Original educational illustration, not a chart of comparative effectiveness.
“Active folate” sounds like an obvious improvement. Yet the name of a chemical form does not establish that a product protects health better. A small trial compared 5-MTHF with folic acid during pregnancy and did not detect a difference in its primary biomarker. That does not prove equivalence, and this was not a trial of neural tube defect prevention.
Discussions about folate often merge three questions: does a substance reach the body, does it change a laboratory measurement, and does it improve the health of the mother or baby? Answering one does not answer the others. This study mainly addresses laboratory parameters. Its results should remain at that level rather than acquire an unsupported claim of clinical superiority.
Who was studied?
The trial took place at two primary care centres in Surabaya, Indonesia. Participants were aged 25–35 and were 9–14 weeks pregnant. Eligibility included a BMI from 18.5 to below 30 and haemoglobin above 11 g/dL. Certain gastrointestinal conditions and medications affecting folate were excluded. This was therefore a selected population, not a cross-section of every pregnant patient. Ernawati and colleagues, methods, PDF pages 7–8.
Of 110 randomised women, 56 received folic acid and 54 received 5-MTHF for 12 weeks. Capsules were identical, and the study was double-blind, with outcome assessment and analysis also blinded. These are meaningful strengths. However, the baseline table reports previous prenatal vitamin use in every participant. The population should not be described as previously unsupplemented. Methods and Table 1, pages 8 and 12.
600 and 1,100 micrograms are not a simple dose comparison
One group received 600 micrograms of folic acid daily; the other received 1,100 micrograms of the glucosamine salt of 5-MTHF. The authors report 1.36 and 1.35 micromoles of folate, respectively. These were approximately equimolar amounts. The larger mass includes the chemical preparation: it does not mean that participants received almost twice as much “active folate”. Intervention, pages 8–9.
This distinction matters when reading labels. Salt mass, the amount of the relevant substance and the way a dose is declared are not automatically interchangeable. Trial doses are not an individual dosing schedule for readers. Do not use this comparison to recalculate, increase or replace a prescribed supplement without discussing your circumstances with your clinician.
The primary outcome was red blood cell folate
The primary endpoint was the change in red blood cell folate over 12 weeks. Secondary measurements included plasma 5-MTHF, homocysteine and unmetabolised folic acid, abbreviated UMFA. These are different laboratory parameters, not four independent demonstrations of healthier babies. Outcome hierarchy matters: a more interesting secondary finding should not replace the prespecified primary question. Outcomes, page 10.
The authors report median increases of 262 ng/mL with folic acid and 244.5 ng/mL with 5-MTHF for red blood cell folate. The unadjusted p value was 0.714. The main comparison therefore did not detect a between-group difference. These are the authors’ reported values; interpretation is further limited by unclear confidence-interval reporting in the same table. Table 2, page 13.
No detected difference is not demonstrated equivalence
“No statistically significant difference” does not mean “proven identical”. An equivalence study needs a predefined acceptable range of differences and an appropriate design and analysis. This trial was designed to detect a difference, not formally demonstrate equivalence. It did not specify the relevant equivalence or non-inferiority margin.
This limitation works in both directions. A neutral primary result cannot advertise 5-MTHF superiority. It also cannot establish that the preparations are interchangeable in every respect or equally effective clinically. The narrower conclusion is that this study did not detect a difference in this biomarker, in these participants, over this observation period.
What about higher plasma 5-MTHF?
The secondary plasma finding may attract marketing attention. Reported increases were 3.12 ng/mL in the folic acid group and 5.67 ng/mL in the 5-MTHF group. The unadjusted p value was 0.027, but after adjustment for baseline it was 0.055. Presenting the result as confirmed superiority after adjustment would omit important information. Table 2 and analysis, pages 13–14.
The issue is not a magical divide between two p values. It is the complete analysis, its uncertainty and the relevance of the measurement. Even a clear plasma difference would not itself demonstrate fewer malformations or better pregnancy outcomes. Those questions require those outcomes to be measured rather than inferred from circulating concentrations.
