Obstetrics, Childbirth and Child Health

Nigeria has one of the highest maternal mortality burdens in the world and a large share of Yoruba births, particularly rural ones, are attended by traditional birth attendants rather than by skilled providers. That combination is the reason this is the most consequential topic in the section: more people die from what happens or fails to happen in this domain than in all the others together.

The evidence here supports a specific and uncomfortable conclusion. Traditional birth attendants handle uncomplicated deliveries adequately, cannot handle the complications that cause maternal deaths, and the decisive variable is not their technique but whether and how fast they refer. Programmes that have improved outcomes have done so by changing referral, not by changing what happens at the birth.

The traditional birth attendant

Agbẹ̀bí is the Yoruba word, from gbà, to receive or catch, plus bí, to give birth: the one who receives the child. The etymology describes the role accurately.

The practice includes abdominal massage and external manipulation to assess and adjust fetal position, management of labour position, delivery, cord cutting and care of the newborn, herbal preparations given during pregnancy and in labour, and the ritual and social management of birth including the observances that lead to the naming ceremony. Attendants are overwhelmingly women, are frequently older women of standing in the community, and typically learned by apprenticeship, often within a family, sometimes reporting a call or dream as the origin of the vocation.

The social position matters for why the practice persists. The attendant is local, known, available at night, paid in instalments or in kind, speaks the language in every sense, and does not require the woman to travel or to be treated by strangers. Against a distant, expensive, and often unwelcoming facility, this is not an irrational choice.

What the outcome evidence shows

The historic policy position and its reversal. Between 1970 and 1990 the World Health Organization promoted and funded the training of traditional birth attendants as a strategy to reduce maternal and neonatal mortality 1. That policy was subsequently abandoned in favour of skilled birth attendance, on the ground that the trials did not show the mortality reductions expected. The current summary of the evidence base is that support for TBA training is limited but promising for reducing maternal deaths 1, which is a much weaker statement than the enthusiasm of the 1970s and a stronger one than the outright dismissal of the 2000s.

Knowledge is better than practice. A study of traditional birth attendants in Osun State examined knowledge and practice for postpartum haemorrhage, the leading direct cause of maternal death. Most attendants were knowledgeable about the causes and warning signs of postpartum haemorrhage, but preventive practices were inadequate and management practices poor 1. That gap between knowing and doing is the characteristic finding in this literature, and it is what training programmes have to close.

Integration can move the numbers. The most instructive Nigerian case is the Agbebiye programme in Ondo State, a Yoruba state, which incorporated traditional birth attendants into the formal maternal health system rather than trying to displace them. The published evaluation reports 5,606 traditional birth attendants incorporated, 14,124 referrals against 142,206 facility deliveries, a referral rate of 9.9 percent, and a 61.8 percent increase in facility births from 33,077 in 2013 to 53,531 in 2016 2. Seven maternal deaths were linked to Agbebiye-registered attendants out of 260 facility deaths statewide during implementation, 2.7 percent 2.

Read that carefully, because the important number is not the referral rate. It is the increase in facility births. The programme's demonstrated achievement is that it moved tens of thousands of deliveries into facilities by making the traditional attendant a route into the system rather than an alternative to it. That is a different theory of change from training attendants to deliver better, and on the evidence it is the one that works.

Attendants can deliver specific clinical functions when supported. A three-arm cluster randomised controlled trial in southwest Nigeria tested training plus case-manager support for traditional birth attendants in linking HIV-positive pregnant women to care 3. This is a narrow function, but it demonstrates the general principle: an attendant embedded in a supported system can perform a defined task that reduces harm.

What helps, what is neutral, what harms

The honest division, stated plainly.

Helps. Presence at the birth of someone competent and trusted, which reduces the number of unattended deliveries. Continuity of care through pregnancy. Recognition of danger signs where the attendant has been trained to recognise them. Referral, where a functioning referral route exists. Positional freedom in labour, which the evidence in obstetrics broadly favours over the supine position. Immediate skin-to-skin contact and early breastfeeding, which are traditional practice and are also current WHO recommendations.

Neutral. Most of the ritual and social apparatus around birth, which does no clinical harm and does real social work, including the naming observances and the recognition of the birth by the lineage.

Harms. Four categories, and they are the ones that kill.

Delay in referral. The dominant cause of maternal death in this setting is not that the wrong thing was done at the birth but that the right thing was done too late somewhere else. Obstructed labour, haemorrhage, eclampsia and sepsis all have short windows. An attendant who persists with her own management, or who refers to a facility hours away with no transport, is the proximate mechanism of a large share of Nigerian maternal deaths.

Herbal preparations in pregnancy and labour. This is the least investigated part of the practice and there is reason for concern. Preparations given to accelerate labour are pharmacologically uterotonic in intent, and an uncontrolled uterotonic in an obstructed labour causes uterine rupture. Abortifacient properties have been demonstrated for commonly used African indigenous antimalarial plants in pregnant mice, with explicit implications noted for maternal and fetal health 4. The plants are not inert; see file 03.

Non-sterile technique. Cord cutting with unsterilised implements and application of substances to the cord stump cause neonatal tetanus and omphalitis. This is a well-understood and entirely preventable mechanism.

Manipulation. External manipulation for malpresentation, performed without imaging or monitoring, can cause abruption and rupture.

The twinning rate and its actual explanation

The Yoruba have the highest rate of twinning recorded in any human population, and this is one of the few places where a striking popular claim about Yoruba distinctiveness is straightforwardly true. It is also routinely explained wrongly.

