A fall, a plateau or a rise in neural tube defects could all be presented as success
Dr Clare Craig
Key points
- No protocol has been published for establishing whether mandating folic acid to the entire population reduces neural tube defects or causes harm.
- The baseline rate of neural tube defects is not stable. It rose and fell by as much as 80 percent within twenty years, for unexplained reasons.
- Because around 80 percent of identified affected pregnancies are terminated the count is dependent on the diagnostics in early pregnancy. English and Scottish researchers attribute a recent rise in recorded anencephaly to changes in its detection.
- Folic acid recommendations, as opposed to mandates, across Europe, Israel and South America did not alter the existing trend. In many cases existing declines in neural tube defect rates slowed after folic acid was mandated in food.
- Although the Scientific Advisory Committee on Nutrition made its support conditional on controlling voluntary fortification and monitoring the population, the regulations have no requirement for either and they do not specify any monitoring requirements.
- No harms have been identified for monitoring, although there are several that would be obvious candidates.
- Because recorded anencephaly is already rising, a fall, a plateau or even a rise after December 2026 could each be presented as consistent with fortification having worked.
Folic acid is being added to non-wholemeal wheat flour with a mandate for all companies beginning on 13 December 2026. HART are opposed to this mandate. If it goes ahead, there are serious questions to be asked about the plans to monitor its effects.
Historically the numbers of neural tube defects have been declining. Other countries have used the background declining numbers as a way of claiming success while assuming that the rates would, in the absence of folic acid fortification, have been constant. The numbers in the UK have also been declining for decades so a similar analysis would be possible, although the rates per 1,000 babies born each year have plateaued.
However there was a slight recent rise in cases of anencephaly (a failure of brain development) which has been attributed to better diagnostics leading to diagnosis of babies that would otherwise have miscarried. This reasoning could be used to ensure that any result of mandating folic acid could be spun into a positive PR story. If the recorded rate of neural tube defects falls, folic acid will of course be credited. If it holds steady, fortification will be said to have prevented a rise that would otherwise have happened. If the rate rises, the changing rates of detection will be blamed with a claim that it “would have been so much worse”. Furthermore, there is evidence that terminations are not always captured in the registry data and a change in the rate of recording could similarly be used as a way of claiming “it would have been so much worse”.
This situation could have been prevented by setting out expectations and a methodology for fair measurement in advance but this has not been done.
The rise in anencephaly
Overall the neural tube defect rate per 1,000 births, which is dominated by spina bifida, has held steady for the last decade in Scotland and in England. However, anencephaly has risen in Scotland by 12 percent a year from 2013 to 2021. In England the same rise appears from 2015 to 2019. The usual assumption is that both spina bifida and anencephaly have the same underlying cause so if the rate of one changes while the other remains steady there may be a change of measurement to account for it.
The early 2010s saw a switch from blood tests in the second trimester to first trimester combined screening which, if positive, would have led to earlier ultrasound confirmation. Where a pregnancy may previously have ended in an undiagnosed miscarriage, with this new earlier testing it would be diagnosed and counted.
Without a method set in advance to separate the changes in detection from changes in underlying incidence, a fall after 2024-2026 could be incorrectly attributed to folic acid fortification.
The underlying rate is not stable
Before antenatal screening even existed the rate in Birmingham nearly doubled, almost halved, then doubled again within twenty years in the absence of any interventions. This same pattern of epidemic waves was recorded in Dublin, Boston and Providence. The cause has never been identified. The Department of Health hopes folic acid in flour will prevent around 20 percent of affected pregnancies. The variation over a decade has often naturally exceeded that in the past.
Having an 80 percent chance of a study detecting a 20 percent reduction in neural tube defects would require seven years of cumulative data after full implementation according to Public Health Scotland. The modelling behind the policy predicts a reduction of 15 to 22 percent. A 15 percent reduction would be even harder to detect.
What other countries found
Whenever attempts are made to measure the impact of flour fortification the background trend in the rate dominates any measurement. When data was pooled from congenital anomaly registries across Europe, Israel and South America, folic acid recommendations for pregnant women (as opposed to mandates) had no impact on the trends up to six years after.
When 1.3 million births were analysed after folic acid fortification in California neural tube defects were already falling by 8.7 cases per 100,000 births each year before the intervention and the decline slowed afterwards.
Analyses that have claimed a reduction as a result of a mandate have excluded terminations at a time when they were increasing and have ignored background falling rates. The Canadian study reporting a 46 percent reduction compared a pre-fortification period with a full-fortification period, treating the earlier rate as though it would otherwise have stayed where it was.
Given these issues, it is paramount that a fair methodology for assessment of the policy should be set out in advance.
Nothing has been specified in advance
The Bread and Flour (Amendment) (England) Regulations 2024 detail what is expected of manufacturers, the minimum dosage in flour and how the requirement will be enforced with criminal threats for millers. Nothing is included about how anyone will establish whether the intervention worked or what negative outcomes it caused.
No clause committing to a review after a certain period is included. There are no details specified for monitoring and no duty to report on any conditions. No methodologies for assessing outcomes, whether benefits or risks, are specified. There is no commitment to publish any data or analysis.
Ministers have stated the intention that monitoring for benefits will take place and that the level of fortification will be reviewed but there is nothing to require any of that. Moreover, there is no intention regarding harm.
The government’s own modelling sets out that at 250 micrograms per 100 grams of non-wholemeal flour with no cap on other sources, 4.4 percent of children aged 1 to 3 exceed their age-specific upper limit. Where a cap was enforced that fell to 0.8 percent. No cap was included in the regulations.
The harms that should be measured
There is no shortage of evidence of harm that could be measured:
- Miscarriage and fetal death. The only randomised trial of a modest dose in low risk women recorded an excess of total fetal deaths. Infant mortality and later deaths were higher in the treatment group but numbers were small and it was not statistically significant. Nevertheless, it would be reckless not to monitor it.
- Gestational diabetes, which tripled in Australia after flour fortification began.
- Multiple births and the associated complications which rose after fortification in more than one country. The rate of twins also rose significantly in the randomised trial, and the increase persisted after excluding women who had taken fertility drugs.
- Immune function, because unmetabolised folic acid circulating in the blood is associated with around 23 percent lower natural killer cell activity which increases risk of viral infections and complications as well as cancer.
- Cancer incidence and cancer mortality. Each cancer should be followed separately.
- Dementia and cognitive decline, and peripheral neuropathy, which follow from folic acid masking a vitamin B12 deficiency while the nerve damage continues.
Children are the group most at risk of over exposure and should be most closely monitored for adverse effects. Follow-up of the babies born in the randomised trial found significantly more atopic dermatitis, asthma and wheezy bronchitis in the first year, and significantly more middle ear infections by the age of six. Most of the affected children had no family history of these conditions.
What should happen now
The outcomes that would reveal harm will take decades to appear, for example colorectal cancer. There has been a failure to establish a pre-fortification baseline series for comparison and changes to screening are being introduced that will muddy the water. This is a national experiment without safeguards and with no way of measuring the results.
A protocol published now on how any benefit will be measured and compared to the baseline and on how assessment of harm will be made would cost nothing.
