Don’t take our word for it. Here’s every fact on this site, with its evidence grade, its caveats, and a link to the source, so you can check us.
We start you on the strong ones: findings that have been replicated or come from one large, well-controlled study. Moderate and weak facts are here too, labeled for what they are. Use “Link to this fact” to share a single card.
Children and pregnancy / Measured in people or real products
In a Finnish birth cohort of 5,131 mother-child pairs, higher maternal thyroid-stimulating hormone (TSH) in early pregnancy was weakly associated with more teacher-rated ADHD symptoms at age 8 in girls (odds ratio about 1.4 per natural-log unit of TSH for combined symptoms, 95% CI 1.07-1.80), but not in boys.
Weak evidence.
Caveats
Single-cohort observational study; teacher-rated symptoms rather than clinical diagnosis; one blood sample; no parental ADHD data; the authors call the association weak, and a high-TSH cutoff analysis was not significant in either sex.
Source: Päkkilä 2014 (NFBC 1986), The Journal of Clinical Endocrinology & Metabolism Link to this fact
Children and pregnancy / Measured in people or real products
The same Finnish cohort found no link between children's ADHD symptoms and maternal hypothyroxinemia (low free T4 with normal TSH; 66 mothers) or thyroid-peroxidase antibodies, although the hypothyroxinemia group was too small to rule out an effect.
Weak evidence: Only 66 exposed mothers; confidence intervals are wide and several girls' cells had zero cases.
Source: Päkkilä 2014 (NFBC 1986), The Journal of Clinical Endocrinology & Metabolism Link to this fact
Children and pregnancy / Other evidence
The same cohort found no adverse association between prenatal urinary MEP (the diethyl phthalate, fragrance-phthalate marker) and child behavior at 6-10 years; the one significant MEP result was inverse (fewer DSM affective-problem symptoms, beta -0.10, 95% CI -0.19 to -0.01).
Weak evidence: Single small cohort; one spot urine; null results are not proof of safety; MEP is a marker of exposure from several sources and no product-use or musk data were collected.
Source: Kobrosly 2014, Environmental Health Perspectives Link to this fact
Children and pregnancy / Other evidence
The sex pattern in Kobrosly 2014 is exploratory: of about 255 coefficients, about 24 reached p <= 0.05 by the page's count, and only a few sex-interaction tests were significant, so the claim that prenatal phthalates affect boys more than girls is not established by this cohort.
Weak evidence: One small cohort; interaction tests reported but many comparisons; MBzP-conduct interaction p printed as 0.01 in the text and < 0.001 in Table 4.
Source: Kobrosly 2014, Environmental Health Perspectives Link to this fact
Children and pregnancy / Other evidence
In the same cohort, children's ADHD-symptom scores were not related to continuous maternal TSH or free T4 levels, to subclinical hypothyroidism, or (for hypothyroxinemia) to oppositional-behavior scores, so only the low-free-T4 group stood out.
Weak evidence: Null results from one cohort; supplement tables (subclinical hypothyroidism) not examined here; the 127-child exposed group limits power.
Source: Modesto 2015 (Generation R), JAMA Pediatrics Link to this fact
Children and pregnancy / Measured in people or real products
In the same study, higher DEHP in carpet dust (a plastics phthalate, not a fragrance one) was more common in homes of the 27 children with developmental delay (odds ratio 2.12, 95% CI 1.10-4.09 per ln-unit), but the authors report the finding did not survive correction for multiple comparisons and may be due to chance.
Weak evidence: 27 cases; exposure measured after the delay was already present (reverse causation possible); DEHP is not a fragrance ingredient; abstract states OR 2.10, table 2.12.
Source: Philippat 2015 (CHARGE), Environmental Health Link to this fact
Children and pregnancy / Other evidence
In a Taiwanese birth cohort of 122 mother-child pairs, higher third-trimester urinary levels of breakdown products of the plastics phthalate DEHP and of dibutyl phthalate (MBP) went with higher parent-rated delinquent and aggressive behavior at age 8-9, roughly 3 to 4 behavior-scale points per 10-fold increase.
