The burden of hypertensive disorders of pregnancy in Africa: A systematic review and meta‐analysis

Abstract

Hypertensive disorders of pregnancy (HDP) are a major contributor to maternal and perinatal morbidity and mortality, especially in resource‐limited settings. Little is known about the magnitude of HDP in Africa. We conducted the first systematic review and meta‐analysis to summarize available data on the prevalence of HDP in Africa. We did a comprehensive literature search to identify review paper published from January 1, 1996, to September 30, 2017, and searched the reference list of retrieved review paper. We used a random‐effects model to estimate the overall and type‐specific prevalence of HDP in Africa. We included 82 studies published between 1997 and 2017 reporting data on a pooled sample of 854 304 women during pregnancy or puerperium. Most studies were hospital‐based, conducted in urban settings across 24 countries. In this population, the overall prevalence of HDP was 100.4‰ (95% CI: 81.4‐121.2). The prevalence was 49.8‰ (95% CI: 32.3‐70.7) for gestational hypertension, 14.7‰ (95% CI: 11.6‐18.2) for chronic hypertension, 9.2‰ (95% CI: 4.2‐16.0) for superimposed preeclampsia, 44.0‰ (95% CI: 36.7‐52.0) for preeclampsia, 22.1‰ (95% CI: 14.8‐30.8) for severe preeclampsia, 14.7‰ (95% CI: 8.1‐23.2) for eclampsia and 2.2‰ (95% CI: 1.2‐3.4) for HELLP syndrome. Prevalence of HDP was significantly higher in Central and Western Africa; there was a consistent tendency of increasing HDP prevalence with income at the country level. In conclusion, the burden of HDP in Africa is high, with about one in 10 pregnancies affected. The higher rate of severe forms of HDP that are associated with significant maternal and perinatal mortality is a major concern in the region.

1. INTRODUCTION

Hypertension is the leading contributor to the global burden of disease, with about 1 billion adults affected worldwide and 9 million associated deaths annually.1, 2 Hypertensive disorders in pregnancy (HDP) are also a major threat to global health.3, 4 They complicate 5.2%‐8.2% of pregnancies globally5 and are associated with an increased risk of adverse fetal, neonatal, and maternal outcomes including premature delivery, fetal growth restriction, intra‐uterine death, renal or hepatic failure, hemorrhage, and stroke.6, 7 Women with HDP also have an increased lifetime risk of cardiovascular disease.8, 9

Albeit a lack of consistency in its definition, HDP refer to a spectrum of conditions of vascular origin and systemic manifestations caused by a mixture of genetic and acquired factors, which occur during pregnancy.5, 7, 10, 11 In the 1990s, the US National High Blood Pressure Education Program Working Group on High Blood Pressure in Pregnancy classified HDP into four entities: chronic hypertension, preeclampsia‐eclampsia, preeclampsia superimposed on chronic hypertension, and gestational hypertension.12, 13 This widely used classification takes into account the time of appearance of the condition in relation to pregnancy.14, 15

African populations have experienced a surge in hypertension prevalence over the past two decades. According to the World Health Organization (WHO), Africa has the highest age‐standardized prevalence of hypertension with 46% of adults older than 25 years being affected.2 Furthermore, many risk factors of HDP including early pregnancies, multiparity, older maternal age, maternal anemia and infections, and low socioeconomic status are highly prevalent in African women,16 suggesting a high prevalence of HDP in the region. More importantly, previous reports have shown that in Africa where access to antenatal care (ANC) and emergency obstetric care services are limited, HDP are among the top leading causes of maternal and perinatal mortality.3, 4, 17

While HDP have been extensively investigated in high‐income settings, little is known about the prevalence of HDP in Africa.16 Yet, African populations have been markedly underrepresented in recent reports on the global burden of HDP.5 Hence, to the best of our knowledge, we conducted the first systematic review and meta‐analysis to estimate the overall and type‐specific prevalence of HDP in Africa, in order to inform policymakers and guide strategies for early detection and appropriate management of these conditions in the region.

2. METHODS

This review is reported in accordance with the Meta‐analyses and Systematic reviews of Observational Studies (MOOSE) guidelines. Its protocol has been published in the PROSPERO International Prospective Register of systematic reviews, registration number: CRD42016047509.

2.1. Literature search

We performed a comprehensible bibliographic search to identify relevant review paper published on the prevalence of HDP in Africa from January 01, 1996 to October 16, 2017, in PubMed/MEDLINE, Excerpta Medica Database (EMBASE), African Journals Online (AJOL), and African Index Medicus. No language restriction was applied. The search strategy for each database is available in Table S1. We scrutinized the reference list of all relevant review paper to identify other potential data sources.

