Why do Bangladeshi and Pakistani mothers in England have such different rates of NTDs?
Dr Clare Craig
Key points
- In Pakistan and Bangladesh, neural tube defects affect several times as many pregnancies as in the ethnically English, at broadly similar rates in the two countries.
- However, in England, mothers of Pakistani ethnicity have a rate that has converged on the white rate. By contrast mothers of Bangladeshi ethnicity have nearly three times the white rate of affected pregnancies.
- A high proportion of Bangladeshi mothers were born there e.g. 74.2 percent giving birth in Birmingham in 2010.
- Is this evidence of an environmental exposure in the country of origin?
The rates before migration
Pakistan does not have national birth defect surveillance but where rates of neural tube defects have been estimated they are high. A systematic review of worldwide incidence found five studies from Pakistan reporting between 3.9 and 12.4 per 1,000 births. The highest figure was the largest recorded in the entire Eastern Mediterranean region. Bangladesh also has no national surveillance. One extrapolation produced an estimate of 4.7 per 1,000 live births. At one tertiary referral hospital the measure was 6.4 per 1,000. 94 percent of these mothers reported no periconceptional folic acid. By contrast the rate among the white English mothers is around 1.1 per 1,000 pregnancies.
The rates in England
Congenital anomaly registers for 2006 to 2011, covering the East Midlands, South Yorkshire and the South West, recorded pregnancies affected by neural tube defects including terminations. Across more than half a million births there were 837 affected pregnancies, of which 655 were terminated and 155 ended in a live birth, the remainder died as either stillbirths or late miscarriages.
After adjustment for maternal age and deprivation, mothers of Bangladeshi ethnicity were 2.86 times as likely as White mothers to carry an affected pregnancy. That is around 3.3 per 1,000, close to the rates recorded in Bangladesh itself. The excess was for spina bifida only not anencephaly (a failure of brain development). For Pakistani mothers the adjusted rate ratio was 1.12 with the actual rate around 1.3 per 1,000 pregnancies. If there was a real excess it was far smaller. There was marked geographical variation, with preliminary West Midlands data putting the Pakistani rate at more than five times the White rate. That difference has never been explained.
The difference between ethnicities in birth rates is heavily skewed by termination. Eighty six percent of White mothers ended an affected pregnancy compared to 56 percent of Pakistani mothers. For spina bifida alone the figures were 81 percent compared to 33 percent.
Based only on births, the Pakistani mothers had three times the risk of white mothers. However, this difference almost disappears once terminations are accounted for. For Bangladeshi mothers the threefold excess remains.
The diets of Pakistani and Bangladeshi immigrants do differ with wheat from chapatti flour far more important in Pakistani diets and rice more dominant in Bangladesh. Some experts have already suggested adding folic acid to rice in addition to our flour, on the grounds that minority ethnic groups who do not regularly eat wheat based foods will otherwise get no benefit.
Comparing Pakistan and Bangladesh
Both countries have a similar background risk. Why have Pakistani mothers in these English regions adopted the white rate while Bangladeshi mothers have kept most of the home country excess. Both communities still include many mothers born abroad, but the proportions differ: in Birmingham in 2010, 74.2 percent of Bangladeshi mothers with a recorded country of birth were born outside the United Kingdom (446 of 601), compared to around 57 percent of Pakistani mothers. One important question is whether the exposure risk is happening before or after migration.
Proposed explanations
Vegetarianism has been proposed as one explanation for an excess of neural tube defects among women of Indian ethnicity (who have an 80 percent increased risk compared to white women). The argument was based on a lack of vitamin B12 rather than folate as animal foods are the main source of B12. The logic was that B12 deficiency produces a functional folate deficiency.
Either way, this explanation cannot explain the Pakistani Bangladeshi discrepancy as both these groups eat meat and fish. Other factors that may play a part include deprivation, periconception folic acid and consanguinity. Adjusting for deprivation barely moved the estimates. There is suggestive evidence that both communities have poor awareness or uptake of periconceptional folic acid. Consanguinity may be a factor as there is an increased risk of genetic anomalies which can include neural tube defects and these cannot be related to the folate pathway. Born in Bradford found around 60 percent of British Pakistani parents were related, while Genes and Health, whose east London cohort is majority British Bangladeshi, finds around a third of participants are the offspring of second cousins or closer. A Birmingham study reported 49.9 percent among Pakistani mothers and 20.8 percent among Bangladeshi mothers, compared to 1.3 percent among White European mothers. Consanguinity has been put forward as a likely contributor for the incidence among the Bangladeshi group but if that were the explanation the effect should be more marked among the Pakistani cohort and the opposite is true.
There are two heavy metal exposures worth considering in this context: arsenic and mercury. Bangladesh has the highest environmental arsenic exposure globally. When the body disposes of arsenic it consumes methyl groups leaving less for gene methylation in the developing embryo. Bangladesh is unusual for the scale of arsenic exposure from groundwater, and rice is another important source (including rice that is imported to the UK and eaten here). The national survey by the British Geological Survey and the Department of Public Health Engineering sampled 3,534 boreholes across 61 of the 64 districts in Bangladesh and found 46 percent above the World Health Organization guideline of 10 micrograms per litre. Nobody has measured arsenic in British Bangladeshi women. Mothers that grow up in Bangladesh before migrating will be at risk of exposure from water. As well as ground water, rice is often contaminated with arsenic and Bangladeshi rice is imported to the UK and consumed by Bangladeshi migrants. Crucially the body handles arsenic by adding a methyl group before disposing of it and this depletes the folate-methylation pathway such that normal methylation of genes could be inhibited.
Mercury is also disposed of via methylation and could have the same impact on depleting the ability of a developing embryonic brain to methylate crucial developmental genes. Skin lightening creams, which may contain mercury, sometimes at high concentrations, are widely used across South Asia. Bangladesh’s government estimates 521 kg of mercury a year entering skin creams, with products on sale there containing between 3,450 and 4,653 parts per million. Creams made in Pakistan have even higher concentrations at up to 33,970 parts per million. A survey of women in Bahawalpur found 51.5 percent used such whitening creams. There is no data on usage in Bangladesh or among migrants to the UK. There is potential for exposure in the UK with London Trading Standards finding mercury in 33 of 87 samples of skin lightening products. These products end up being seized under Trading Standards law but many remain on sale in small stores. Furthermore, medicines used in Pakistan and Bangladesh still sometimes contain mercury. A study of mercury levels in women of reproductive age could be enlightening.
The comparisons that should be made
Three simple comparisons could shed light on the causative factors in these cases. Firstly, UK based Bangladesh born mothers could be compared to British born Bangladeshi mothers to look for evidence of the significance of an exposure in early life in Bangladesh. Secondly, the rates among Bangladesh born mothers in Britain compared to women who remain in Bangladesh could test for the effects of migration and any ongoing environmental exposures. Finally, measuring Bangladeshi mothers in Canada or Australia compared to Britain could test whether folic acid fortification alters the excess. More detailed studies could follow quantifying arsenic exposure through imported rice and mercury exposure from face cream usage or even based on blood and hair testing.
Ethnic differences in neural tube defects have previously been attributed to differences in folic acid uptake, deprivation and consanguinity. Examining the situation more closely reveals two communities one of which has lost its excess after migration and one of which has not. The crucial comparison of the excess among first generation Bangladeshi mothers compared to the British born cohort, has never been done. Surely, the answer to this question should be sought before mass medicating the entire population.