Weak evidence.
Caveats
Association from a small, selectively followed cohort (122 of 610 enrolled), one urine sample, many unadjusted comparisons, no correction for co-exposures; DEHP and DBP are not fragrance phthalates.
Source: Lien 2015, Taiwan, Environmental Health Perspectives Link to this fact
Children and pregnancy / Other evidence
In the same Taiwanese cohort, the urinary marker of the fragrance-solvent phthalate DEP (MEP, geometric mean about 62 ug/g creatinine) was not associated with children's behavior scores at age 8-9 in adjusted models.
Weak evidence: Small study; one spot urine for a short-half-life metabolite; the authors' Discussion describes MEP as associated with clinical-range externalizing scores based on an 8-case cell.
Source: Lien 2015, Taiwan, Environmental Health Perspectives Link to this fact
Children and pregnancy / Measured in people or real products
The one phthalate-thyroid result that held in repeated-measure models was a small drop in free T3 (2.9% per interquartile range) with MCPP, a non-specific metabolite of several high-molecular-weight phthalates; the confidence interval nearly touched zero.
Weak evidence: P = 0.049, one of 48 models, MCPP cannot be traced to a single product or parent compound.
Source: Johns 2015, Reproductive Biology and Endocrinology Link to this fact
Children and pregnancy / Measured in people or real products
In 546 French mother-son pairs, boys whose mothers had higher urinary bisphenol A (BPA) in mid-pregnancy had slightly higher mother-rated peer-problem scores at age 3 (about 11% higher per doubling) and hyperactivity-inattention scores at age 5 (about 8% higher per doubling).
Weak evidence.
Caveats
One spot urine at 22-29 weeks; boys only; no association survived false-discovery-rate correction (lowest corrected p 0.42); BPA is not a fragrance ingredient; association, not causation.
Source: Philippat 2017, Environmental Health Perspectives Link to this fact
Children and pregnancy / Measured in people or real products
The same French cohort found the fragrance-solvent marker MEP (from diethyl phthalate) was not associated with any behavior score in boys at 3 or 5 years in the main models.
Weak evidence: Single spot urine; MEP may mark fragranced-product use rather than DEP itself; a null in boys from one cohort does not show fragrance is safe.
Source: Philippat 2017, Environmental Health Perspectives Link to this fact
Children and pregnancy / Measured in people or real products
Higher prenatal urinary MnBP (a dibutyl phthalate metabolite) was associated with slightly higher internalizing, emotional and peer-problem scores in 3-year-old boys (about 6% per doubling), but not at age 5.
Weak evidence: Not FDR-robust; one cohort; MnBP is not specific to fragrance; association only.
Source: Philippat 2017, Environmental Health Perspectives Link to this fact
Children and pregnancy / Other evidence
In a pooled analysis of 4 of the 6 studies of maternal hypothyroxinemia (normal TSH, low free T4) in early pregnancy, the odds of ADHD symptoms in children were about 1.5 times higher (OR 1.54, 95% CI 1.02-2.33).
Weak evidence.
Caveats
Lower CI bound barely above 1; four studies, two from the same Dutch cohort; free T4 by immunoassay in pregnancy can misclassify; the review's own Discussion calls the hypothyroxinemia evidence mixed.
Children and pregnancy / Review of other studies
In a 2019 review, newborn total thyroxine from heel-prick screening cards did not predict a later ADHD diagnosis (ORs 0.99 and 1.08, CI 0.95-1.24), and newborn TSH did not predict ADHD symptoms at ages 4-5.
Weak evidence: Few studies; sample timing after birth was not adjusted; some studies may share participants.
Children and pregnancy / Other evidence
Higher pregnancy levels of low-molecular-weight phthalate metabolites (which include the DEP marker MEP) went with more self-reported attention problems (about 1.5 T-score points per doubling) and hyperactivity (about 0.8) at age 16 in the CHAMACOS cohort.
Weak evidence.
