Determinants of preeclampsia among pregnant and laboring mothers managed at Wolaita Sodo university comprehensive specialized hospital, South ethiopia: an unmatched case-control study

Abstract

Background

Worldwide, preeclampsia continued to be one of a leading cause of maternal and perinatal morbidity and mortality. Indeed, developing countries like Ethiopia face significant public health challenges because of scarce data and limited applications of determinants of preeclampsia. Therefore, the aim of this research was to identify determinants of preeclampsia among pregnant and laboring mothers who came for antenatal care and labor and delivery services.

Methods

We conducted an unmatched case-control study among women diagnosed with preeclampsia and admitted to Wolaita Sodo University Comprehensive Specialized Hospital. Two hundred seventy women with case to control ratio of 1:2 were participated. We computed descriptive statistics like frequencies, percentages, mean, and standard deviation. An independent sample t-test was computed to compare the difference in means of cases and controls. Chi-squared test was used to compare categorical variables between cases and controls. To assess the association between dependent and independent variables, we carried out bivariable and multivariable logistic regressions. Variables whose p-value < 0.05 were considered as having strong association and strength of association was measured using an odds ratio (OR) with its corresponding 95% confidence level (CI).

Results

Pre-pregnancy overweight (OR = 6.7; 95%CI: 1.1–41.3 p = 0.0035), hypertension in the family member (OR = 3.1, 95%CI: 1.4–6.8, p = 004), multifetal gestation (OR = 5.4, 95% CI: 1.1–26, p = 003) and increased mid-gestation diastolic blood pressure of ≥ 80mmgh (OR = 7.6, 95%CI: 2.54–22.7, p = 0.002) were determinants of preeclampsia. Drinking coffee during pregnancy (OR = 0.27, 95%CI: 0.07, 0.94, p = 0.015) and iron and folic acid supplementation during pregnancy (OR = 0.48, 95%CI: 0.25–0.96, p = 0.041) were identified as protective factors.

Conclusion

Odds of preeclampsia was higher in women with pre-pregnancy overweight, family history of hypertension, multifetal gestation, and high mid-gestation diastolic blood pressure, whereas supplementation of iron and folic acid lowers the risk of preeclampsia. Therefore, health care providers should emphasis on preconception care and thorough clinical evaluation during pregnancy, which may provide opportunity to detect women likely to develop preeclampsia. Nutritional intervention and conducting further research on protective factors like coffee intake is recommended to clarify the inconsistent findings.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12884-025-08176-w.

Introduction

Preeclampsia is pregnancy specific condition characterized by the presence of hypertension and significant proteinuria (2 + + or more on urine dipstick measurement) or evidence of biochemical or hematological derangement occurring for the first time after 20 weeks of gestation. Diagnostic criteria for preeclampsia is presence of blood pressure (BP) ≥ 140/90 or more on two occasions at least 4 h apart with proteinuria or BP ≥ 160/110 mm Hg on a single measurement after 20 weeks of gestation [1].

According to a multicountry survey by the World Health Organization, 18% of pregnant women die because of hypertensive disorders in pregnancy [2]. Globally, preeclampsia is one of the most common contributor to maternal and perinatal morbidity and mortality. In about 30–40% of women developed adverse outcome while about 25% of cases had poor perinatal outcome [3, 4].

The prevalence of preeclampsia is not clearly figured in Ethiopia, but studies from different regions show that it is about 12%. From this, it is responsible for 40% of poor maternal outcome and 25% of perinatal outcome. Together with other hypertensive disorders of pregnancy, this proportion of poor outcome contributed to about 16.9% maternal death in Ethiopia [5–7].

High body weight, age, parity, multifetal gestation, history of preeclampsia, history of hypertension, pregestational or gestational diabetes mellitus, preeclampsia among family members were found to be risk factors for preeclampsia [8, 9]. Nutritional counseling during antenatal care (ANC) visits, fruit and vegetable consumption and antenatal follow up were risk reducing factors [9, 10]. In some studies, drinking coffee during pregnancy is found to be a protective factor whereas; others fail to find any association between drinking coffee and preeclampsia. Because of this inconsistency, they recommended further broader studies [11, 12]. Supplementation of folic acid is protective and iron supplementation has conflicting findings in different studies on preeclampsia [13, 14].

Eventhough indications to use aspirin during pregnancy is subject of controversy, with risk assessment early initiation of low dose aspirin is highly preventive of preterm preeclampsia [15–17]. The influence of calcium in reducing incidence of pre-eclampsia is dependent on the baseline calcium intake of the population and pre-existing risk factors (teenagers, age ≥ 40, and previous history of preeclampsia or HTN). Therefore, calcium supplementation is particularly important for women in developing countries compared to developed countries and those with underlying risk factors. Even if calcium supplementation is particularly important for women in developing countries, still it has suboptimal global implementation. This could be explained by difficulty in assessment of dietary calcium intake, cost, storage, transport and the likes [18, 19].

