Sex differences in children’s cognitive functions and phthalates exposure: a meta-analysis

Background Phthalates exposure might affect children’s intelligence development. This study aimed to determine (1) whether sex and age affect cognitive function and (2) whether sex differences in cognitive performance are wider with higher phthalate concentrations. Methods Data were collected from PubMed (1998–2022), PROQUEST (1997–2022), and SpringerLink (1995–2022). The study followed the PRISMA process. The included articles were followed by PECO framework. The GRADE applied to assess the certainty of evidence. Of 2422 articles obtained, nine were selected using inclusion criteria. The random-effects model was used to estimate the pooled effects. Results Our meta-regression indicated a significant difference between sex differences with age at phthalate concentration assessment (β = −0.25; 95% CI = −0.47, −0.03) and MEHP concentration (β = −0.20; 95% CI = −0.37, −0.03). Conclusions The limitation of the current article is it only provides information on intelligence level rather than other aspects of cognitive function. Thus, the sequelae of phthalate exposure on attention and executive function are still unclear. Our analysis shows significant difference between sex differences in cognitive function scores associated with age at phthalate concentration assessment. Girls might be more resilient in cognitive function at a younger age or during lower concentrations of phthalates metabolites. Impact This is the first meta-analysis to evaluate the pooled estimates of sex differences in objective cognitive functions among children with phthalate exposure. The female might be a protective factor when exposed to toxic plasticizers while the concentration is low. This study captures the possible role of sex in cognitive functioning and plasticizer exposure through a meta-analysis of children’s sex, cognitive scores, and plasticizer exposure.


5-6
Selection process 8 Specify the methods used to decide whether a study met the inclusion criteria of the review, including how many reviewers screened each record and each report retrieved, whether they worked independently, and if applicable, details of automation tools used in the process.

6
Data collection process 9 Specify the methods used to collect data from reports, including how many reviewers collected data from each report, whether they worked independently, any processes for obtaining or confirming data from study investigators, and if applicable, details of automation tools used in the process.

6
Data items 10a List and define all outcomes for which data were sought.Specify whether all results that were compatible with each outcome domain in each study were sought (e.g. for all measures, time points, analyses), and if not, the methods used to decide which results to collect.

10b
List and define all other variables for which data were sought (e.g.participant and intervention characteristics, funding sources).Describe any assumptions made about any missing or unclear information. 6 Study risk of bias assessment 11 Specify the methods used to assess risk of bias in the included studies, including details of the tool(s) used, how many reviewers assessed each study and whether they worked independently, and if applicable, details of automation tools used in the process.

7
Effect measures 12 Specify for each outcome the effect measure(s) (e.g.risk ratio, mean difference) used in the synthesis or presentation of results.

6-7
Synthesis methods 13a Describe the processes used to decide which studies were eligible for each synthesis (e.g.tabulating the study intervention characteristics and comparing against the planned groups for each synthesis (item #5)).

6-7
13b Describe any methods required to prepare the data for presentation or synthesis, such as handling of missing summary statistics, or data conversions.

6-7
13c Describe any methods used to tabulate or visually display results of individual studies and syntheses.6-7 13d Describe any methods used to synthesize results and provide a rationale for the choice(s).If meta-analysis was performed, describe the model(s), method(s) to identify the presence and extent of statistical heterogeneity, and software package(s) used.

6-7
13e Describe any methods used to explore possible causes of heterogeneity among study results (e.g.subgroup analysis, meta-regression).

7
13f Describe any sensitivity analyses conducted to assess robustness of the synthesized results.7 Reporting bias assessment 14 Describe any methods used to assess risk of bias due to missing results in a synthesis (arising from reporting biases).

7
Certainty assessment 15 Describe any methods used to assess certainty (or confidence) in the body of evidence for an outcome.7

Study selection 16a
Describe the results of the search and selection process, from the number of records identified in the search to the number of studies included in the review, ideally using a flow diagram.

7
16b Cite studies that might appear to meet the inclusion criteria, but which were excluded, and explain why they were excluded.

Study characteristics
17 Cite each included study and present its characteristics.8 Risk of bias in studies 18 Present assessments of risk of bias for each included study.9-10

Results of individual studies
19 For all outcomes, present, for each study: (a) summary statistics for each group (where appropriate) and (b) an effect estimate and its precision (e.g.confidence/credible interval), ideally using structured tables or plots.

8
Results of syntheses 20a For each synthesis, briefly summarise the characteristics and risk of bias among contributing studies.8-10 20b Present results of all statistical syntheses conducted.If meta-analysis was done, present for each the summary estimate and its precision (e.g.confidence/credible interval) and measures of statistical heterogeneity.If comparing groups, describe the direction of the effect.9-10 20c Present results of all investigations of possible causes of heterogeneity among study results.9-10 20d Present results of all sensitivity analyses conducted to assess the robustness of the synthesized results.9-10 Reporting biases 21 Present assessments of risk of bias due to missing results (arising from reporting biases) for each synthesis assessed.

