Sutapa Agrawal, Gagandeep K. Walia, Eleonora Staines-Urias, Juan P. Casas,3, Christopher Millett

1. Public Health Foundation of India, Gurgaon, New Delhi NCR,India

2. Department of Epidemiology and Population Health, London School of Hygiene and Tropical Medicine, London, UK

3. Institute of Cardiovascular Science, University College London,London, UK

4. School of Public Health, Imperial College, London, UK

Prevalence of and risk factors for eclampsia in pregnant women in India

Sutapa Agrawal1, Gagandeep K. Walia1, Eleonora Staines-Urias2, Juan P. Casas2,3, Christopher Millett4

1. Public Health Foundation of India, Gurgaon, New Delhi NCR,India

2. Department of Epidemiology and Population Health, London School of Hygiene and Tropical Medicine, London, UK

3. Institute of Cardiovascular Science, University College London,London, UK

4. School of Public Health, Imperial College, London, UK

Objective:Eclampsia is a potentially fatal disorder in pregnant women and remains an important cause of maternal and perinatal child morbidity and death worldwide. We aimed to assess the prevalence of and risk factors for convulsions (not occurring from fever) during pregnancy in Indian women. Convulsion is a key symptom suggestive of eclampsia.Methods:Cross-sectional data from India’s third National Family Health Survey, conducted during 2005—2006 were used. Self-reported information on convulsions during pregnancy was obtained from 39,657 women aged 15—49 years who had a live birth in the 5 years preceding the survey. Multiple logistic regression analysis was used to explore risk factors for convulsion in pregnancy.Results:Overall, 1 in 10 women reported having convulsions in their most recent pregnancy.The prevalence was significantly higher in women living in rural areas compared with those living in urban areas (11.3% vs. 7.4%;P＜0.0001), with marked state and geographic variation. The odds of convulsions were significantly higher in women with a twin pregnancy [odds ratio (OR)2.12; 95% confidence interval (CI) 1.45—3.11], a previously terminated pregnancy (OR 1.32; 95%CI 1.20—1.45), diabetes (OR 1.37; 95% CI 0.99—1.89), or asthma (OR 1.85; 95% CI 1.35—2.54), in women who were alerted to pregnancy complications (OR 2.78; 95% CI 2.50—3.08), in Sikh women (OR 1.73; 95% CI 1.28—2.33), in women in a low social group (OR 1.40; 95% CI 1.25—1.58),and in women residing in central India (OR 1.51; 95% CI 1.31—1.74) or eastern India (OR 1.33;95% CI 1.14—1.54) with reference to their counterparts.Conclusion:Our findings from a large population-based nationally representative sample of Indian women indicate a high prevalence of convulsions, a symptom suggestive of eclampsia, and its association with several maternal, lifestyle risk factors and sociodemographic characteristics.

Convulsions; eclampsia; risk factors; women; Indian

Introduction

‘Eclampsia’ refers to the occurrence of newonset, generalized, tonic—clonic seizures or coma in a woman with preeclampsia. It is the convulsive manifestation of preeclampsia and one of several clinical manifestations at the severe end of the preeclampsia spectrum[1, 2]. It is a potentially fatal disorder in pregnant women and remains an important cause of maternal and perinatal child morbidity and death worldwide, accounting for more than 50,000 maternal deaths annually[3—7]. World Health Organization estimated, at least 16%of maternal deaths in low- and middle-income countries(LMICs) result from hypertensive disorders of pregnancy, of which eclampsia is the primary contributor [8]. Preeclamptic women in LMICs are also three times likelier to progress to eclampsia than women in high-income countries [9]. Largely on the basis of clinical data, the incidence of eclampsia ranges between 2% and 10%, depending on the population studied and the definition of eclampsia used [10]; clinical studies suggest that the proportion of deliveries impacted by eclampsia in Indian women ranges from as low as 0.9% to as high as 7.7% [11—14]. However, these clinical studies likely suffer from selection bias on the basis of severity of the condition,especially among populations with limited access to prenatal care, and therefore may underestimate the prevalence of the condition. Precise country-specific population-level estimates of eclampsia prevalence are largely unavailable.

Risk factors for eclampsia reported in high-income countries include young and old maternal age, obesity before pregnancy, being unmarried, excessive weight gain during pregnancy, multiple gestations, nulliparity, chronic hypertension, low socioeconomic status, prolonged birth interval,lack of prenatal care, and current smoking [15—21]. Very few population-based epidemiological studies of eclampsia have been conducted in LMICs. Prior studies were largely undertaken in clinical settings [13, 22—24], and do not provide representative data on risk factors as they are based on highly selected samples. Available information suggests that predisposing factors in LMICs may include poverty,illiteracy, low educational attainment, lack of health awareness, and poor access to antenatal care (ANC) during pregnancy. Previous studies have also found that beliefs about seeking medical advice during pregnancy may also be a factor in some LMICs, and can result in delayed diagnosis and inappropriate treatment of patients with preeclampsia or eclampsia [25, 26].

India is in the midst of a demographic, epidemiological,and nutrition transition characterized by a growing population,increasing urbanization, a shift in the patterns of diseases,and changes in lifestyle [27]. The past decade has seen a dramatic increase in lifestyle-related noncommunicable diseases,including obesity, diabetes mellitus, hypertension, coronary heart disease, stroke, and cancers [28]. Given that preeclampsia shares many risk factors with cardiovascular disease (e.g.,obesity, type 2 diabetes, and hypertension), it is expected that the increase in cardiovascular disease risk factors to occur in women of childbearing age will translate into a higher incidence of preeclampsia and its complications (e.g., eclampsia).The objective of this study was thus to quantify the prevalence and predictors of convulsions (not occurring from fever), a symptom suggestive of eclampsia, with use of nationally representative data among Indian women.