2.2. Selection of studies for inclusion in the review

We included all cross‐sectional, case‐control, and cohort studies reporting on the prevalence (or enough data to compute this estimate) of chronic hypertension, gestational hypertension, superimposed preeclampsia, preeclampsia, severe preeclampsia, and/or eclampsia in women residing in African countries, or providing enough data to compute it. We excluded studies conducted among populations of African origin residing outside Africa, letters, reviews, commentaries, and editorials. We also excluded case‐control studies in which participants were selected on the basis of the diagnosis of a HDP, in order to avoid selection bias. For studies published in more than one paper, the most comprehensive one reporting the largest sample size was considered. Titles and abstracts of review paper retrieved from literature search were independently screened by two investigators (JJN and JJB), and full‐texts of those potentially eligible were obtained and further assessed for final. Disagreements were resolved through consensus.

2.3. Strategy for data compilation and definition of HDP

The definition of HDP was either clearly reported by authors, in which case we relied on it; or not explicitly stated, in which case we assumed that the definition used was derived from the widely used classification of HDP.13 Thus, criteria used to define PE in most studies included in this review were an elevated blood pressure (≥140 mm Hg systolic or ≥90 mm Hg diastolic) after 20 weeks of gestation and a high level of protein in the urine (proteinuria >0.3 g/24 hours or a urinary protein/creatinine ratio of ≥0.3, or a urine dipstick protein of 1+ if a quantitative measurement is unavailable), associated or not with various systemic manifestations.13 However, since some studies reporting on preeclampsia did not differentiate between patients with and without severe features, we pooled data from these studies into a category referred to as “PE”, irrespective of the disease severity. To reduce the risk of selection bias, studies focusing on patients presenting with severe features of preeclampsia were pooled separately in a category referred to as “severe PE.” Similarly, HELLP (hemolysis, elevated liver enzyme, low platelets) syndrome, a variant of severe PE, was analyzed separately because its definition is mainly based on biological parameters.18 Studies focusing only on ECP were also pooled distinctly. Most studies reporting on superimposed PE did not provide enough information to differentiate between patients with and without severe features and were compiled in the meta‐analysis. Since very few studies reported on the overall prevalence of HDP (all four entities combined), we pooled data from eligible studies reporting on at least three of the four entities of HDP (CH, GH, PE/ECP, superimposed PE/ECP) to get a rough estimate of the overall prevalence of HDP in Africa.

2.4. Assessment of methodological quality and reporting of data

Studies were critically appraised using the tool developed by Hoy and colleagues.19 A score of 1 (yes) or 0 (no) was assigned for each item, and scores were summed across items to generate an overall quality score that ranged from 0 to 10. Studies were then classified as having a low (>8), moderate (6‐8), or high (≤5) risk of bias. Two investigators (UFN and AMJ) assessed methodological quality of selected studies, and their assessments were independently reviewed by a third investigator (JJN) with disagreements being resolved through consensus.

2.5. Data extraction and management

A preconceived and standardized data extraction form was used to collect information on first author's name, study country, African subregion (according to the United Nations’ classification), year of publication, type of study, study design, timing of data collection, number of participants, study population, mean or median age of the population, diagnostic criteria for each HDP type, number of cases by HDP type. Two investigators (UFN and AMJ) extracted the data from individual studies, and all extracted data were cross‐checked by a third investigator (JJN) with disagreements being resolved through consensus.

2.6. Statistical analysis

Prevalence was expressed for each and any type of HDP per 1000 (‰) pregnant or parturient women, whatever data were available. For cohort studies, we used baseline data and for case‐control studies, we used the number of cases of HDP and the total number of pregnant or parturient women reported during a given period of time. Data were analyzed using the “meta” packages of the statistical software R (version 3.4.4; The R Foundation for statistical computing, Vienna, Austria). Unadjusted prevalence of HDP was recalculated based on crude numerators and denominators provided by individual studies. To minimize the effect of studies with extremely small or extremely large prevalence estimates, the variance of the study‐specific prevalence was stabilized with the Freeman‐Tukey double arcsine transformation before pooling the data within a random‐effects meta‐analysis model.20 Egger's test served to assess the presence of publication bias.21 A P‐value <0.10 on the Egger test was considered indicative of statistically significant publication bias. Heterogeneity was evaluated by the chi‐square test on Cochrane's Q statistic,22 which was quantified by H and I² values. The I² statistic estimates the percentage of total variation across studies due to actual between‐study differences rather than chance. Generally, I² values <60%‐70% indicate the presence of substantial heterogeneity.23

Subgroup analyses were performed to assess the potential effect of some variables including the African subregion (Central, Eastern, Northern, Southern, Western), the number of study sites (single vs multisite), the study area (urban vs rural), the timing of data collection (prospective vs retrospective), the income level determined at the country scale (low, lower‐middle or upper‐middle), and the proportion of primiparous (≤50% vs >50%) in the study population, on the prevalence of the various forms of HDP in Africa.