Caveats
Association only; one cohort, self-report at one age, no multiple-comparison correction; parent reports were null and teacher reports opposite; the signal shrank after adjusting for PBDEs, pesticides or manganese; ΣLMW also includes dibutyl phthalates, and MEP cannot be separated from the fragranced-product mixture.
Source: Hyland 2019 (CHAMACOS), Environmental Health Perspectives Link to this fact
Children and pregnancy / Measured in people or real products
In the same cohort, methylparaben in newborns' first stool went with about 4.8 times the odds of preterm birth (95% CI 2.29-10.10) and a birth about 0.6 weeks earlier, from only 13 preterm births (6 in the exposed group).
Weak evidence.
Caveats
Very few events; crude OR 3.97 (1.29-12.27); association only; weighted intervals are narrower than crude ones and the paper does not say whether weights' uncertainty was accounted for. Other cohorts cited in the paper found longer gestation with methylparaben.
Source: Baker 2020, Environment International Link to this fact
Children and pregnancy / Review of other studies
One systematic review summarized in the 2022 scoping review found 14 of 16 studies reported a phthalate association with ADHD in children, but the authors of that review said study limitations made definite conclusions difficult and did not pool the results.
Weak evidence: Vote count of directions across mixed designs including cross-sectional studies; no risk-of-bias grading; not independent confirmations; relayed second-hand by the scoping review.
Source: Moore 2022 Link to this fact
Children and pregnancy / Other evidence
In 491 mother-child pairs in southern Taiwan, higher second-trimester maternal urine levels of MnBP (a breakdown product of the plasticizer DBP) went with higher odds of a clinical-range behavior score at age 1.5-3 for total problems (OR about 8.8 per 10-fold rise, 95% CI 1.8-43), anxiety and sleep problems.
Weak evidence: One cohort, single spot urine, very wide CIs (anxiety 2-521, sleep 1-1633), no multiple-comparison control for outcomes, and the abstract misprints the total-problems OR as 19.32.
Source: Huang 2024, Toxics Link to this fact
Children and pregnancy / Other evidence
In the same Taiwanese cohort, the child's own urinary MnBP (98 children) went with clinical-range total problems (OR 7.06, 95% CI 1.01-49.05) and internalizing problems (OR 11.04, 1.27-95.72), but the paper's claim that prenatal and postnatal exposure have 'distinct' effects is not tested.
Weak evidence.
Caveats
Small subsample, child urine apparently concurrent with the behavior rating, CIs overlap the prenatal ones, and child MBzP and MEHP showed implausible inverse ORs for externalizing.
Source: Huang 2024, Toxics Link to this fact
Children and pregnancy / Other evidence
In this cohort MEP, the urinary marker of the fragrance phthalate DEP, did not predict clinical-range behavior problems prenatally or postnatally (all 95% CIs include 1); it showed only uncorrected linear associations with ADHD-problems and pervasive-developmental-problems T-scores (coefficients 0.12-0.13), so the study gives no basis for a perfume-and-toddler-behavior claim.
Weak evidence: MEP is a product-use marker that cannot be separated from the fragrance mixture it travels with; one spot urine; the T-score signals are not corrected for 300 tests.
Source: Huang 2024, Toxics Link to this fact
Children and pregnancy / Other evidence
The review lists a prenatal triclosan study in which cord TT4 was lower and thyroid antibodies (TPOAb) were reported as 328% higher at age 3.
Weak evidence: One study relayed without confidence interval; triclosan is a personal-care ingredient, not fragrance.
Source: Street 2024, Front Endocrinol, Frontiers in Endocrinology Link to this fact
Children and pregnancy / Animal study
The same rat study found changes in the brain's dopamine machinery in the prefrontal cortex (less TH, DDC, VMAT2, DRD1 and DRD2; more of the reuptake transporter DAT), and a similar pattern in human neuroblastoma cells at 1-10 micromolar triclosan.
Weak evidence: Rat plus cell-line data from one lab; protein and mRNA only; cell doses (up to 10 uM) are well above typical blood levels; DRD1 moved opposite ways in rats and cells.
