Fathers' vitamin B12 before conception tracked with birth defects too

Design

Prospective cohort study with weighted logistic regression and standardised predicted prevalences (Shanghai Preconception Cohort)

Population

3,032 couples sampled within four months before conception (73 birth defect cases) and 17,765 women sampled at or before four months' gestation (327 cases)

Primary outcome

Birth defects among live births, stillbirths and abortions for fetal abnormality, by parental red-cell folate and vitamin B12

Effect

Maternal preconception B12 lowest to highest category: 49.6 to 16.2 cases per 1000 pregnancies; paternal 55.5 to 24.0; postconception red-cell folate 906-1131 nmol/L 16.5 vs 25.7 below 453

Folic acid before conception prevents neural tube defects and that is settled. What vitamin B12 does, and whether the father's status matters at all, is not. The Shanghai Preconception Cohort measured red-cell folate and B12 in 3,032 couples within four months before conception (73 birth defect cases) and in 17,765 women at or before four months' gestation (327 cases).

Higher preconception maternal B12 was associated with fewer birth defects in a dose-response pattern: predicted prevalence fell from 49.6 cases per 1000 pregnancies in the lowest category (under 148 pmol/L) to 16.2 in the highest (590 pmol/L or above). Paternal B12 showed the same direction, attenuated - 55.5 down to 24.0 per 1000. Preconception red-cell folate showed imprecise patterns. After conception, maternal red-cell folate between 906 and 1131 nmol/L was associated with 16.5 cases per 1000 against 25.7 below 453 nmol/L, with little further reduction at higher levels.

This is observational and the authors name residual confounding as the limitation. B12 status tracks with diet, socioeconomic position, and much else that independently affects pregnancy outcomes, and a paternal association is exactly the kind of finding that shared household diet could manufacture.

Still, two things are usable. The first is the plateau in the folate curve: more is not better above about 1100 nmol/L, which argues against the high-dose supplementation that has crept into private antenatal practice. The second is that B12 deficiency is common in India, particularly with vegetarian diets, and is not routinely checked before conception even where folic acid is prescribed reflexively. Testing and correcting it preconceptionally is low cost and defensible on other grounds already; this adds a reason.

• Check B12 preconceptionally in vegetarian or vegan patients rather than assuming folate covers it
• Continue standard folic acid - nothing here supports higher doses, and the curve plateaus
• Treat the paternal association cautiously: shared diet is the obvious alternative explanation
• Correct documented B12 deficiency before conception, not in the second trimester
• Metformin and long-term proton pump inhibitors lower B12 - ask about them at the preconception visit

The statistics, in plain English

These are predicted prevalences from a weighted logistic model, not observed rates - the model estimates what the rate would be at each biomarker level after adjustment, which makes the numbers cleaner than the raw data and more dependent on the model being right. With 73 cases among 3,032 couples, the preconception estimates rest on very few events, which is why the folate patterns are described as imprecise. A dose-response gradient is more persuasive than a single cut-off comparison, but observational dose-response can still be produced entirely by a confounder that also follows a gradient.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.