Methods

Study setting and participants

Cross-sectional data from the third National Family Health Survey (NFHS-3), conducted in India in 2005—2006, was used for this study. NFHS-3 was designed on the lines of the Demographic and Health Surveys (available at www.measuredhs.com), which have been conducted in many LMICs since the 1980s. The National Family Health Survey has been conducted in India for three successive rounds, each at an interval of 5 years. NFHS-3 collected demographic, socioeconomic, and health information from a nationally representative probability sample of 124,385 women aged 15—49 years residing in 109,041 households. The sample is a multistage cluster sample with an overall response rate of 98%. All states of India are represented in the sample (except the small union territories), covering more than 99% of the country’s population. Full details of the survey have been published [29] and are available at www.nfhsindia.org. To assess symptoms suggestive of eclampsia, we restricted the sample to those women who had had a live birth in the 5 years preceding the survey.We further restricted our analyses to data pertaining to the most recent birth, both to minimize recall bias and to draw on antenatal care (ANC) measures, which were available only for the most recent pregnancy. This resulted in a final sample size of 39,657 participants.

Outcome measures

To assess the occurrence of eclampsia, we constructed a measure based on women’s self-reports of symptoms during pregnancy. Specifically, mothers were asked: “During this pregnancy, did you have convulsions not from fever?”.

The response options were “Yes,” “No,” and “Don’t know.”Following World Health Organization [30] and National Institute for Health and Care Excellence [31] guidelines, we created a dichotomous indicator of eclampsia: women who reported experiencing convulsions (not from fever) were coded as eclamptic. However, it was not possible to confirm clinical diagnosis of these symptoms. Data on physicianreported diagnosis of convulsions (not from fever) or clinical test results were not available in the NFHS-3 to verify a selfreported diagnosis.

Predictor variables

The following maternal reproductive risk factors were evaluated in the adjusted analysis: total number of children ever born (one, two or three, four or more); preceding birth interval ( first birth order, interval ＜2 years, interval 2—3 years,interval ＞3 years); ANC visit during pregnancy (no visits,one visit, two visits, three visits, four of more visits); blood pressure measured during pregnancy (no, yes); received advice on pregnancy nutrition during ANC visit (no, yes);alerted to pregnancy complications such as convulsions(no, yes); type of pregnancy (singleton, twin); ever had a terminated pregnancy (no, yes); and anemia level (not anemic, mild anemia, moderate anemia, severe anemia). The biological and lifestyle factors included body mass index(BMI) categories (Indian adult population standard) [32] —18.4 kg/m2or less (underweight), 18.5—22.9 kg/m2(normal),23.0—24.9 kg/m2(overweight), 25.0 kg/m2or more (obese);current tobacco smoking (no, yes); alcohol consumption(no, yes); self-reported diabetes (no, yes); and self-reported asthma (no, yes). Sociodemographic predictors included age(15—29 years, 30—39 years, 40—49 years); education (no education, primary education, secondary education, higher education); religion (Hindu, Muslim, Christian, Sikh, others);caste (Scheduled Castes, Scheduled Tribes, Other Backward Class, others); employment status (not working, working);wealth index (measured by an index based on household ownership of assets and graded as lowest, second, middle, fourth, and highest) that was computed by previously described methods [29]; place of residence (urban, rural);and geographic regions (north, northeast, central, east, west,south). For the definition of some variables, see Table 1.

Statistical analyses

We first examined observed and expected regional and rural/urban differences in the prevalence of seizures and then estimated associations with eight socioeconomic and demographic variables, nine maternal factors, and five BMI- and lifestyle-related and disease-related factors. We did this analysis since we anticipated a striking difference in the prevalence of eclampsia by state since all the states in India are in different levels of demographic, epidemiological, and fertility transition and the differences in health-care and healthseeking behavior among women is also discernible. To do this analysis, firstly we ran a multiple regression model (not including the following geographic variables: state and region and place of residence) but including the identified potential risk factors. By using the “predict” command in the logistic regression, we ‘estimated’ the risk (log odds) of eclampsia for every participant in this study. Then, using this estimated risk,we obtained the average by each state and then compared the estimates with the observed prevalence (see Table 2). We then plotted the results as bar plots sorted from the lowest to the highest prevalence of eclampsia, and for each state we have the observed and the estimated rates based on our model. This is a simple way to try to quantify what accounts for the massive difference in the states.

Potential risk factors were selected on the basis of previous knowledge of their association with seizures presented in earlier literature. Lastly we used multiple logistic regression models to estimate the prevalence odds ratios (ORs) for each of these risk factors, adjusted for the above-mentioned confounders and risk factors. As certain states and certain categories of respondents were oversampled, sample weights were used to restore the representativeness of the sample [29].

The results are presented as ORs with 95% confidence intervals (CIs). The estimation of CIs takes into account design effects due to clustering at the level of the primary sampling unit. We assessed the possibility of multicollinearity between the covariates. In the correlation matrix of covariates, all pairwise Pearson correlation coefficients were less than 0.5, suggesting that multicollinearity did not affect the findings. Analyses were conducted with the IBM SPSS Statistics version 19 (IBM, Armonk, NY,United States).

Table 1. Characteristics of the study participants and reported prevalence of eclampsia during pregnancy for the most recent live birth among women aged 15—49 years (n=39,657) who had a live birth in the 5 years preceding the survey according to selected characteristics, India,2005—2006

Table 1 (continued)

Table 1 (continued)

Table 1 (continued)

Ethical considerations

The NFHS-3 survey received ethics approval from the International Institute for Population Science’s Ethical Review Board and the Indian Government. Prior written informed consent was obtained from each respondent. The analysis presented in this study is based on secondary analysis of existing survey data with all identifying information removed.