Source: Cui 2024, Ecotoxicology and Environmental Safety Link to this fact
Children and pregnancy / Measured in people or real products
In the same Baltimore children, recent scented candle use went with about 2.6 times higher urine methyl paraben (2.57, 95% CI 1.09-6.06) and scented carpet powder with about 2.8 times higher 2,5-dichlorophenol (2.77, 1.21-6.34), but only in models adjusting for other products; the candle result was not significant on its own (2.32, 0.95-5.64).
Weak evidence: Wide intervals, many comparisons, and the authors say carpet powder may only mark carpeted homes; no ingredients checked.
Source: Fandiño-Del-Rio 2024, Journal of Exposure Science & Environmental Epidemiology Link to this fact
Children and pregnancy / Measured in people or real products
Black children in this Baltimore sample (86% of 110) were reported by the authors to have higher levels of some EDCs tied to scented products than other US children, but lower benzophenone-3, consistent with less sunscreen use; the study had no comparison group of non-Black children.
Weak evidence: The comparison is to NHANES 2015-2016 published data, not to a matched group; unmeasured product use, diet and housing could explain differences.
Source: Fandiño-Del-Rio 2024, Journal of Exposure Science & Environmental Epidemiology Link to this fact
Children and pregnancy / Measured in people or real products
In the same cohort, higher prenatal BPA was associated with lower cognitive scores in boys (about 1.4 points per scaled log unit, 95% CI -2.54 to -0.23) but not girls, while the overall association was null.
Weak evidence.
Caveats
One of 36 exploratory interaction tests (p = 0.021), no multiplicity correction; one urine sample; single cohort; direction of BPA sex differences conflicts across other cohorts. BPA is not a fragrance ingredient.
Source: Zhou 2025, Environmental Research Industry-funded Link to this fact
Children and pregnancy / Measured in people or real products
The only paraben signal was methyl-paraben (log MC 0.042, 95% CI 0.013 to 0.071) with attention variability; ethyl-, propyl- and butyl-paraben, BPA, triclosan and oxybenzone showed no clear association with attention in the same children.
Weak evidence: Single cohort, uncorrected for multiple testing; parabens are personal-care preservatives, not fragrance ingredients.
Source: Fabbri 2025 (HELIX), Environmental Research Link to this fact
Children and pregnancy / Review of other studies
The review names early pregnancy (roughly weeks 6-14) as the most vulnerable window for thyroid disruption by EDCs and says boys show larger IQ and behavior deficits than girls, citing cohort studies rather than its own data.
Weak evidence: Secondhand; sex difference not consistent across our phthalate cohorts.
Source: Eraslan Aydemir & Sahin 2025, Turk J Med Sci, Turkish Journal of Medical Sciences Link to this fact
Children and pregnancy / Measured in people or real products
In boys only, higher prenatal benzophenone-3 (sunscreen filter) went with about 3.2 points worse parent-rated sleep disturbance (top vs lowest tertile, 95% CI 0.27 to 6.14) and mid-level BPA with about 5.7 points more sleep-related impairment; neither was seen in girls.
Weak evidence.
Caveats
Boys-only subgroup estimates from many uncorrected tests; the BP3-by-sex interaction was not significant (p = 0.71), the BPA CI lower bound is printed with inconsistent signs in the table, and the mixture was null. Exploratory; benzophenone-3 and BPA are not fragrance ingredients.
Source: Geiger 2025, Frontiers in Pediatrics Link to this fact
Children and pregnancy / Other evidence
In 217 newborns from two US birth cohorts (Puerto Rico and Illinois), higher pregnancy urine levels of 2,4-dichlorophenol and propylparaben went with weaker sucking strength at 1-8 weeks of age (propylparaben: about 0.6 cmH2O lower amplitude per log unit, 95% CI 0.1 to 1.1); the study is the first to look at phenols and infant sucking.
Weak evidence.
Caveats
Small exploratory sample of 217 across two very different cohorts; 72 single-chemical tests; one NNS session per infant; no evidence yet that altered sucking predicts later outcomes. Not fragrance ingredients (dichlorophenol, a herbicide and mothball-chemical metabolite; propylparaben, a preservative).