Why the tables need caution
The version reviewed is a Journal Pre-proof, not the final version of record. In Table 2, the reported 95% confidence intervals lack a clearly explained scale connecting them to the raw changes in ng/mL. A trustworthy effect-size chart cannot be reconstructed from those entries. Readers should not silently “repair” them or label them as intervals for the raw difference.
There are also uncertainties in homocysteine reporting and subgroup descriptions. These justify more cautious conclusions, not an accusation of intentional misconduct. In particular, different p values within BMI subgroups do not substitute for a treatment-by-BMI interaction test. They cannot establish a rule that one form works better for a particular body size.
The analysis included 96 women, not all 110
Forty-eight participants in each group completed the study, giving 96 analysed women. Fourteen of 110 were lost from the final analysis, approximately 12.7%. The analysis was per-protocol without imputation, rather than an intention-to-treat analysis retaining all randomised participants. Attrition can weaken the protection provided by randomisation. Participant flow and analysis, pages 10–11.
The sample-size calculation required at least 50 participants per group; 48 were analysed. This does not automatically invalidate every result, but it adds caution when interpreting a neutral finding as ruling out an important difference. A small, short trial also cannot provide a strong answer about uncommon adverse effects.
What was not measured?
The study did not assess neural tube defect prevention, newborn outcomes or children’s development. It was not a trial of preconception effectiveness. Starting supplementation at 9–14 weeks does not make this a direct test of prevention around neural tube closure. The article must not be presented as demonstrating that a particular product better protects the baby.
MTHFR genotypes were not measured either. This trial therefore cannot show that a preparation is better for a particular genetic variant. Personalised claims cannot be added to the results simply because they appear plausible in a broader biological discussion. Mechanism, hypothesis and demonstrated clinical benefit remain distinct levels of evidence.
Funding belongs in the interpretation
The study was funded by Simex Pharmaceutical Indonesia, the manufacturer of the tested preparation. The authors state that the funder had no role in design, data collection, analysis, interpretation or the publication decision, and report no personal financial conflicts. Access to data on reasonable request nevertheless requires sponsor approval. These disclosures should be presented together. Declarations, page 26.
Industry funding is not itself proof of error, but it should not be invisible. The context matters when results may promote a product. The paper gives registration number NCT06935630; the presence of a number alone does not verify that every relevant plan was registered prospectively before recruitment began. That timing was not independently confirmed in this review.
UMFA does not mean proven toxicity
No significant between-group difference in UMFA change was demonstrated. Clinical toxicity was not assessed. A laboratory measurement therefore cannot become a claim about “toxic folic acid”. Detecting a substance does not establish harm, its magnitude or its relevance to an individual. Equally, a neutral comparison over a short period does not demonstrate identical safety.
What should patients take from this?
Do not stop folate or increase the dose because of this comparison. Choice and dosing should reflect individual indications and advice from the treating clinician. This paper does not settle every special clinical situation. It also does not support the claim that ordinary folic acid has stopped working and every pregnant woman must buy another form.
Before accepting a superiority claim, ask what it rests on: a blood concentration, another biomarker or actual health outcomes? Were comparable amounts tested? Was superiority or equivalence properly assessed? Does the account include attrition and funding? These questions are more useful than an appealing adjective on a package.
The defensible summary: a small RCT did not detect a difference in its primary biomarker between approximately equimolar folate preparations. This proves neither equivalence nor superiority in neural tube defect prevention. Reporting uncertainties further limit confidence.
Source
Ernawati and colleagues. Comparative Bioavailability of 5-Methyltetrahydrofolate and Folic Acid in Pregnant Women: A Randomized Controlled Trial. Clinical Nutrition ESPEN, 2026, Journal Pre-proof. DOI: 10.1016/j.clnesp.2026.105107. Educational commentary based on the full text, not individual medical advice or product advertising.