9-10
Certainty of evidence 22 Present assessments of certainty (or confidence) in the body of evidence for each outcome assessed.10

Probably high
The participants were recruited within the same time frame using the same inclusion and exclusion criteria and were of similar age and health status.However, there is insufficient information provided about the subject sources.

Low
The pregnant participants who were above 18 years were recruited from 3 cities in Korea Seoul, Cheonan, and Ulsan.

Probably low
The comparison groups are recruited from the same area.The selected area is regarded as representative of the general Swedish population.

Probably low
The covariates were based on selfreported information given during the study interview.The variables were categorical and modeled using categories, including birth weight, breastfeeding time, household income, and maternal education.

Probably high
Researchers employed the general linear model to assess differences in endocrine-disrupting chemicals and gonadal hormones between ADHD and control groups, adjusting for age, urinary creatinine levels, and specific gravity as covariates, although indirect evidence suggests that primary covariate and known confounder distribution differences between groups were not adequately adjusted for in the final analyses.

Probably high
Based on information obtained from the questionnaires applied in the different aligned studies, variables corresponded across the studies and data/variables.Potential confounders in neurodevelopment to the three cohorts were the following: the highest education level of the household of the child, body mass index (BMI) z-score, and sex of the child, was modeling.

Probably low
The variable was adjusted for child sex, maternal nutrition index, energy, age at birth, pre-pregnancy weight, education level, maternal IQ, and smoking status.These variables are considered appropriate adjustments.
The study did not adjust for potential confounding factors, such as prematurity and birth weight.

Attrition/ Exclusion bias
Were outcome data complete without attrition or exclusion from analysis?

Probably high
There is indirect evidence of incomplete outcome data, as inconsistencies exist between the number of recruited and outcome participants without an explanation for participant attrition.

Definitively low
Clearly explain there are two participants were excluded due to extremely dilute urine (creatinine < 10 mg/dL).There is no evidence of exclusion bias.

Probably Low
A consistent number of participants reported on the outcome results.There were no other significant differences between boys and girls.

Probably high
There is indirect evidence that the exclusion of subjects from analyses was not adequately addressed.Only the source of the subject was mentioned.

Probably high
There is indirect evidence of incomplete outcomes of data.The inconsistent number of participants was provided on the recruited and outcome participant without explaining the reasons for participant attrition in the outcome report.

Probably high
The article only partially explains the number of participants and the reasons for inclusion and exclusion, but the number of partial exclusions remains unspecified.

Probably low
From the methods, eligibility criteria for cases were all incident; the researcher excluded those patients who had a comorbid autism spectrum disorder, intellectual disability, psychotic disorders, major depressive disorder, bipolar disorder, or neurological disorders.

Probably high
No specified exclusion biases.No state about the loss to follow-up of participants.

Low
There is no evidence of exclusion bias, as the method description indicates that data for mother-child pairs with complete data were selected from the Swedish Environmental Longitudinal Mother and Child, Asthma and Allergy (SELMA) pregnancy cohort.

Detection bias
1. Can we be confident in the exposure characterization?

Probably low
Phthalates were measured using urine sampling, while maternal and children's cognitive functions were assessed separately, with the children being individually evaluated by a licensed clinical psychologist with 20 years of experience.

Definitively low
There is direct evidence that exposure was consistently assessed.The phthalates were measured by urine sampling of creatinine concentration.

Probably low
The study collected spot urine samples from maternal participants during the third trimester of pregnancy, with each analytical run including calibration standards, reagent blanks, and quality control samples, while specific gravity correction for urinary dilution, as recommended for phthalates, was measured using a handheld refractometer.

Probably low
Indirect evidence suggests that acceptable methods were used to assess outcomes, including urinary sampling methodology with highperformance liquid chromatography coupled with tandem mass spectrometry (HPLC-MS/MS) and limits of detection (LOD), indicating consistent exposure assessment through well-established, direct measurement techniques.

Probably low
Data from the prospective Polish Mother and Child Cohort Study (REPRO_PL), established in 2007, was approved by the Nofer Institute of Occupational Medicine's Ethical Committee and included informed consent from pregnant women and children's parents for scheduled cohort phases, providing direct evidence of consistent exposure assessment.

Probably low
The article describes in detail the geographic distribution and number of participants.There is direct evidence that exposure was consistently assessed.

Low
The phthalates were measured using urine sampling.The children's cognitive function was measured separately and measured by a senior psychiatrist and child psychologist.

Low
The phthalates were measured using urine sampling.In the three cohorts, the neurodevelopment of children was assessed by trained psychologists using the WISC test.