Source: Woodbury 2025 (ECHO), Environmental Epidemiology Link to this fact
Children and pregnancy / Other evidence
In the same study, the combined effect of seven pregnancy phenols (two dichlorophenols, benzophenone-3, BPA, BPS, methyl- and propylparaben) was clearest for more sucking bursts per minute (+0.72 per quartile, 95% CI 0.10 to 1.34); other mixture estimates had confidence intervals spanning zero.
Weak evidence.
Caveats
Only one mixture outcome excluded zero; Bayesian kernel machine regression importance scores were below 0.50 for bursts and cycles per burst; association only; none of the compounds is a fragrance ingredient.
Source: Woodbury 2025 (ECHO), Environmental Epidemiology Link to this fact
Children and pregnancy / Other evidence
In a Korean cohort of 242 healthy first-trimester women, a single urinary bisphenol F level was associated with lower TSH (B -0.122 per ln-unit, 95% CI -0.223 to -0.020), but BPA and BPS were not associated with TSH.
Weak evidence.
Caveats
Single small cohort, one spot urine, no multiple-comparison correction (about 2 false positives expected in 45 tests), BPF detected in only 42% of samples, TSH only, no health outcome; BPF direction opposes the authors' own mechanism.
Source: Kim 2025, Environmental Analysis Health and Toxicology Link to this fact
Children and pregnancy / Other evidence
In the same Korean first-trimester cohort, triclosan, benzophenone-3 and methyl-, ethyl-, propyl- and butylparaben (personal-care-product chemicals) were not associated with TSH or with the oxidative-stress markers MDA and 8-OHdG.
Weak evidence: A null from one small cohort with one spot urine per woman; it does not show safety, and it contradicts earlier cohorts the authors cite (propylparaben, BPS).
Source: Kim 2025, Environmental Analysis Health and Toxicology Link to this fact
Children and pregnancy / Other evidence
The PAH metabolites 1-hydroxypyrene (lower TSH) and 2-hydroxyfluorene (higher TSH) were associated with first-trimester TSH in opposite directions, with wide confidence intervals nearly touching zero.
Weak evidence.
Caveats
Opposite signs for correlated combustion metabolites, no correction for 45 tests, and PAH sources (cooking smoke, air pollution) were not apportioned; not a fragrance or candle finding.
Source: Kim 2025, Environmental Analysis Health and Toxicology Link to this fact
Children and pregnancy / Other evidence
The authors write that oxidative stress mediates EDC effects on thyroid function, but their data show no such link: none of the three TSH-associated chemicals was associated with MDA or 8-OHdG, and no mediation analysis was run.
Weak evidence: Absence of association in one underpowered cohort is not evidence against a ROS pathway either; the point is that the paper did not test it.
Source: Kim 2025, Environmental Analysis Health and Toxicology Link to this fact
Children and pregnancy / Other evidence
In a Shenzhen survey of 5,602 mother-child pairs, mothers who recalled a thyroid problem during pregnancy (3.9%) had children with about 54% higher adjusted hazard of screening positive for ADHD on a parent-rated scale (HR 1.54, 95% CI 1.05-2.26).
Weak evidence.
Caveats
Self-reported thyroid yes/no with no hormone tests; the ADHD outcome is a mother-completed screen, not a clinical diagnosis; shared-reporter and recall bias; no sex split; lower CI limit near 1.05.
Source: Feng 2025, European Child & Adolescent Psychiatry Link to this fact
Children and pregnancy / Other evidence
The same Shenzhen study reported that mothers rating first-trimester stress as significant had children with higher ADHD-screen risk than mothers rating it moderate (HR 0.64) or minor (0.57), but its continuous stress model and group means point the other way, so the stress result is unreliable.
Weak evidence: Non-validated single-item stress ratings recalled 3-7 years later; the scale direction is stated inconsistently in the paper; no mediation or interaction with thyroid tested.