Results

Observed and expected prevalence of convulsions in pregnancy by area of residence and state. Overall, 1 in 10 participants (10.3%; n=4071) in our sample reported convulsions(not from fever) during pregnancy, symptoms suggestive of eclampsia (Table 2). The observed prevalence was significantly higher in women living in rural areas than in those living in urban areas (11.3% vs. 7.4%; P＜0.0001; Figs. 1—3).There was marked state-level variation in the observed prevalence of convulsions, with rates ranging from 1.1% in Tamil Nadu to 24.5% in Jharkhand (Fig. 4). The observed prevalence of convulsions was greater than 20% in Uttarakhand, Bihar,Jharkhand, Arunachal Pradesh, and Sikkim and between 10%and 20% in Punjab, Madhya Pradesh, Uttar Pradesh, Orissa,Nagaland, and Tripura.

Characteristics of the study participants

Table 1 presents the characteristics of the study participants. One quarter of the mothers had given birth once,31% of the births were preceded by an interval of more than 3 years, and 1% of the pregnancies resulted in multiple births. Approximately one fifth (18.5%) of the mothers reported having terminated a pregnancy. Malnutrition rates were high: almost one fifth of the mothers were moderately or severely anemic, 31% were underweight, and 22% were either overweight or obese. More than half the mothers had had four or more ANC visits during their last pregnancy,for 68% blood pressure was measured during an ANC visit during their last pregnancy, 66% reported receiving advice on pregnancy nutrition during an ANC visit, and only 15%were alerted to the possibility of pregnancy-related complications such as convulsions. Very few were current smokers(1.5%) or alcohol drinkers (2.3%). The prevalence of diabetes (1.3%) and asthma (1.2%) was low. Most mothers (almost three quarters) were aged between 15 and 29 years (the mean age being 34.89 years), and almost half (47.4%) had no education. Most of the mothers (four in five) were identified as Hindu, and two fifths belonged to Other Backward Class.Seventy percent of the mothers were not working and one quarter belonged to the poorest households. More than 70%of the mothers were residing in rural areas, whereas 28%were residents of central India.

Prevalence of convulsions in pregnancy

The prevalence of reported convulsions was highly significant (P＜0.0001) in the following groups: parity greater than four (13.3%); a preceding birth interval of less than 2 years(11.3%); no ANC (13.5%); women alerted to pregnancy complications such as convulsions (16.2%); women with a twin pregnancy (15.9%); women with a previously terminated pregnancy (12.6%); women with moderate anemia (11.4%)to severe anemia (10.0%); underweight (11.0%) and normalweight women (11.1%); current smokers (15.0%); current alcohol drinkers (12.7%); women with diabetes (16.6%) or asthma(19.4%); women with no education (12.1%); Sikh women(13.7%); women belonging to Scheduled Tribes and Other Backward Class; women belonging to households in the lowest wealth quintile (13.6%); women who do not work (10.7%);women residing in rural areas (11.3%); and among women living in the eastern part of India (15.1%).

Table 2. Observed versus expected prevalence of eclampsia during pregnancy for the most recent live birth among women aged 15—49 years(n=39,657) who had a live birth in the 5 years preceding the survey, by state and residence, India, 2005—2006

Fig. 1. Prevalence of eclampsia in India, 2005—2006.

Associations between convulsion and predictor variables

The adjusted odds for convulsion were significantly higher in women with a twin pregnancy compared with a singleton (OR 2.12; 95% CI 1.45—3.11), in women with a previous terminated pregnancy (OR 1.32; 95% CI 1.20—1.45), among women who were alerted to pregnancy complication such as convulsions during an ANC visit (OR 2.78; 95% CI 2.50—3.08), in women with diabetes (OR 1.37; 95% CI 0.99—1.89) or asthma (OR 1.85; 95% CI 1.35—2.54), among Sikh women (OR 1.73; 95%CI 1.28—2.33) compared with Hindu women, among women belonging to Other Backward Class (OR 1.40; 95% CI 1.25—1.58) compared with women belonging to Scheduled Castes,women residing in rural areas (OR 1.22; 95% CI 1.09—1.37)or in central (OR 1.51; 95% CI 1.31—1.74) or eastern (OR 1.33;95% CI 1.14—1.54) regions of India compared with women residing in northern regions (Table 3).

Women with at least one ANC visit during pregnancy (OR 0.68—0.77; 95% CI: 0.59—0.79), overweight (OR 0.86; 95% CI:0.73—1.02) or obese women (OR: 0.70; 95%CI: 0.58—0.85),women older than 20 years of age, women who are employed(OR 0.85; 95% CI 0.79—0.92), women with higher educational attainment (OR 0.68; 95% CI 0.54—0.87), women with higher household wealth status, and women living in north eastern (OR 0.74; 95% CI 0.56—0.98), western (OR 0.38; 95% CI 0.31—0.47),and southern (OR 0.44; 95% CI 0.36—0.52) regions of India had a lower likelihood of having symptoms suggestive of eclampsia.

Discussion

In this nationwide large-scale cross-sectional study we identified three main sets of findings relating to (1) overall selfreported prevalence of convulsions (not from fever), which are the main symptoms of eclampsia, (2) massive geographic differences in prevalence, and (3) risk factors for prevalence.We found that the prevalence of reported convulsions (10.3%)was high compared with that in earlier studies of clinical eclampsia in Asian populations, notably in Singapore [34].

Fig. 2. Prevalence of eclampsia in rural India, 2005—2006.

The prevalence in our study is higher than not only that of high-income countries but also that of most LMICs and the incidence is higher than that reported in other parts of India.This may be due to the cross-sectional symptomatic nature of the study rather than clinical confirmation. The only other study that gives an incidence similar to that of our study is from the Dhaka Medical College and Hospital, Bangladesh,which is the largest tertiary referral government hospital in Bangladesh and deals mostly with referral cases, where the incidence of eclampsia is 9% [25]. Therefore from the study we can safely say that approximately 10% of the Indian women surveyed who had been pregnant during the previous 5 years self-reported convulsions. The diagnosis of convulsions was suspected and was not confirmed by a clinician. The cause of the convulsions was not specified. Some (perhaps a small minority of the seizures) were associated with eclampsia.