2.7. Role of the funding source

There was no funding source for this study. The corresponding author had full access to all study data and had final responsibility for the decision to submit the paper for publication.

3. RESULTS

3.1. Review process and study characteristics

We identified 3252 records; after elimination of duplicates, 3078 remained. After screening titles and abstracts, we retained and assessed 163 full‐text papers for eligibility. Finally, 82 full‐texts were included (Figure S1). The inter‐rater agreement for study selection was high (κ = 0.92). Twenty‐four (29.3%), 54 (65.8%), and four (4.9%) studies had low, moderate, and high risk of bias, respectively.

The characteristics of included studies are summarized in the Tables S2 and S3. Data were from 24 countries including 854 304 participants. All studies were hospital‐based with most of them conducted in gynecology/obstetric units. Most studies were conducted in a single center and in urban areas. The timing of data collection was often retrospective, the sampling method was often consecutive, and included women at any age of pregnancy, until 6 weeks postpartum. Most of studies did not clearly define the condition and there was high heterogeneity in the definitions of HDP across studies.

3.2. Overall and type‐specific prevalence estimates of HDP in Africa

The pooled prevalence of HDP (all types combined) was 100.4‰ (95% CI: 81.4‐121.2; Figure 1). The most frequent forms of HDP were GH and PE with prevalence rates of 49.8‰ (95% CI: 32.3‐70.7; Figure 2) and 44.0‰ (95% CI: 36.7‐52.0; Figure 3), respectively. The pooled prevalence of severe PE was 22.1‰ (95% CI: 14.8‐30.8; Figure 3), while ECP (Figure 4) and CH (Figure 5) had similar prevalence rates of 14.7‰ (95% CI: 11.6‐18.2 for ECP and 8.1‐23.2 for CH). The less frequent forms of HDP were superimposed PE (9.2‰; 95% CI: 4.2‐16.0; Figure S2) and HELLP syndrome (2.2‰; 95% CI: 1.2‐3.4; Figure S3). While there was substantial heterogeneity (Table 1), no publication bias was found, both for overall (Figure S4) and all type‐specific HDP (Table 1, Figures S5‐S10).

Figure 1.

Meta‐analysis of the prevalence of overall hypertensive disorders of pregnancy

Figure 2.

Meta‐analysis of the prevalence of gestational hypertension

Figure 3.

A, Meta‐analysis of the prevalence of preeclampsia. B, Meta‐analysis of the prevalence of severe preeclampsia

Figure 4.

Meta‐analysis of the prevalence of eclampsia

Figure 5.

Meta‐analysis of the prevalence of chronic hypertension

Table 1.

Summary statistics of the prevalence of hypertensive disorders in pregnancy in Africa

	Prevalence ‰ (95% confidence interval)	N studies	N participants	I2	P heterogeneity	P Egger
Overall	100.4 (81.4‐121.2)	23	139 916	99.3	<0.0001	0.930
Gestational hypertension	49.8 (32.3‐70.7)	23	85 088	99.3	<0.0001	0.242
Preeclampsia	44.0 (36.7‐52.0)	39	373 344	99.1	<0.0001	0.320
Severe preeclampsia	22.1 (14.8‐30.8)	15	105 927	98.5	<0.0001	0.698
Superimposed preeclampsia	9.2 (4.2‐16.0)	9	51 713	97.3	<0.0001	0.122
HELLP syndrome	2.2 (1.2‐3.4)	18	212 674	96.3	<0.0001	0.834
Eclampsia	14.7 (11.6‐18.2)	57	574 691	99.0	<0.0001	0.202
Chronic hypertension	14.7 (8.1‐23.2)	14	153 585	99.2	<0.0001	0.654

HELLP, hemolysis, elevated liver enzymes, low platelet count.