Source: Feng 2025, European Child & Adolescent Psychiatry Link to this fact
Children and pregnancy / Measured in people or real products
In three US cohorts (878 mother-child pairs), a mixture of 10 phenols and 13 phthalate metabolites measured in pregnancy urine was not associated with any preschool behavior score when boys and girls were analyzed together.
Weak evidence: One exploratory WQS analysis, spot urines averaged across pregnancy, mother-reported checklist at about age 3; uncorrected for multiple comparisons.
Source: Nakiwala 2025 (ECHO), Environmental Research Link to this fact
Children and pregnancy / Measured in people or real products
When split by sex, the same mixture index went with about 0.04 T-points higher externalizing behavior in boys (95% CI 0.00 to 0.08) and about 0.05 to 0.06 points lower total and externalizing scores in girls, on a scale where one standard deviation is 10 points.
Weak evidence: Tiny effects, boys' interval touches zero, about 45 uncorrected tests, and a sensitivity subset of 591 children showed positive results in both sexes; association only, not cause.
Source: Nakiwala 2025 (ECHO), Environmental Research Link to this fact
Children and pregnancy / Measured in people or real products
Urinary MEP, the fragrance-solvent (DEP) marker, was one of 23 components in the mixture and was named only as a driver of the girls' externalizing result, which pointed toward lower, not higher, behavior scores.
Weak evidence: MEP in a mixed index cannot separate DEP from the fragrance mixture or other sources; weights come from the text, supplementary figures were unavailable.
Source: Nakiwala 2025 (ECHO), Environmental Research Link to this fact
Children and pregnancy / Other evidence
In the ECHO cohorts, prenatal paraben and benzophenone levels showed stronger associations with eczema in girls than boys (girls: parabens OR 1.21, 1.04-1.42; benzophenones 1.18, 1.00-1.38; bisphenol S 1.21, 1.03-1.43), a pattern the authors call consistent across chemicals.
Weak evidence: Stratified estimates with no correction for multiple tests and no interaction p reported for the eczema comparisons in the main text; subgroup finding.
Source: Miller 2025 (ECHO), Environment International Link to this fact
Children and pregnancy / Measured in people or real products
In an Illinois cohort, mothers with higher ethylparaben (a preservative in some cosmetics) in urine pooled across pregnancy had children with somewhat higher mother-reported ADHD-type scores at age 2 and oppositional-defiant scores at age 3.
Weak evidence.
Caveats
Small cohort (255-285 children per age); over a hundred uncorrected tests, flagged at p<0.1; no effect at age 4; CBCL scales are not diagnoses; association only; parabens are not fragrance ingredients.
Children and pregnancy / Measured in people or real products
The same study found all three parabens measured (methyl, ethyl, propyl) went with lower withdrawn scores in boys at age 2, which the authors read as more outward behavior, not a benefit.
Weak evidence: Sex-stratified post-hoc estimates; ethylparaben's interval included zero; no multiple-comparison correction.
Children and pregnancy / Measured in people or real products
Propylparaben pulled paraben-mixture associations in the opposite direction from ethyl- and methylparaben, so the three-paraben mixture showed only one significant behavior outcome (oppositional defiant at age 3).
Weak evidence: Descriptive pattern within one cohort; mixture methods were applied only to outcomes already linked to a single paraben.
Children and pregnancy / Other evidence
The positive estimates for the pregnancy chemical mixture came mainly from the Cincinnati HOME cohort, which had higher exposure levels (behavior symptom index about 5 points higher per quartile, CI -1.6 to 11.6), while the Canadian MIREC cohort showed essentially none (-0.06, CI -1.65 to 1.52).
Weak evidence: Small HOME subsample (n=139), wide CIs including zero, results as reported in the text rather than checked in the supplement.
Source: Puvvula 2026, Environ Int, Environment International Link to this fact
Children and pregnancy / Other evidence
In a small 2026 Turkish clinic study of 80 teenagers, those with ADHD had higher urine levels of the low-weight phthalate dimethyl phthalate (DMP) than healthy teens, but DMP was found in only about a third of samples.
Weak evidence.