In this study we used the occurrence of convulsions (not from fever) as a proxy for eclampsia. There are other causes of nonfebrile seizures in pregnancy and it is not justified to make that presumption without our talking about the general prevalence of epilepsy in the general population. In various studies the overall prevalence of epilepsy in India is estimated to be between 2.5 and 11.9 per 1000 population [35—39]. The prevalence is reported to be about 1% of the overall population[35], being higher in the rural population (1.9%) than in the urban population (0.6%) [40, 41]. However, there are almost to 1.5 million women with epilepsy in the reproductive age group (15—49 years) in India [42]. There are very few incidence studies from India, and the most recent one suggests an agestandardized incidence rate of 27.3 per 100,000 per year [43].

Fig. 3. Prevalence of eclampsia in urban India, 2005—2006.

Fig. 4. Observed and expected prevalence of convulsions (not from fever) among women in the last pregnancy, India 2005—2006.

Table 3. Unadjusted and adjusted odds ratios and 95% confidence interval for the risk of eclampsia during pregnancy for the most recent birth among women aged 15—49 years who had a live birth in the 5 years preceding the survey, India, 2005—2006

Table 3 (continued)

Table 3 (continued)

Convulsions (during pregnancy) are a common manifestation of epilepsy and eclampsia [44], and in the case of no history of epilepsy before conception, eclampsia remains the major cause of convulsions that are a result of untreated hypertension during pregnancy. Apart from eclampsia and idiopathic causes, the various other causes of convulsions during pregnancy include antiphospholipid syndrome, cerebral vein thrombosis, thrombotic thrombocytopenic purpura, cerebral infarction, drug and alcohol withdrawal, and hypoglycemia, but these are very rare compared with eclampsia [45].However, convulsions during pregnancy that are unrelated to preeclampsia need to be distinguished from eclampsia. Such disorders include seizure disorders as well as brain tumor,aneurysm of the brain, and medication- or drug-related seizures. Since convulsions during pregnancy could be a clinical manifestation of a number of medical disorders and thus a different pathophysiology, it is recommended to carefully diagnose the cause of convulsions in pregnant women and women of childbearing age and then accordingly manage them to avoid any future complications.

The striking geographic variations in the prevalence of eclampsia between specific states in India in our study needs specific mention. There was a more than 20-fold variation in eclampsia rates between the states with the lowest (Tamil Nadu 1.1%) and highest (Jharkhand 24.5%) prevalence. These substantial statewise differences in the prevalence of eclampsia clearly warrant further investigation. State-specific analysis using multilevel methods could be conducted to explore the substantial differences in prevalence in Indian states. One possible explanation for the state differential may be related to how well the question on convulsion was delivered and/or understood in the different states. Other potential explanations for these differences are that in high-prevalence states there are high rates of diabetes cases, more terminated pregnancy cases among women, and a high Scheduled Tribe population coupled with poorer access to health care services (expect for Kerala) compared with the rest of India [29]. Also this can be attributed to health care infrastructure, health service availability, and health-seeking behavior. An alternative explanation may be related to climatic differences across Indian regions.Some studies in the West have reported a higher incidence of eclampsia associated with conception during the spring and summer months [46]. Potential mechanisms include seasonal variation in exposure to infections, dietary changes, and alteration in vitamin D regulation and calcium metabolism as a consequence of exposure to sunlight, which are, in turn, associated with blood pressure levels [46, 47].

Thirdly, we identified a number of specific traditional and nontraditional risk factors for symptoms suggestive of eclampsia. Some of the risk factors for eclampsia among Indian women are similar to those found among Asian women [48] and other ethnic groups, whereas some others differ. In line with published reports, symptoms of eclampsia were associated with a maternal age of more than 35 years,nulliparity, multiple pregnancies, poor socioeconomic conditions, and poor education [13, 22, 23]. The evidence for young maternal age as an independent risk factor for preeclampsia is still controversial [24—26]. The age distribution of the women in our study is similar to that in other reported studies and suggests that eclampsia is, probably, a disease of low-parity young women [49]. However, a study done in Saudi Arabia showed that women at extremes of maternal age, nulliparous women, and high-parity women are at an increased risk of developing eclampsia [50]. The difference between our findings and those of other studies could be due to differences in the population-based and hospital-based settings.

It is well established that the risk of eclampsia is greater in twin rather than in singleton pregnancies, and we found similar result in our study. Various studies have reported the incidence of eclampsia in twin pregnancies to be between 13%and 37%, which is two to three times higher than in singleton pregnancies [51—54], and about 24.3% in the case of triplet and quadruplicate pregnancies [55]. In a clinical study, analysis of 37 cases of twin pregnancy complicated by eclampsia showed that the incidence of twins in the total 1030 cases of eclampsia was three times the incidence in the general population [56].Our findings were similar to those of other studies regarding the association of a history of miscarriage or a terminated pregnancy [57] with eclampsia risk.

Underlying medical conditions such as diabetes [52],asthma, or a previous preeclampsia [58] are associated with higher prevalence odds of eclampsia, and our study findings are consistent with earlier reports from high-income countries.A contrasting finding from our study that obese women are at lower risk of eclampsia is different from the findings of studies conducted in the West [59], but the mechanisms involved are not known. Women with the lowest BMI are relatively protected against eclampsia [60], which is also confirmed in our study (OR 0.93; 95% CI 0.85—1.02).