3.3. Subgroup analysis of HDP prevalence in Africa

Subgroup analysis of HDP prevalence by African subregion showed a significantly higher pooled prevalence in Central (84.2‰; 95% CI: 80.3‐88.2 for PE, and 12.4‰ [4.0‐25.2] for ECP) and Western (67.6‰ [24.7‐129.1] for GH, 47.2‰ [32.5‐64.5] for PE, and 20.7‰ [15.3‐26.9]) Africa (Tables S4‐S6). The lowest prevalence estimates were observed in Northern (15.6‰ [9.7‐22.9] for GE, 38.6‰ [22.2‐59.2] for PE, and 3.1‰ [0.3‐8.5] for ECP) and Eastern (38.7‰ [14.1‐74.2] for GE, 36.5‰ [29.9‐43.8] for PE, and 10.2‰ [6.3‐8.6] for ECP) Africa (Tables S4‐S6).

Although the difference was not statistically significant, there was an overall trend of increasing prevalence of HDP with the country's level of income in Africa. Indeed, HDP prevalence in low‐income, lower‐middle income, and higher‐middle income countries was 26.4‰ (95% CI: 11.5‐46.8), 47.6‰ (16.5‐93.6), and 79.8‰ (35.8‐138.9), respectively for GH (P = 0.083); 40.1‰ (29.1‐52.8), 44.7‰ (32.7‐58.3), and 48.7‰ (31.7‐68.7), respectively for PE (P = 0.738); and 11.8‰ (8.4‐15.6), 9.6‰ (1.3‐24.8), and 17.5‰ (12.7‐23.1), respectively for ECP (P = 0.182; Tables S4‐S6).

Further subgroup analysis showed a significantly higher prevalence of PE in urban dwellings (49.1‰, 95% CI: 36.8‐63.2) beside rural settings (32.6‰ [28.5‐37.0], P = 0.011; Table S5). However, prevalence of ECP tended to be higher in rural areas (36.8‰ [9.5‐80.8]) than in urban settings (12.8‰ [9.3‐16.8]; Table S6).

The remaining variables included in the subset analysis—timing of data collection, number of study sites, place of patients’ enrollment, and proportion of primiparous—did not influence significantly the prevalence of HDP in Africa (Tables S4‐S6).

4. DISCUSSION

This systematic review suggests that about 10% of women in Africa develop any form of HDP at some point during pregnancy or puerperium. Among the various clinical types of HDP, PE (5.3%) whether superimposed (0.9%) or not (4.4%), and regardless of the degree of severity, was found to be the commonest form of HDP in Africa. Estimates also varied substantially according to subregion, with significantly higher prevalence of HDP among women in sub‐Saharan Africa beside Northern Africa.

The prevalence of HDP in Africa obtained in this review (10.0%) exceeds the global estimates (5.2%‐8.2%) described in a recent report.5 Additionally, this prevalence seems to be higher than that observed in the United States (6%‐8%)24, 25; in the UK (5.4%‐6.7%),5, 26 and in Australia (8.2%).5, 27 Many factors could explain these differences. First, most studies included in this review were hospital‐based unlike reports from other settings, suggesting that the prevalence of HDP in Africa may have been overestimated as pregnant women in limited resource settings seem less likely to utilize ANC and emergency obstetric care services when they develop or have a history of pregnancy‐related complications.28, 29 However, since many forms of HDP (GH, CH, and PE without severe features), may be clinically unapparent at onset,6 a significant number of pregnant women with subclinical HDP could also have been missed by studies included in this report, in a context where utilization of ANC and institutional delivery services is relatively low.30, 31 Yet, among pregnant women who attend ANC in Africa, the timing of visits, considered as a practical way of preventing adverse pregnancy outcomes, is not always adequate33; even when adequate, the quality of ANC care is often suboptimal, as blood pressure and proteinuria measurement may not be routinely available.34 Second, the uneven distribution of risk factors for HDP between high‐income countries and Africa, where women are more frequently of low socioeconomic background, are more likely to have a maternal infection or anemia,16 could partly explain the higher rate of HDP in the region.

The prevalence of PE in Africa (5.3%), seems slightly higher than the 3%‐5% rate obtained in the United States.6, 35 Importantly, this review shows that about half cases of PE in Africa are diagnosed with severe features, which is substantially greater compared to the United States where severe PE accounts for about 25% of all cases of PE.36 In addition, on average, one case of ECP is diagnosed for every 3‐4 cases of PE in Africa, which is outstandingly higher than the ratio in the United States where one case of ECP occurs in 100‐200 cases of PE.37 Although variations in study methodology and population structure could contribute to explain the differences observed between Africa and the United States, these findings may also reflect the poor access of pregnant women to health services in Africa, resulting in a late diagnosis of the condition in affected women, further contributing to the high burden of maternal and perinatal morbidity and mortality observed in the region.3, 17