Low
The phthalates were measured using blood and urine sampling.The children's cognitive function was measured by a trained psychiatrist.
2. Can we be confident in the outcome assessment?

Probably high
There is indirect evidence that the cross-sectional design cannot rule out the possibility of misclassification between exposure and outcome.

Probably low
Indirect evidence suggests that acceptable methods were used to assess outcomes; however, the single spot urine sample collected during one trimester may not represent exposure throughout pregnancy due to phthalates' short biological half-lives, and while measurement error in phthalate metabolite concentration is suspected to be non-differential, it is still relevant to important qualities such as child's neurologic development, phthalate exposure, and confounding variables.

Probably low
The phthalates were measured using urine sampling for the same length of time in all study groups.The outcome assessor was assessed using an acceptable method.

Probably low
The phthalates were analyzed with an LC-ESE-MS/MS (Agilent 1200/ API4000).The subsequence procedure used well-established methods.

Probably low
The phthalates were measured using urine sampling for the same length of time in all study groups.The outcome assessor was assessed using acceptable methods.

Probably low
Phthalate metabolite measurements followed the Centers for Disease Control and Prevention (CDC) Laboratory Procedure Manual using urine samples, while cognitive function assessment was conducted by trained examiners with high interrater reliability (kappa value > 0.8) in quiet rooms for 30 to 45 minutes.

Probably low
There is indirect evidence that the cross-sectional design cannot rule out the possibility of misclassification between exposure and outcome.

Probably high
The data were obtained from European Human Biomonitoring Initiative (HBM4EU) aligned studies.
There is indirect evidence that the outcome was assessed using acceptable methods.

Probably low
From the text, phthalates and cognitive data were obtained from the SELMA pregnancy cohort.There is direct evidence that the outcome was assessed using well-established methods.

Selective reporting bias
Were all measured outcomes reported?

Probably low
The outcome has been reported in methods and abstract with 95%CI.This outcome provides sufficient detail to be included in the metaanalysis.

Probably low
The outcome has been reported in methods and abstract with 95%CI.This outcome provides sufficient detail to be included in the metaanalysis.

Probably Low
Metabolites were detected in 84-100% of urine samples, and as expected, the four DEHP metabolite concentrations, adjusted for specific gravity, showed high correlation (Spearman's r values .68-.97), leading to their conversion to molecular weights and summed (ΣDEHP); outcomes were reported in methods and abstract with 95%CI, providing sufficient detail for inclusion in the meta-analysis.

Probably low
The research protocol was approved by the Research Ethics Committee of the National Health Research Institutes (No. EC1000903) and the collaborating hospitals.
The outcome has been reported in the methods and abstract.This outcome provides sufficient detail to be included in the meta-analysis.

Probably low
The outcome has been reported in methods and abstract with 95%CI.This outcome provides sufficient detail to be included in the metaanalysis.

Probably high
The outcome has been reported in methods and abstract with 95%CI.This outcome provides sufficient detail to be included in the metaanalysis.

Probably low
The outcome has been reported in the abstract and method, following the protocol of the study.This outcome provides sufficient detail to be included in the meta-analysis.

Probably low
The results of this study protocol evaluation lack significant and consistent evidence of inverse associations between this group's exposures and FSIQ.All outcomes have been reported crudely and adjusted β and 95% CI of all linear regression models.

Probably high
Data were analyzed using random splits (e.g., 40% for weight estimation and 60% for hypothesis testing) and potentially multiple random splits to address generalizability, with the ensemble's strength lying in the increased identification of concerning agents, albeit at the cost of forgoing model parsimony.

Figure 2 .
Figure 2. Correlation plots of MEHP concentration (µg/L) and (A) age at phthalate concentration assessment and (B) age at intelligence assessment.

Table 2 . Risk of bias of studies based on the NTP/OHAT risk of bias tool (NTP/OHAT, 2019). Domain of risk of bias\Study Cho et al., 2010, Doherty et al., 2017 Whyatt et al., 2012 Huang et al., 2017 Jankowska et al., 2019a Kim et al., 2011 Tsai et al., 2020 Rosolen et al., 2022 Gennings et al., 2022 Selection bias
explain any amendments to information provided at registration or in the protocol.
n/a Support 25 Describe sources of financial or non-financial support for the review, and the role of the funders or sponsors in the review.20 Competing interests 26 Declare any competing interests of review authors.20 Availability of data, code and other materials 27 Report which of the following are publicly available and where they can be found: template data collection forms; data extracted from included studies; data used for all analyses; analytic code; any other materials used in the review.15 From: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al.The PRISMA 2020 statement: an updated guideline for reporting systematic reviews.BMJ 2021;372:n71.doi: 10.1136/bmj.n71