Our findings are similar to those of other studies that evaluated the risk factors for eclampsia in other populations regarding socioeconomic status [23, 61] such as ethnicity [62],education, employment status, and place of residence. Studies from high-income countries have been largely inconclusive on this, given the low incidence of eclampsia, but in our study we found a clear pattern: no difference from other health conditions. We found women belonging to the Sikh religion and Other Backward Class and women residing in rural areas were at more risk of symptoms suggestive of eclampsia. The reason for this increased risk is unclear but may be associated with maternal poverty and social deprivation. The high odds of eclampsia risk among Sikh women in our study might be mediated through obesity, which is highest among Sikh women(4.7% vs. 1.2% among Hindu women for obesity defined as BMI ＞30 kg/m2). Maternal prepregnancy BMI classified as overweight and obese has long been associated as a modifiable risk factors for preeclampsia/eclampsia [5, 6]. Obese pregnant women (BMI ＞30 kg/m2) experience a nearly threefold increase in the risk of developing preeclampsia/eclampsia compared with women of normal weight (BMI 18.5 to ＜25.0 kg/m2) [7], which has been confirmed in several large-scale US studies. In India a recent study assessing the risk factors for early onset of severe preeclampsia and eclampsia among north Indian women found overweight (＞120%—150% of prepregnancy ideal body weight, adjusted OR 4.65) and lower socioeconomic class (Kuppuswamy class III—V) (adjusted OR 3.00)to be related to the increased risk [23]. We also found a link with education, with a 32% reduced risk in women who had a higher level of education (9 years or more). These novel risks were independent of other risk factors, including adequacy of prenatal care.

The new information linking symptoms suggestive of eclampsia risk to the mother’s socioeconomic characteristics is an important first step toward identifying new, nontraditional risk factors. Although traditional risk factors (mostly clinical)for eclampsia are well recognized, these diseases remain unknowable, and there is no known effective way to reduce their incidence. Therefore nontraditional risk factors may be of great importance in the design of future interventions to prevent the occurrence of eclampsia or eclampsia symptoms in a low-resource setting such as India, but data regarding these risk factors are scarce in LMICs.

Our study indicates higher prevalence of symptoms suggestive of eclampsia in women who never visited an antenatal clinic during the pregnancy or who never received advice on pregnancy nutrition (OR 1.07, P=0.080) and were not alerted to pregnancy complications such as convulsions during their ANC visit. This may also be due to ‘reverse causation.’ For example, ANC personnel treat pregnant women at high risk of preeclampsia/eclampsia or who report symptoms such as blurred vision and headache, thus resulting in a positive association. The identification of eclampsia and counseling of women with eclampsia relies fundamentally on the frequency of ANC [63] and if blood pressure was measured during the visit. Globally, the absence of ANC is strongly associated with eclampsia and death [18]. Despite this, the absence of preconception care coupled with a lack of effective and universal ANC remains a serious challenge in LMICs. Many women with preeclampsia, particularly, at the community level, are missed because of the lack of ANC. These women are likelier to develop serious complications such as eclampsia during the latter part of the pregnancy. ANC utilization is approximately 68% in LMICs compared with 98% in high-resource settings[64]. The regions of the world with the lowest levels of use are South Asia, where only 54% of pregnant women have at least one ANC visit [64] and India (22.8%) [29]. Not surprisingly,there is marked urban—rural differential in accessing ANC in LMICs, including India. Whereas 86% of women in urban settings will have one antenatal visit, only 65% of women in rural settings will have the same [64]. For repeated antenatal visits,62.4% of women in urban India report four or more antenatal visits compared with 27.7% of rural women [29].

Strength and limitations of the study

The strengths of our study include the large nationally representative study sample allowing comparisons to be made between states and urban versus rural settings, and the ability to examine socioeconomic and lifestyle patterning of symptoms suggestive of eclampsia risk in a population-based survey. Further, the large sample size provided adequate power to identify the potential risk factors and compensated for the ethnic variations in Indian populations. We could evaluate the association of well-known risk factors as potential confounders and effect modifiers, including birth intervals, maternal age, type of pregnancy, diabetes, asthma, BMI, and tobacco smoking. Furthermore, the survey was conducted with an interviewer-administered questionnaire in the native language of the respondent with a local, commonly understood term for all the health problems during pregnancy. Eighteen languages were used, with back translation to English to ensure accuracy and comparability.

However, because of the general challenges of measuring hypertensive disorders in population-based studies, the measurement of symptoms suggestive of eclampsia in the National Family Health Survey also has apparent limitations.These include the cross-sectional study design, thereby preventing conclusions regarding causality. Second, an important drawback of the study is that we report on seizures/convulsions during pregnancy and used this outcome as a proxy for eclampsia, which is likely to be a poor proxy, but these are the best available data we have at the national level. The 10%prevalence of ‘eclampsia’ is too high for a population-based survey, and might be due to overreporting, probably because some women do not understand the question. Also, NFHS-3 data did not have as a variable hypertension, which is considered the hallmark for the diagnosis of eclampsia, and convulsions sometimes may be due to cerebrovascular accidents or some brain lesion or medication- or drug-related seizures;these may be some of the reasons for the higher prevalence reported in this study. Third, the symptoms suggestive of eclampsia were all self-reported by women, and are therefore subject to bias. Case ascertainment was based on self-reported convulsions/seizures during the last pregnancy rather than a clinical assessment. Although we cannot exclude misclassification within this context, it is unlikely that we missed severe symptoms suggestive of eclampsia cases. Fourth, we could not identify the gestational onset of eclampsia. Fifth, no information is available on the prepregnancy eclampsia risk factors of the women as health problems during pregnancy were assessed only for the most recent birth within 5 years preceding the survey. So there was a time gap between the information on all the covariates/risk factors and pregnancyrelated health problems, which may be one of the reasons that we could not find any substantial association between some of the important risk factors (e.g., obesity, age of women) and eclampsia in this population-based survey which was otherwise proven in clinical studies. Sixth, the study database did not include information on some potentially important factors;for example, we could not obtain maternal prepregnancy BMI[65], familial aggregation [66], and genetic factors [67], which are important risk factor for eclampsia as proved in studies from high-income countries.

This study has potential implications for the possible prevention of pregnancy-related complications in India. Pregnant women in LMICs, including India, are among the most vulnerable populations in the world. Hypertensive disorders in pregnancy such as preeclampsia, eclampsia, and gestational hypertension are one of the three (apart from hemorrhage and sepsis) leading causes of maternal death and morbidity[68—72]. Community health care workers, specifically female health care workers, are an integral part of the health care force in many LMICs and can be employed to provide timely care to women with preeclampsia and eclampsia. Prevention strategies should be applied to every pregnant woman since we cannot predict who will develop preeclampsia and eclampsia given the limitation in resources. Measuring blood pressure and urinary protein levels is challenging in LMICs because of the financial cost and lack of training. A detection tool that is affordable and can be easily applied is needed.

