Periconceptional maternal diet quality influences blood heavy metal concentrations and their effect on low birth weight: the Japan Environment and Children's Study

Okubo and Nakayama (2023) analysed 72,317 mother-child pairs from the Japan Environment and Children’s Study (JECS) to test whether periconceptional diet quality, scored using a modified Balanced Diet Score (BDS) anchored to the Japanese Food Guide Spinning Top, modifies both maternal blood metal concentrations measured in the second/third trimester and the association between blood Pb and low birth weight (LBW, birth weight <2,500 g). Higher periconceptional BDS was associated with modestly lower blood Pb (adjusted β = −0.03 per 10-point BDS increase, 95% CrI −0.04, −0.03) and Cd (adjusted β = −0.010, 95% CrI −0.016, −0.004) but higher Hg (adjusted β = +0.05, 95% CrI 0.05, 0.06), the Hg signal reflecting greater fish and shellfish consumption in higher-quality diets. In Bayesian logistic regression adjusted for maternal age, height, pre-pregnancy BMI, gestational weight gain, parity, education, smoking, alcohol, folate supplementation, physical activity, gestational age, and child sex, each log2-doubling of blood Pb was associated with elevated LBW risk (adjusted OR 1.50, 95% CrI 1.36, 1.64); equivalently, Q4 vs Q1 blood Pb (≥7.33 ng/g vs ≤4.71 ng/g) adjusted OR was 1.44 (95% CrI 1.34, 1.56). Higher BDS attenuated this association (BDS × log2-Pb interaction OR 0.991, 95% CrI 0.983, 0.999; multiple-imputation sensitivity OR 0.993, 95% CrI 0.986, 0.999), with the effect more visible at higher blood Pb concentrations. No effect modification of BDS was found on Cd-LBW or Hg-LBW relationships.

Key numbers

Blood metal concentration medians (ng/g whole blood) stratified by balanced diet score quartile (Table 2):

Effect size on the original (ng/g) scale per +10-point BDS: Hg +1.04, Pb −0.98, Cd −0.99 ng/g.

LBW outcomes (Table 3, n=72,317, overall LBW prevalence 7.6%):

Effect modification by diet quality on the blood Pb–LBW association:

• BDS × log2-Pb interaction OR: 0.991 (95% CrI 0.983, 0.999)
• Multiple-imputation sensitivity analysis: OR 0.993 (95% CrI 0.986, 0.999)
• Predicted LBW probability at high blood Pb (Fig. 2B) is visibly lower in the highest BDS tertile (median 54.9) than the lowest (median 39.7); no clear modification by BDS for Cd or Hg.

Analyte detection limits (whole-blood ICP-MS, Agilent Technologies, Tokyo, Japan):

• Hg: 0.049 ng/g
• Pb: 0.129 ng/g
• Cd: 0.0234 ng/g

All measured whole-blood concentrations of Hg, Pb and Cd were above the analyte LOD. BDS observed quartile cutoffs span 13.1 (Q1 lower bound) to 70.0 (Q4 upper bound, which is also the theoretical maximum); median 48.0 (IQR 42.3–53.0); central 98% of the distribution (1st–99th percentile) is 27.7–62.6 (paper §3.1, p.5). Hg measured as total Hg (no speciation); reported here as tHg per wiki convention.

Methods (brief)

Whole-blood samples (200 µL) collected from a peripheral vein at a prenatal-care visit during the second/third trimester were diluted 1:5 (v/v) in a buffer of 2% (v/v) 1-butanol, 0.1% tetramethylammonium hydroxide, 0.5 g/L polyoxyethylene (10) octylphenyl ether, and 0.5 g/L ethylenediaminetetraacetic acid, then vortex-mixed and analysed by inductively coupled plasma–mass spectrometry (Agilent Technologies, Tokyo, Japan; instrument model not specified in this paper, methods described previously in Nakayama et al. 2019). Pb, Cd and tHg reported as wet-weight ng/g whole blood.

Habitual diet was assessed via self-administered food-frequency questionnaires (FFQs) at the first trimester (covering the year before enrolment, used here as the periconceptional measure) and again at the second/third trimester. The Balanced Diet Score (BDS), modified from the Japanese Food Guide Spinning Top scoring system per Kuriyama et al. 2016, summed scores across seven components — grain dishes, vegetable dishes, fish-and-meat dishes, milk, fruits, snacks-and-alcoholic-beverages, and sodium from seasonings — each scored 0-10 proportionally, for a total range of 0-70. Upper cut-offs for over-consumption of the five dish categories were dropped for the pregnant-women population. LBW defined as birth weight <2,500 g.

Statistical analysis: Bayesian inference of univariate and multiple linear regression for diet-metal associations; Bayesian logistic regression (R package brms) for LBW risk, with blood metals categorised by quartile and modelled continuously as log2-transformed concentrations. Effect modification by BDS tested via interaction term. Adjusted models included maternal age, height, pre-pregnancy BMI, gestational weight gain, parity, education (<13, 13-14, ≥15 years), smoking status, alcohol drinking, routine folate supplementation, physical activity (low/moderate/high), gestational age at birth, and child sex; metal-on-LBW models added BDS as a confounder. Multiple imputation by chained equations (R miceRanger, m=10, random-forest, majority-vote method) handled missing covariates (11.0% of mothers had one or more missing values). STROBE-compliant reporting.