Caveats
One small single-center cross-sectional study; non-detects set to half the detection limit; CI spans 30-fold; ADHD group had much higher family income and education (clinic selection); medication not reported; reverse causation cannot be excluded; no multiple-comparison correction; DMP, not MEP, and not a fragrance-specific finding.
Source: Terzioğlu 2026, Psychiatry Investigation Link to this fact
Children and pregnancy / Review of other studies
The review treats thyroid disruption by endocrine-disrupting chemicals in early pregnancy as a route to fetal neurodevelopmental harm, but it presents no test of that chain and does not cite the Norwegian MoBa analysis in which maternal thyroid hormones did not mediate the DEHP and child ADHD association (indirect OR 1.00).
Weak evidence.
Caveats
Engel 2018 used one mid-pregnancy sample, total hormones, no thyroid antibodies and one outcome, so it does not rule out a thyroid route; the review's chain is plausible but untested.
Children and pregnancy / Other evidence
In a Denver study of 26 children with ADHD and 30 typically developing children, parents rated the ADHD group as having more sensory difficulties on 6 of 7 Short Sensory Profile scales, including taste/smell sensitivity (mean 3.49 vs 4.64, lower is worse).
Weak evidence.
Caveats
Small clinic sample, parent report by parents who knew the diagnosis, no IQ or comorbidity control, one research group; the taste/smell scale is a behavior checklist, not an odor-intolerance test.
Source: Mangeot 2001, Developmental Medicine & Child Neurology Link to this fact
Children and pregnancy / Other evidence
Children with ADHD in this study showed a larger first skin-conductance response to sensory stimuli (smell, sound, light, touch, tilt) than controls, which faded to similar levels over 8 repeats; the overall group difference was only borderline (p=0.056).
Weak evidence: Marginal main effect, responses pooled across five stimuli so the wintergreen odor is not reported separately, sample counts differ between text (26/30) and Table IV (30/31).
Source: Mangeot 2001, Developmental Medicine & Child Neurology Link to this fact
Children and pregnancy / Other evidence
Within children with ADHD (n=19), parent-reported touch sensitivity and sensation seeking went with more aggressive behavior (r about -0.66 to -0.73), so sensory over-responsiveness varies widely among children with ADHD (SSP totals 77 to 170; 24% in the normal range).
Weak evidence: Cross-sectional correlations in 19 children, same parent reporting both scales, no causal direction.
Source: Mangeot 2001, Developmental Medicine & Child Neurology Link to this fact
Children and pregnancy / Other evidence
Father's smoking in pregnancy predicted child ADHD symptoms about as strongly as mother's smoking when the child was related to the father, but not when the father was unrelated (sperm or embryo donor), pointing to inherited liability rather than fetal toxicity.
Weak evidence: Small unrelated groups; results for the paternal analysis stated as available from the first author; single study.
Source: Thapar 2009, Biological Psychiatry Link to this fact
Children and pregnancy / Other evidence
In those schools, each extra liter per second per person of ventilation went with about 2.7-2.9 more percentage points of fifth graders passing state math and reading tests (math 2.85, 95% CI about 0.9-4.8; reading 2.71, 0.5-4.9), after adjusting for district, free-lunch, English-learner, gifted and mobility rates, among the 87 classrooms at or below 7.1 L/s.
Weak evidence.
Caveats
Observational, school-level (no student data), exposure measured on one day, and the 13 schools above 7.1 L/s were dropped before the result became significant; the same analysis on all 100 schools was not significant (P=0.18 math, 0.16 reading). Part of the funding came from HVAC industry groups. Association only; the responsible agent (CO2, bioeffluents, VOCs, temperature) is not identified.
Source: Haverinen-Shaughnessy 2011, Indoor Air Link to this fact
Children and pregnancy / Other evidence
The ventilation rate was not meaningfully correlated with the school's percent free-lunch students (Spearman 0.19), limited-English students (-0.14) or mobility (-0.11), so the authors judged poverty indicators did not explain the ventilation-score link.
Weak evidence: School-level proxies only; district was a strong predictor (about 13 points) and unmeasured differences (teachers, funding, building age) could still confound.
