This study has identified significant associations between maternal BMI and the stage of pregnancy women book for antenatal care in England. Women with an overweight or obese BMI were more likely to access antenatal care later in pregnancy than women with an underweight or recommended BMI. A slightly increased association with booking beyond the first trimester was observed for women with an obese BMI, whereas the analysis of booking by trimester showed that the strongest association with maternal weight status and late booking was among women with obesity and booking very late in pregnancy, in the third trimester. Women who access antenatal care late are at increased risk of adverse outcomes, and this presents a double burden for women with obesity and their offspring who face risks due to both their weight status and due to late access. Accessing care late in pregnancy misses opportunities for routine screening such as the fetal anomaly scan at 20 weeks gestation [
4]. As there is an established association between maternal obesity and congenital anomalies [
18] this presents a missed opportunity for early detection for this population of women. Despite the significant associations between maternal BMI and late access to care identified in our study, these were not observed to the same extent as other socio-demographic factors including women from BME groups, teenagers and unemployed mothers. Other studies in England have reported similar associations with late access among BME groups, young women and unemployment [
11‐
13]. Similar to obesity, both unemployment and BME groups are overrepresented among maternal deaths in the UK, [
6,
9], identifying further inequalities in maternal and perinatal risk for these populations. In our study, 36.5% of women accessed antenatal care late which is comparable to local datasets for East London and Sheffield (37.5 to 49.9%) when taking into consideration national variation in BMI and other socio-demographic factors [
11,
12]. However, there is limited comparative robust research specifically investigating maternal BMI and late access. A study in the USA found that women with an obese BMI accessed care 0.2 weeks later than women with a recommended BMI, although this was not statistically significant [
37]. Similar to our findings, the authors identified that late or no access to antenatal care was more likely among women from BME groups, teenagers and multiparous mothers [
37]. A further London-based study also identified that teenagers and multiparous women accessed care beyond 18 weeks, but the authors reported a high level of missing BMI data [
38].
A recent review of the pathophysiology of obesity and menstrual disorders identified that obesity, especially central adiposity, was associated with increased oestrogen levels, circulating free testosterone, and with insulin levels which stimulates the production of androgens in ovarian tissue which can cause disruptions to normal ovulation and menstrual bleeding [
32]. Additionally, the association with menstruation disturbances was stronger for early onset obesity potentially due to the leptin levels which regulates the gonadotropin surge initiating pubertal stages [
32]. The association with menstrual disturbances may further contribute to late access to antenatal care due to delayed realisation about conception, particularly among women with central adiposity, or those who developed obesity during childhood. Qualitative research has also identified complex reasons for late access among socially excluded “hard to reach” groups of women. Haddrill et al. [
39] reported three themes to describe reasons for delayed access to antenatal care, including
“not knowing” where women reported a lack of realisation about the pregnancy or beliefs that they were not pregnant (e.g. due to contraception use or maternal age);
“knowing” including women who avoided or postponed access to antenatal care (e.g. due to fears or lack of perceived value of antenatal care); and being
“delayed” including health professional and healthcare system failures (e.g. mis-estimation of gestational age, delays with referrals and appointments). A study for the UK Department of Health also grouped women into two distinct typologies: those who embraced their pregnancy such as women from Asian, Muslim, Somali and Romany communities; and those who were anxious about their pregnancy such as women who were homeless, drug and alcohol dependent, with learning difficulties and teenagers [
40]. The authors reported different reasons for delaying access to antenatal care between groups. Among those who embraced their pregnancy the family was a significant factor in their lives, and there were culturally defined roles of motherhood within society. Women were expected to continue their daily routine rather than seeking medical help as pregnancy was considered to just be part of life, or an act of God which was out of their hands. There was an emphasis on seeking medical advice from the family and distrust among certain groups of medical professionals [
40]. Our study identified the strongest associations with late access among women from BME groups, and previous studies in the UK report that South Asian women had fewer antenatal appointments and waited longer before seeking antenatal care when compared with White women [
10,