Eclampsia has remained a significant public health threat in both high-income countries and LMICs, contributing to maternal and perinatal morbidity and mortality globally[69—72]. The Millennium Development Goals have placed maternal health at the core of the struggle against poverty and inequality as a matter of human rights. Ten percent of women have high blood pressure during pregnancy, and eclampsia and preeclampsia complicate 2%—8% of pregnancies [72].Increasing awareness of maternal death as a public health priority, for both maternal and child health, has been important,and has helped implementation of improved health services.Millennium Development Goal 5 calls for a reduction by three quarters, between 1990 and 2015, in the maternal mortality ratio (www.un.org/millenniumgoals).

Our study findings may serve as an important call for health care providers to increase their awareness of the increased population-level risk of eclampsia. An increase in the risk of conditions as potentially dangerous as eclampsia underlines the importance of regular health care during the preconception, interconception, and antenatal periods. However, more epidemiological research in India should focus on uncovering preventable causes of eclampsia, and public health practice and policy must promote improved access to health care and mandatory ANC visits and reduction of traditional and nontraditional risk factors.

To conclude, the high prevalence of symptoms suggestive of eclampsia observed in the present analyses highlights that hypertensive disorders of pregnancy remain unaddressed in India, with a huge epidemiological burden especially when we are aware that ANC is poorly utilized in Indian settings.Our findings from a large nationally representative sample of Indian women indicate that modifiable lifestyle risk factors related to different geographic ethnic groups, regions, and locations exist that can be addressed while specific preventive and management strategies are developed. We found that higher education level, high household wealth status, and the mother’s employment status acted as protective factors for eclampsia among Indian mothers. Further research to examine the clinically diagnosed prevalence of eclampsia is needed in LMICs.

With the target of the Millennium Development Goals in sight, hypertensive disorders of pregnancy (including eclampsia and preeclampsia) should be identified as a priority area in reducing maternal and infant morbidity and mortality in India. This calls for coordinated efforts and close involvement of the community, governmental/nongovernmental organizations, clinicians, nurses, and paramedical staff. Since for most Indian mothers ANC is still the main source of blood pressure measurement, nutrition education,and counseling on the signs of pregnancy complications such as convulsions, ensuring universal provision of mandatory comprehensive ANC, as emphasized in India’s 11th 5 year plan (2007—2012) [73], is vital for improving maternal health in India.

Sutapa Agrawal and Gagandeep K. Walia are supported by Wellcome Trust Strategic Award grant no. WT084674 to Shah Ebrahim and a Wellcome Trust Capacity Strengthening Strategic Award-Extension phase to the Public Health Foundation of India and a consortium of UK universities(WT084754/Z/08/A). The data for this research were collected by the Demographic and Health Surveys Program (www.dhsprogram.com), under a contract from the US Agency for International Development. The support of Macro International(Calverton, MD, United States) and the International Institute for Population Sciences (Mumbai, India) in providing access to the 2005—2006 Indian National Family Health Survey 3 data is greatly acknowledged. Critical comments by Shah Ebrahim and Pat Doyle on an earlier draft are greatly acknowledged.

Data availability

The authors confirm that all data underlying the findings are fully available without restriction. The data are publicly available from the Demographic and Health Surveys website: http://dhsprogram.com/what-we-do/survey/survey-display-264.cfm.

Conflict of interest

The authors declare no conflict of interest.

1. Lockwood CJ, Pedley TA, editors. Eclampsia [Internet]. Wolters Kluwer: release: 24.3 — C24.142 [last updated 2016 May;accessed 2015 Dec 28]. Available from: www.uptodate.com/contents/eclampsia#subscribeMessage.

2. Sibai BM. Diagnosis, prevention, and management of eclampsia.Obstet Gynecol 2005;105(2):402—10.

3. Kuklina EV, Ayala C, Callaghan WM. Hypertensive disorders and severe obstetric morbidity in the United States. Obstet Gynecol 2009;113(6):1299—306.

4. Kullima AA, Kawuwa MB, Audu BM, Usman H, Geidam AD.A 5-year review of maternal mortality associated with eclampsia in a tertiary institution in northern Nigeria. Ann Afr Med 2009;8(2):81—4.

5. Wen SW, Huang L, Liston R, Heaman M, Baskett T, Rusen ID.Severe maternal morbidity in Canada, 1991—2001. Can Med Assoc J 2005;173(7):759—64.

6. Chen CY, Kwek K, Tan KH, Yeo GS. Our experience with eclampsia in Singapore. Singapore Med J 2003;44(2):88—93.

7. Chames MC, Livingston JC, Ivester TS, Barton JR, Sibai BM.Late postpartum eclampsia: a preventable disease? Am J Obstet Gynecol 2002;186(6):1174—7.

8. Khan KS, Wojdyla D, Say L, Gülmezoglu AM, Van Look PF.WHO analysis of causes of maternal death: a systematic review.Lancet 2006;367(9516):1066—74.

9. Dolea C, AbouZahr C. Global burden of hypertensive disorders of pregnancy in the year 2000. In: Global burden of diseases, evidence and information for policy (EIP). Geneva: World Health Organization; 2003.

10. Geographic variation in the incidence of hypertension in pregnancy. World Health Organization International Collaborative Study of Hypertensive Disorders of Pregnancy. Am J Obst Gynecol 1988;158(1):80—3.

11. Swain S, Ojha KN, Prakash A, Bhatia BD. Maternal and perinatal mortality due to eclampsia. Indian Pediatr 1993;30(6):771—3.

12. Arora R, Ganguli RP, Swain S, Oumachigui A, Rajaram P. Determinants of maternal mortality in eclampsia in India. Aust NZ J Obstet Gyn 1994;34(5):537—9.