Source: Haverinen-Shaughnessy 2011, Indoor Air Link to this fact
Children and pregnancy / Other evidence
The review reports that in the authors' own child cohort, urinary phthalate exposure was inversely associated with serum T3 and with height attainment, and that DBP and DEHP metabolites were inversely associated with T4 or T3 in two small adult and pregnancy studies.
Weak evidence: Secondhand summary with no numbers; later cohorts show phthalate-thyroid associations of both signs. None of these phthalates is DEP, the fragrance phthalate.
Source: Boas 2012, Molecular and Cellular Endocrinology Link to this fact
Children and pregnancy / Other evidence
Milder ADHD-like symptoms in Swedish 9-year-olds did drift up from 2004 to 2014: mean score about half a symptom higher on a 0-19 scale (1.79 to 2.29) and subthreshold prevalence 10.8% to 14.8%.
Weak evidence: Could reflect real change or parents rating more symptoms with greater awareness; subthreshold trend not significant in the one-twin-per-pair sensitivity analysis (p=.403).
Source: Rydell et al. 2018, J Child Psychol Psychiatry, Journal of Child Psychology and Psychiatry Link to this fact
Children and pregnancy / Other evidence
In a 2020 Taiwanese study of 98 boys with ADHD and 42 control boys aged 6-12, urinary MnBP (a dibutyl phthalate breakdown product) was higher in the boys with ADHD (p = 0.014; about 20 vs 13 ng/mL, approx., read from Fig. 1); girls (32 with ADHD, 26 controls) showed no EDC differences.
Weak evidence: Small single-site cross-sectional clinic-vs-community comparison, one spot urine, p-values only, no multiplicity correction (authors say all findings would become non-significant).
Source: Tsai 2020, Toxics Link to this fact
Children and pregnancy / Other evidence
The same Taiwanese study found no ADHD-versus-control difference in MEP (the diethyl phthalate marker), MMP, MBzP, MEHP, BPA or any measured sex hormone in either sex.
Weak evidence: Small samples (girls n = 32 and 26) limit power; single spot urine.
Source: Tsai 2020, Toxics Link to this fact
Children and pregnancy / Other evidence
The paper reports a paraben difference in ADHD boys inconsistently: the abstract says ethylparaben was higher, the Results say propylparaben (p = 0.032), and Fig. 1 puts the only asterisk on ethylparaben with controls higher.
Weak evidence: Compared across the paper's own text and figure; parabens are preservatives, not fragrance.
Source: Tsai 2020, Toxics Link to this fact
Children and pregnancy / Other evidence
The ADHD consensus reports artificial food dyes were linked to a small increase in child hyperactivity in two meta-analyses (219 and 794 participants), and the second found it only in parent ratings, not teacher or other observer ratings.
Weak evidence: Both meta-analyses are small (under 1,000 participants); effect described as small or very small; rater-dependent.
Source: Faraone 2021, Neuroscience & Biobehavioral Reviews Link to this fact
Children and pregnancy / Other evidence
The ADHD consensus lists one prenatal phthalate study: mothers in the highest quintile of phthalate metabolites had children about three times as likely to have ADHD (Norwegian MoBa nested case-control, 297 cases, 553 controls).
Weak evidence: A single study, not graded for risk of bias; EPA's graded review rated most phthalate-behavior evidence slight or indeterminate; no fragrance or DEP content.
Source: Faraone 2021, Neuroscience & Biobehavioral Reviews Link to this fact
Children and pregnancy / Other evidence
In pregnancy cohorts the review cites, paraben associations with maternal free T3 and free T4 changed with gestational week (methylparaben higher T3 and lower fT4 after about 20 weeks in Puerto Rico and Boston), which the authors read as timing-dependence.
Weak evidence: Repeated-measures cohorts with many comparisons; study labels and reference numbers in the review's tables are inconsistent; association only.
Source: Azeredo 2023, International Journal of Molecular Sciences Link to this fact
Information, not medical advice. See also: myths we won’t tell you.