41], and women from BME groups report being insufficiently involved in decisions about their maternity care to have confidence and trust in the staff [
42]. However, among the population categorised as being anxious, reasons for late access related to difficulties accepting the pregnancy, or that accessing care was less important than other priorities such as finding housing [
40]. Similar to the group of women who embraced pregnancy, there was some distrust of medical professionals, plus fears of being labelled or referred to social services and having their baby taken into care. Extremely anxious women waited longer to seek antenatal care, particularly teenager mothers who reported feeling concerned about the associated stigma, fearful of health professionals informing their parents of the pregnancy, and relinquishing control [
39,
40,
43]. Previous studies have reported that teenage mothers are less likely to keep appointments or attend antenatal classes, and access maternity care later [
44,
45], which is also reflected in our analyses of teenage mothers and late access to care. Additional barriers to accessing care are inability to travel to appointments, language barriers for women who do not speak English, women with no fixed address who are not registered with a GP, and women with learning difficulties report being embarrassed to seek help as they did not understand written information leaflets [
40,
43]. Although we were able to explore teenage pregnancy and maternal ethnic group in our analyses, the dataset did not include variables reflecting additional factors such as ability to speak English, learning difficulties or homelessness and therefore we could not explore these factors.
Delivery of maternity services and targeted public health support should encourage early access to antenatal care for optimal pregnancy outcomes for both mothers and babies. The findings of this research, and the existing qualitative research, highlight the complexity of socio-demographic inequalities associated with late access to antenatal care which are often interrelated. For example, women who are unemployed are more likely to have a higher BMI, as are women from more deprived areas and among BME groups [
23,
25]. It is, therefore, important to try to tackle these inequalities in access to antenatal care in a way which holistically targets disadvantaged populations. Future epidemiological research investigating predictors of late access to antenatal care, and maternity or public health interventions to improve access to care, should consider the relationships between these complex factors particularly maternal BMI, employment, teenage pregnancy and ethnic minority groups.
Strengths and limitations
This is the first national-level study to explore maternal BMI and other socio-demographic factors which may be associated with late access to antenatal care in the UK, considering the influence of confounders. The research has a large sample size of over 600,000 births, which is comparable to the national average pregnancy population in terms of maternal characteristics such as age [
24]. The dataset used in this study is the only existing national, general maternity dataset (i.e. not restricted to mortality or other sub-population data) which incorporates maternal BMI data over a prolonged time period. A limitation of this study is the lack of data available post-2007. When comparing maternal BMI distribution in this dataset with recent UK cross sectional booking data, maternal overweight and obesity prevalence appear to have increased (obesity reported as 21% of all women booking in July in 2015, [
46]). Therefore, the prevalence of late access relating to overweight and obesity is likely to be higher today than that reported in this study. A further limitation of this study relates to the nature of secondary analysis of a dataset which was collected for a different purpose. While existing datasets provide relatively instant access to a wealth of data for research which would take decades to prospectively collect, they are often limited as they lack the entire group of variables that would be required to fully answer the research question. For example, in this study it was not possible to explore the impact of some potentially important confounders that have been identified in the existing literature, such as women’s ability to speak English or the influence of a learning disability. Additionally, reliance on routine maternity data means it is not possible to identify whether the BMI data represents self-reported pre-pregnancy weight or measured booking pregnancy weight. UK guidelines state that the booking BMI should be measured [
47]. However, a national report highlighted that 16% of maternity units used self-reported height and weight measurements rather than measured [
25]. Self-reported weight is often underestimated and this may have caused an underrepresentation of overweight and obesity in the sample. Alternatively, if measured weights were used to define booking BMI then this would over-estimate overweight and obesity prevalence among women accessing care late in pregnancy as the measurements would also incorporate the naturally incurred weight gain of pregnancy (including fat gain, as well as the healthy weight of the fetus, placenta and fluids). Attempts to limit the effect of false positives of overweight and obesity were made by adjusting the BMI data for women who booked after the first trimester of pregnancy.