Our results are indicative that HDP prevalence in Africa tends to increase with the country's level of income. Moreover, the prevalence of PE was significantly higher in urban settings (49.1‰, 95% CI: 36.8‐63.2), compared to rural areas (32.6‰ [28.5‐37.0]), unlike the prevalence of ECP which appeared to be three times higher in rural areas. This interesting finding could be explained by the fact that in more developed countries and in urban settings in Africa, health systems are usually more advanced and provided with skilled health care providers; furthermore, populations are generally better educated, and utilization of maternity services is likely to be higher. Conversely, in less developed countries and in rural settings, access of women to ANC and emergency obstetric care is limited; health facilities are poorly equipped, and most women show up in health facilities when symptoms are serious.

Given the significant proportion of severe forms of HDP which contribute to the high maternal and perinatal adverse outcomes in rural settings, integration of culturally suitable policies to early detect and adequately manage women with HDP into programs to reduce mother and child mortality should be a priority in Africa. Special attention should be given: (a) to raise awareness of the population on the need for pregnant women to have at least one consultation during the first half of the pregnancy for a systematic blood pressure measurement and a timely administration of calcium or low‐dose (60‐150 mg/d) oral aspirin to prevent PE and some perinatal adverse outcomes in eligible women38; (b) to search actively for warning signs in the presence of which pregnant women should immediately receive further investigation and appropriate care; and (c) on the importance of primary prevention, that is, maintaining a healthy weight and diet, to avoid anemia, to control and monitor blood pressure and blood sugar levels prior to and throughout pregnancy.

Results of this review showed that HDP prevalence seems significantly higher in sub‐Saharan African countries than in Northern Africa, where the number of blacks is proportionally lower. Although this association could not be assessed at the individual level, this finding suggests that blacks may have some genetic features that predispose them to HDP. While racial differences in the prevalence of PE have been difficult to assess because of confounding by socioeconomic and cultural factors, some reports in the United States have shown that black women had higher risk of developing HDP than whites.39, 40 Indeed, Hypertension in Africans (and African Americans) is associated with salt sensitivity, which correlates with a decrease in Nitric Oxide (NO)‐dependent vasodilation.41 And this decrease in NO levels was found to be more pronounced among blacks compared to other races.42

This study has some limitations. The definition of HDP was not consistently reported, and even when reported, criteria for defining HDP were not always clearly described or always include information about the device used to measure BP and/or proteinuria, the technique for BP measurement. This is mostly due to the fact that most of the included studies (~55%) were retrospective, and therefore, it was not possible for authors to indicate the definition used to define HDP. Cases of HDP were those labeled as such by the treating physician. Furthermore, we found substantial heterogeneity common in meta‐analysis of prevalence studies, which was partly explained by some study‐specific characteristics as demonstrated by subgroup analyses. This highlights the need to standardize the definition of HDP in Africa.

In conclusion, this review showed that the prevalence of HDP in Africa is among the highest worldwide. Given the high burden of HDP in Africa on one hand and the significant proportion of severe forms contributing to the adverse maternal and perinatal outcomes on the other, integration of culturally suitable policies to early detect and adequately manage women with HDP into programs to reduce mother and child mortality should be a priority in Africa. Pending the development of a valid and reproducible algorithm to screen and identify women at higher risk of HDP, emphasis should be made on primary prevention, particularly among women of low socioeconomic background or living in rural areas. To that end, primary health care providers in the area should be adequately trained on the diagnosis and management of HDP, and medications should be made available in primary health care facilities. Given the high variability in definition of HDP across African settings, it is also imperative to harmonize the criteria to define this condition on the continent.

CONFLICT OF INTEREST

We declare no competing interests.

AUTHOR CONTRIBUTION

JJN, JJB, and JFD conceived the study, and together with JRN, ADK, and AMJ developed the protocol. JJN and JJB did the literature search. JJN and JJB selected the studies. UFN, AMJ, and JJN extracted the relevant information. JJN and JJB synthesized the data. JJN, JJB, and JFD wrote the first draft of the paper. JJN, JJB, UFN, ADK, AMJ, JRN, and JFD critically revised successive drafts of the paper and approved its final version. JJN is the guarantor of the review.

Supporting information

 

Noubiap JJ, Bigna JJ, Nyaga UF, et al. The burden of hypertensive disorders of pregnancy in Africa: A systematic review and meta‐analysis. J Clin Hypertens. 2019;21:479–488. 10.1111/jch.13514