13. Singh S, Behera A. Eclampsia in eastern India: incidence, demographic profile and response to three different anticonvulsant regimes of magnesium sulphate. Int J Gynecol Obstet 2010;15(2):1—7.

14. Abalos E, Cuesta C, Grosso AL, Chou D, Say L. Global and regional estimates of preeclampsia and eclampsia: a systematic review. Eur J Obstet Gynecol Reprod Biol 2013;170(1):1—7.

15. Chesley LC. History and epidemiology of preeclampsia-eclampsia. Clin Obstet Gynecol 1984;27(4):801—20.

16. Saftlas AF, Olson DR, Franks AL, Atrash HK, Pokras R. Epidemiology of preeclampsia and eclampsia in the United States,1979—1986. Am J Obstet Gynecol 1990;163(2):460—5.

17. Douglas KA, Redman CW. Eclampsia in the United Kingdom. Br Med J 1994;309:1395—400.

18. Ansari MZ, Mueller BA, Krohn MA. Epidemiology of eclampsia. Eur J Epidemiol 1995;11(4):447—51.

19. Baeten JM, Bukusi EA, Lambe M. Pregnancy complications and outcomes among overweight and obese nulliparous women. Am J Public Health 2001;91(3):436—40.

20. Zwart JJ, Richters A, Ory F, de Vries JI, Bloemenkamp KW, van Roosmalen J. Eclampsia in the Netherlands. Obstet Gynecol 2008;112(4):820—7.

21. Coghill AE, Hansen S, Littman AJ. Risk factors for eclampsia: a population-based study in Washington State, 1987—2007. Am J Obstet Gynecol 2011;205(6):553.e1—7.

22. Pal A, Bhattacharyya R, Adhikari S, Roy A, Chakrabarty D,Ghosh P, et al. Eclampsia-scenario in a hospital — a ten years study. Bangladesh Med Res Counc Bull 2011;37(2):66—70.

23. Nanjundan P, Bagga R, Kalra JK, Thakur JS, Raveendran A. Risk factors for early onset severe pre-eclampsia and eclampsia among north Indian women. J Obstet Gynaecol 2011;31(5):384—9.

24. Yaliwal RG, Jaju PB, Vanishree M. Eclampsia and perinatal outcome: a retrospective study in a teaching hospital. J Clin Diagn Res 2011;5(5):1056—9.

25. Begum MR, Begum A, Quadir E, Akhter S, Shamsuddin L. Eclampsia: still a problem in Bangladesh. MedGenMed 2004;6(4):52.

26. Ghulmiyyah L, Sibai B. Maternal mortality from preeclampsia/eclampsia. Semin Perinatol 2012;36(1):56—9.

27. Shetty PS. Nutrition transition in India. Public Health Nutr 2002;5(1A):175—82.

28. World Health Organization. Diet, nutrition and the prevention of chronic diseases: a report of a Joint WHO/FAO Expert Consultation. Geneva; 2003 [accessed 2015 Dec 28]. Available from:http://whqlibdoc.who.int/trs/WHO_TRS_916.pdf.

29. International Institute for Population Sciences and Macro International. National Family Health Survey (NFHS-3), 2005—06:India. Volume 1. Mumbai: International Institute for Population Sciences; 2007.

30. World Health Organization. Integrated management of pregnancy and childbirth: managing complications in pregnancy and childbirth: a guide for midwives and doctors. Geneva: Department of Reproductive Health and Research; 2007.

31. National Institute for Health and Care Excellence. Hypertension in pregnancy: diagnosis and management. NICE guidelines[CG107]. 2010 [accessed 2015 Jun 24]. Available from: www.nice.org.uk/guidance/CG107/chapter/1-Guidance.

32. Indian Consensus Group. Indian consensus for prevention of hypertension and coronary heart disease. A joint scientific statement of Indian Society of Hypertension and International College of Nutrition. J Nutr Environ Med 1996;6:309—18.

33. Centers for Disease Control and Prevention (CDC). Recommendations to prevent and control iron deficiency in the United States. Morbidity and Mortality Weekly Report 1998;47(RR-3):1—29.

34. Tan KH, Kwek K, Yeo GS. Epidemiology of pre-eclampsia and eclampsia at the KK Women’s and Children’s Hospital, Singapore. Singapore Med J 2006;47(1):48—53.

35. Sridharan R, Murthy BN. Prevalence and pattern of epilepsy in India. Epilepsia 1999;40(5):631—6.

36. Sridharan R. Epidemiology of epilepsy. Curr Sci 2002;82(6):664—70.

37. Goel D, Agarwal A, Dhanai JS, Semval VD, Mehrotra V,Saxena V, et al. Comprehensive rural epilepsy surveillance programme in Uttarakhand state of India. Neurol India 2009;57(3):355—6.

38. Tripathi M, Jain DC, Gourie Devi M, Jain S, Saxena V. Need for a national epilepsy control program. Ann Indian Acad Neurol 2012;15(2):89—93.

39. Bharucha NE. Epidemiology and treatment gap of epilepsy in India. Ann Indian Acad Neurol 2012;15(4):352—3.

40. Leonardi M, Ustun TB. The global burden of epilepsy. Epilepsia 2002;43(Suppl 6):21—5.

41. Pahl K, de Boer HM. Atlas: epilepsy care in the world. Geneva:World Health Organization; 2005. Epilepsy and rights; pp. 72—3.

42. Thomas SV. Managing epilepsy in pregnancy. Neurol India 2011;59(1):59—65.

43. Banerjee TK, Ray BK, Das SK, Hazra A, Ghosal MK, Chaudhuri A, et al. A longitudinal study of epilepsy in Kolkota, India. Epilepsia 2010;51(12):2384—91.

44. Pandey R, Garg R, Darlong V, Punj J, Khanna P. Recurrent seizures in pregnancy-epilepsy or eclampsia: a diagnostic dilemma?A case report. AANA J 2011;79(5):388—90.