The figures. Nylander's studies in the late 1960s and early 1970s documented twinning rates of 45 to 53 per 1,000 maternities in southwestern Nigeria 56. Comparing mothers of different ethnic origin delivering in two Ibadan hospitals, he recorded rates of 45, 45, 31 and 21 per 1,000 maternities for mothers of Western, Eastern, Mid-Western and Northern Nigerian origin respectively 6. A later analysis of hospital data from 1995 to 2004 across four southwestern towns recorded 46.5 per 1,000 at Ilesa, 46.2 at Ile-Ife, 38.5 at Ogbomoso and 22.1 at Ado-Ekiti, an overall average of 40.2 per 1,000, which the authors state ranks among the highest recorded rates in the world 7. Igbo-Ora in Oyo State is the town most often cited, at 45 per 1,000, roughly one twin birth in 22 8. The Yoruba rate is more than fourfold that of European-descended populations 5.

The actual explanation, in three parts.

It is dizygotic, not monozygotic. This is the single most important fact and it is the one popular accounts omit. Nylander's analyses of placentation and zygosity established that the excess is entirely in dizygotic twinning, while monozygotic rates, determined by sex, placentation, blood groups and other genetic markers, vary very little between the ethnic groups compared 56. Monozygotic twinning is a near-constant of about 3 to 4 per 1,000 in every human population; dizygotic twinning varies enormously. So the question is never "why do Yoruba women have more twins" but the much narrower "why do Yoruba women release more than one ovum per cycle more often."

The mechanism is hyperovulation, and it is hormonal. Nylander reported that mean follicle-stimulating hormone levels were substantially higher in western Nigerian women with two sets of twins than in those with one set 6. Higher FSH produces multiple follicular recruitment, which produces multiple ovulation, which produces dizygotic twins. The pathway is understood; what drives the FSH difference is not settled.

Maternal age and parity contribute but do not explain it. Dizygotic twinning rises with maternal age everywhere, peaking in the late thirties. The southwestern Nigerian data show the highest occurrence in the 25 to 29 age group and the lowest in the 45 to 49 group 7, which does not follow the standard age curve and indicates that age structure is not the driver.

The dietary hypothesis, handled honestly. The most repeated explanation is that yams or cassava contain phytoestrogens that stimulate the ovaries. It is not established.

What is documented: community members in Igbo-Ora themselves most frequently name ìlasa, a soup made with okra leaves prepared with local water, together with àmàlà made from cassava, as the likely dietary cause; the qualitative study found the community emphasis to be on ìlasa rather than on yam, which is the version that circulates internationally 8. Nylander proposed in 1978 that dietary substances might raise FSH, and this remained speculative 8. The 2020 study's conclusion is that although several postulations exist, no study has been conclusive on a definite causative agent, and that neighbouring communities consume the same foods while showing lower twinning rates, which points towards environmental or epigenetic factors requiring directed investigation 8.

So the honest statement is: the excess is real, it is dizygotic, it runs through elevated FSH and hyperovulation, it has a familial component, and the specific environmental or dietary trigger is unidentified despite fifty years of looking. Anyone who tells you it is the yams is stating a hypothesis as a finding.

Why the rate matters clinically. Twin pregnancy carries substantially higher perinatal mortality, and Nylander's work also contributed to the study of perinatal mortality in multiple gestation 5. A population with a twinning rate above 40 per 1,000 delivered largely outside facilities has a serious obstetric problem, because twin deliveries are precisely the ones that need skilled attendance for malpresentation, second-twin complications and postpartum haemorrhage.

Ìbejì. The elaborate Yoruba cult of twins, the ère ìbejì carvings and the observances for a deceased twin, are treated in 06-orisa/16-ibeji and 07-arts/04-ibeji. The connection worth making here is causal in the right direction: the extraordinary cultural elaboration around twins is a response to their extraordinary frequency, plus the historically high mortality among them. A society where one birth in twenty-two produces twins, and where twins died often, is a society that will develop institutions around twins.

Infant and child care

Breastfeeding is prolonged and near-universal, with early initiation, and this is straightforwardly protective.

Carrying on the back with a wrapper provides continuous contact, thermoregulation and responsive feeding. There is no evidence of harm and good reason to think it beneficial.

Ọmọ tuntun practice. The newborn period includes bathing, massage and stretching of the infant's limbs, which is widely practised and for which this compiler located no outcome evidence in either direction. Vigorous manipulation of infants carries theoretical risk; the practice has not been evaluated.

Substances applied to the cord are a documented cause of neonatal infection, as above.

Herbal preparations given to infants are common, are given for conditions including ibà, fever, and convulsions, and are of particular concern because dosing in an infant is unforgiving and because the contamination problems in the commercial herbal trade documented in file 12 apply with full force to a two-kilogram patient.

Ilà, facial marks. Cutting performed by the olola, treated in 07-arts/09-body-arts. As a medical matter it is a clean-instrument question and a tetanus question, and the practice has declined sharply.

What follows

The direction the evidence supports is not the elimination of the traditional birth attendant, which is not achievable in the medium term and which the WHO's own reversal on the question demonstrates. It is the Ondo model: register attendants, connect them to facilities, pay for referral, and treat the attendant as the entry point to a system rather than as its rival 2. The measurable output of that approach was a 61.8 percent rise in facility deliveries over three years 2, and moving births into facilities is the intervention with the strongest evidence behind it in global maternal health.

The herbal preparations given in pregnancy remain the largest unexamined risk in Yoruba medicine, and the one this compiler would nominate for research priority if asked. They are administered to millions of people in a physiological state of maximal vulnerability, and almost nothing is known about them.