Implications

Certification: This is one of the largest prospective birth cohort studies (n=72,317) demonstrating a Pb-LBW dose-response across blood Pb concentrations in the 1.2–110 ng/g range, with significant LBW associations down to the second quartile of blood Pb (4.72–5.85 ng/g; adjusted OR 1.18, 95% CrI 1.07, 1.30 vs Q1 ≤4.71 ng/g). The per-doubling adjusted OR of 1.50 means each doubling of maternal blood Pb is associated with roughly 50% greater odds of LBW; the diet-quality interaction attenuates but does not eliminate this risk. For HMTc literature-baseline use, this study contributes prospective-cohort evidence that maternal blood Pb below the former CDC blood-Pb level of concern (50 ng/g = 5 µg/dL) is associated with measurable LBW risk.

Courses: Useful case study for the interaction between dietary quality and metal toxicity. The Hg-fish-consumption pattern illustrates a co-exposure trade-off: higher BDS reflects more fish and shellfish, which raises Hg but lowers Pb and Cd through better mineral status (Ca, Fe, Zn) and lower intestinal Pb/Cd absorption. Relevant to prenatal-nutrition modules and to dietary-counselling content that addresses the fish-Hg trade-off in pregnancy.

App: Diet-quality × metal interactions are not directly modelable from packaged-food ingredient lists, but the JECS Pb-LBW dose-response provides a real-world calibration anchor for blood-Pb exposure in a low-exposure Japanese population (BDS-stratified blood-Pb medians 5.71–6.05 ng/g). The reported per-doubling OR is a useful sanity check for any future model that estimates LBW-relevant Pb exposure from food contamination data.

Verification notes

Merge-enhanced 2026-05-18 from the PCMF PDF (raw/Papers Cube Manual Fetch/Periconceptional maternal diet quality influences blood heavy metal concentrations and their effect on low birth weight: the Japan Environment and Children's Study.pdf, SHA-256 bd8cbc44...264f). Page was created 2026-05-14 with raw_handle: papers-cube placeholder and was in routing_malformed.csv (advisory: empty optional routing fields).

Defects corrected:

• Title: prior page used “and the association between blood lead levels and low birth weight”; published Environment International title is “and their effect on low birth weight: the Japan Environment and Children’s Study”. Corrected.
• raw_handle: papers-cube placeholder → PCMF_periconceptional-maternal-diet-quality-influences- (PCMF_ convention, 50-char window matching sibling pages).
• raw_path: filename in prior raw_path (“…and the association between blood lead levels and low birth weight.pdf”) did not exist on filesystem; corrected to actual PDF filename in PCMF folder.
• Pb-LBW OR label: prior page asserted “adjusted OR=1.50 per quartile Q4 vs Q1, 95% CI 1.36-1.64” — this conflated two distinct statistics. Per Table 3, the Q4 vs Q1 adjusted OR for blood Pb is 1.44 (1.34, 1.56) and the per-log2-doubling adjusted OR is 1.50 (1.36, 1.64). Both now reported separately with correct labels.
• “23% lower odds of Pb-associated LBW” for BDS Q4 vs Q1: not supported by Table 3. Q4 BDS adjusted OR for LBW = 0.87 (0.79, 0.96), which is a 13% lower-odds figure, not 23%. Removed.
• Methods “Agilent 7700” model: paper specifies “Agilent Technologies, Tokyo, Japan” without instrument model and refers to Nakayama et al. 2019 for full method detail. Model number removed to avoid fabricated specificity.
• Hg → tHg: paper measures total Hg without speciation (no separate MeHg). Per Part 14 convention, the speciated label tHg is used on this page; metals frontmatter updated.
• access_url and raw_sha256 added for provenance.

Preserved per v2 skill rule: cite_key, doi, evidence_tier: A, license: CC BY-NC-ND 4.0, source_type: peer-reviewed, jurisdictions: [JP], sample_n: 72317, near_duplicates: [], matrices: [blood, maternal-blood], products: [], ingredients: [] (paper measures blood biomarkers and dietary patterns, not food contamination; product/ingredient frontmatter correctly empty — advisory-only routing_malformed entry expected).

Audit subagent (2026-05-18, fresh-context general-purpose): REVISE verdict. Three findings; two applied, one rejected as false positive.

• ⚠️ Check 1 BDS-range phrasing — applied. Tightened “BDS range across the cohort: 13.1–70.0” to distinguish observed quartile cutoffs (13.1–70.0) from the 1st–99th percentile range (27.7–62.6, per §3.1 p.5).
• ⚠️ Check 5 “supports stringent dietary-Pb minimisation” — applied. Reworded Implications-Certification closing to data-contribution framing per Part 2 (literature-reports, does-not-advocate).
• ❌ Check 3 “Agilent 7700 instrument-model regression” — rejected as false positive. Re-read of PDF p.2 §2.2 confirms the published paper text reads “…inductively coupled plasma-mass spectrometry analysis (Agilent Technologies, Tokyo, Japan)” with no model number. The subagent’s quoted text “(Agilent 7700, Agilent Technologies, Tokyo, Japan)” does not appear in this paper; the model “Agilent 7700” likely came from a citing reference (Nakayama et al. 2019 is cited for full method detail) and was incorrectly attributed to the present paper. The wiki page’s correction (removing “Agilent 7700”) is faithful to the source.

Wiki pages updated on ingest

• lead
• cadmium
• mercury
• vulnerable-populations
• reproductive-outcomes

Page history

The five most recent substantive edits to this page. The full version history lives in git; when DOI minting comes online (see schema docs), each entry below will also link to a version-pinned DataCite DOI.