45. Hart LA, Sibai BM. Seizures in pregnancy: epilepsy, eclampsia,and stroke. Semin Perinatol 2013;37(4):207—24.

46. Rudra CB, Williams MA. Monthly variation in preeclampsia prevalence: Washington State, 1987—2001. J Matern Fetal Neonatal Med 2005;18(5):319—24.

47. Phillips JK, Bernstein IM, Mongeon JA, Badger GJ. Seasonal variation in preeclampsia-based on timing of conception. Obstet Gynecol 2004;104(5 Pt 1):1015—20.

48. Lee W, O’Connell CM, Basket TF. Maternal and perinatal outcomes of eclampsia: Nova Scotia, 1981—2000. J Obstet Gynecol Can 2004;26(2):119—23.

49. Cleary-Goldman J. Impact of maternal age on obstetric outcome.Obstet Gynecol 2005;105(5):983—90.

50. Lawoyn TO, Ami F. Epidemiologic aspects of pre-eclampsia in Saudi Arabia. East Afr Med J 1996;73(6):404—6.

51. Ros HS, Cnattingius S, Lipworth L. Comparison of risk factors for preeclampsia and gestational hypertension in a population-based cohort study. Am J Epidemiol 1998;147(11):1062—70.

52. Sibai BM. Risk factors, pregnancy complications, and prevention of hypertensive disorders in women with pregravid diabetes mellitus. J Matern Fetal Med 2000;9(1):62—5.

53. Wong LFA, Stuart B, Gleeson N. Triploidy partial mole and proteinuric hypertension. J Obstet Gynaecol 2007;27(4):424—5.

54. Bdolah Y, Lam C, Rajakumar A, Shivalingappa V, Mutter W,Sachs BP, et al. Twin pregnancy and the risk of preeclampsia:Bigger placenta or relative ischemia? Am J Obstet Gynecol 2008;198(4):428.e1—6.

55. Cassell KA, O’Connell CM, Baskett TF. The origins and outcomes of triplet and quadruplet pregnancies in Nova Scotia: 1980 to 2001. Am J Perinatol 2004;21(8):439—45.

56. Lopez-Llera M, De la Luna-Oslen E, Nis-Ramjos J. Eclampsia in twin pregnancy. J Reprod Med 1989;34(10):802—6.

57. Trogstad L, Magnus P, Stoltenberg C. Pre-eclampsia: risk factors and causal models. Best Pract Res Clin Obstet Gynaecol 2011;25(3):329—42.

58. Duckitt K, Harrington D. Risk factors for pre-eclampsia at antenatal booking: systematic review of controlled studies. Br Med J 2005;330(7491):565.

59. Bodnar LM, Catov JM, Klebanoff MA, Ness RB, Roberts JM. Prepregnancy body mass index and the occurrence of severe hypertensive disorders of pregnancy. Epidemiology 2007;18(2):234—9.

60. Belogolovkin V, Eddleman KA, Malone FD, Sullivan L, Ball RH, Nyberg DA, et al. The effect of low body mass index (BMI)on the development of gestational hypertension and preeclampsia. J Matern Fetal Neonatal Med 2007;20(7):509—13.

61. Halimi S, Halimi SM. Eclampsia and its association with external factors. J Ayub Med Coll Abbottabad 2010;22(3):110—2.

62. Dempsey JC, Williams MA, Luthy DA, Emanuel I, Shy K.Weight at birth and subsequent risk of preeclampsia as an adult.Am J Obstet Gynecol 2003;189(2):494—500.

63. Shennan AH, Redman C, Cooper C, Milne F. Are most maternal deaths from preeclampsia avoidable? Lancet 2012;379(9827):1686—7.

64. Wang W, Alva S, Wang S, Fort A. Levels and trends in the use of maternal health services in developing countries. DHS comparative reports no. 26. Calverton, MD, United States: ICF Macro;2011.

65. Svensson E, Reas DL, Sandanger I, Nyg?rd JF. Urban-rural differences in BMI, overweight and obesity in Norway (1990 and 2001). Scand J Public Health 2007;35(5):555—8.

66. Esplin MS, Fausett MB, Fraser A, Kerber R, Mineau G, Carrillo J. Paternal and maternal components of the predisposition to preeclampsia. N Engl J Med 2001;344(12):867—72.

67. Laivuori H, Lahermo P, Ollikainen V, Widen E, H?iv?-M?llinen L, Sundstr?m H. Susceptibility loci for preeclampsia on chromosome 2p25 and 9p13 in Finnish families. Am J Hum Genet 2003;72(1):168—77.

68. Deley L. Maternal mortality associated with hypertensive disorders of pregnancy in Africa, Asia, Latin America and Caribbean.Br J Obstet Gycecol 1992;99(7):547—53.

69. World Health Organization. The world health report 1998: life in the 21st century; a vision for all. Geneva: World Health Organization; 1998.

70. Ghosh MK. Maternal mortality: a global perspective. J Reprod Med 2001;46(5):427—33.

71. World Health Organization. The hypertensive disorders of pregnancy. Report of a WHO study group. WHO technical report series 758. Geneva: World Health Organization; 1987.pp. 63—9.

72. Say L, Chou D, Gemmill A, Tun?alp ?, Moller AB, Daniels J,et al. Global causes of maternal death: a WHO systematic analysis. Lancet Glob Health 2014;2(6):e323—33.

73. Government of India. Eleventh five year plan 2007—2012. Volume 2. New Delhi: Planning Commission, Government of India;2007.

Sutapa Agrawal, PhD

Epidemiologist, Centre for Chronic Conditions and Injuries, Public Health Foundation of India, Plot no. 47, Sector 44,Gurgaon, Haryana 122002, India Tel.: +91-124-4781400 Ext 4419;+91-9650-155334

E-mail: sutapa.agrawal@phfi.org;sutapaiips@rediffmail.com

19 October 2015;Accepted 6 February 2016