and lifestyle choices [25].

Various policy solutions have been proposed to minimize sugar intake in healthy populations, including modifications to the purchasing pattern and educational interventions. Studies' conclusions showed that the majority of interventions were educational and that most of them were successful in lowering sugar intake. It appears that focused training combined with additional treatments can successfully alter behavior, despite the doubts of many experts regarding the efficacy of unaccompanied educational interventions [51,52]. Additionally, studies on tax increases were successful in reducing the consumption of sugar-sweetened beverages [[53], [54], [55]]. Reducing sugar intake tends to be more successful when all the strategies are used at once [56]. However, it is important to notice that proposed interventions are influenced by resources, limitations, and infrastructure of various countries.

3.1. Strengths and limitations

To the best of our knowledge, this study is the first dose-response meta-analysis examining the relationship between SSB intake and levels of FPG, FSI, and HOMA-IR in children and teens. To improve the search, a thorough search plan and a manual search of reference lists were used. The strength of our findings was increased when the risk of bias analysis revealed that all of the included studies received excellent quality scores. However, this study has some limitations. One was the measurement error inherent in evaluating SSB consumption using various instruments as recall of attendees greatly influence their responses.

Additionally, the intake assessment was measured using a variety of instruments and units and was self-reported. Second, it was not possible to subgroup based on the types of SSBs since different types of SSBs were taken into consideration in the included publications. Furthermore, various papers had different confounding variables that may change the outcome. Besides, one of the main limitations of the included studies was the absence of information regarding the puberty stage, which made it impossible to subgroup participants in this analysis depending on their pubertal stage.

4. Conclusion

The results of this meta-analysis showed that higher SSB consumption in children and adolescents is associated with higher HOMA_IR. However, it is advised that more comprehensive prospective long-term studies be carried out in order to validate the relations. The clinical implications of this research should compel policymakers to employ educational interventions in addition to other community- and school-based initiatives, such product reformulation or providing healthy beverages (like water) rather than sugar-filled beverages at schools.

CRediT authorship contribution statement

Mahdieh Abbasalizad Farhangi: Writing – original draft, Methodology, Conceptualization. Zeinab Nikniaz: Writing – review & editing, Software, Methodology, Investigation, Data curation. Seyedeh-Tarlan Mirzohreh: Writing – original draft, Methodology, Investigation. Leila Nikniaz: Writing – review & editing, Writing – original draft, Supervision, Methodology, Conceptualization.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

If requested, data will be available at https://hsri-en.tbzmed.ac.ir/

Funding

This research received no specific grant from any funding agency.

Declaration of competing interests

I declare that the authors have no competing interests as defined by BMC, or other interests that might be perceived to influence the results and/or discussion reported in this paper.

Handling Editor: Dr D Levy

Appendix A. Supplementary data

The following is/are the supplementary data to this article.