In Ethiopia, preeclampsia is among the leading causes of maternal and perinatal morbidity and mortality. To alleviate this problem, the government of Ethiopia has put various strategies into practice. Among these, increasing demand for services, creating awareness, access to emergency obstetric care and improving quality of health services were included. Addition of more health facilities, increased availability of supplies and goods; and training and placement of competent health professionals were implemented plan of actions [10, 20]. Poor documentation and lack of evidence on predictors of preeclampsia, late ANC initiation and less follow up, in Ethiopia, made difficult the controlling strategies of its complication. Giving due concern to preeclampsia related maternal mortality could fasten the reduction of maternal mortality [21].

Contrasting to direct causes of maternal deaths, deaths related with preeclampsia appear to be growing and remain a major problem in both maternal and infants morbidity and mortality. Therefore, this study was sought to identify determinants of preeclampsia in Wolaita, South Ethiopia, at Specialized University Hospital (WSUCSH). The findings help produce evidences, which are most important for screening and detection of women likely to develop preeclampsia and provide facts for health policies, programs and strategies.

Methods and materials

Study design, period, and setting

We conducted an unmatched case-control study from October to December 2021 in WSUCSH. The hospital is found in Wolaita Zone located 328 km from Addis Ababa, Ethiopia’s capital. The university hospital provides delivery services for 600 parturient per month.

Participants and eligibility criteria

The reference population was all women attending for ANC follow up and labor and delivery services in WSUCSH while all the selected women were our study participants. Participants included were those who fulfill criteria for case definition, with new-onset Hypertension without proteinuria but any of the severity features of preeclampsia present and women with chronic hypertension with super-imposed preeclampsia. Cases were those who have preeclampsia. Postpartum women who develop preeclampsia within 48h of delivery were included in the study. Controls were women on ANC follow up, women admitted to labor and delivery unit without preeclampsia. Mothers with renal disease (e.g. nephrotic, nephritic syndrome), acute or chronic liver diseases not attributable to preeclampsia, women with serious medical conditions and eclamptic patients if convulsion occurred outside hospital or at arrival or at ward before diagnosis of preeclampsia were not included in study. Control participants who develop preeclampsia after 48 h of postpartum period were excluded from study considering it as atypical preeclampsia. In place of excluded controls another controls were included.

Sample size determination and sampling procedures

We calculated the required sample size by using double population formula in OpenEpi (https://www.openepi.com/Menu/OE_Menu.htm) taking fruit consumption as a protective factor from the previous study [9]. Accordingly, the proportion of preeclamptic mothers who did not consume fruit 6.2% with the odds ratio of 3.36. The ratio of controls to cases was 2, 95% confidence level with the 5% marginal error and power of 80% giving sample size 246. After taking non-response rate of 10%, the final sample size was 270 (90 cases and 180 controls).

Cases were recruited consecutively as they were diagnosed with preeclampsia until the calculated sample size was achieved. We selected two controls for each case with simple random sampling technique at time of selection of a case. Selection of control participants was made through lottery method. A total of 90 cases and 180 controls (1:2) participants were included in the study. Senior residents working in the hospital evaluate the patient clinically and laboratory tests and revising documents to diagnose preeclampsia. The consultants of obstetrics and gynecology decide for management options either conservative or termination. The investigators reviewed women’s document and women were included as cases if they fulfilled the previously mentioned diagnostic criteria. Senior residents and obstetricians in the hospital saw all cases and controls. Both cases and controls were followed for at least 48 h in ward before discharge. The data collectors are independently assigned to different service units. To prevent double data collection from same patient and records coding of patient and charts anonymously was used.

Data collection procedure and tools

A structured and pretested questionnaire was implemented to gather information through interviewing each women face to face. The questionnaires were prepared with careful review of literatures from different sources and it included information associated to socio-demographic factors, reproductive and obstetrics factors, medical and nutritional factors, personal/behavioral and family factors [9, 10, 20, 22, 23]. Besides the questionnaire, patients’ medical document was cross-checked to abstract important variables related with laboratory, clinical and obstetrics data. Harvard university food frequency questionnaire was used to determine the fruit and vegetables intake status of mothers [24]. Gathering data was performed in ANC clinic, labor and delivery unit and postnatal ward; and conducted by trained three midwives while assisted by residents and supervised by two senior residents and principal investigator.

Operational definitions

Cases: are patients who fulfil criteria for definition of preeclampsia.

Controls: are mothers who doesnot fulfil definition of preeclampsia.

Preeclampsia: Characterized by elevated BP of ≥ 140/90 mmHg in two occasions at least 4 h apart but should not be more than seven days or BP ≥ 160/110 mmHg on single measurement and occurrence of proteinuria at least + 1 and above on dipstick after 20 weeks of gestation and after 5 months of amenorrhea in women with unknown date. We considered clinical and laboratory features to diagnose case in mothers with no protein detected in urine. Clinical manifestations were respiratory compromise due to pulmonary edema, cerebral or visual disturbances, and severe persistent upper abdominal pain with vomiting. Laboratory features were low platelet below 100,000/µl, abnormalities in liver function (elevated transaminase enzymes ≥ 2 times upper limit of normal), impaired renal function (evidenced by oliguria of < 30 ml/hr, elevated creatinine (> 1.1 mg/dl) or doubled from its normal value [1, 25].

Super-imposed preeclampsia on chronic hypertension: Defined by presence of chronic underlying hypertension in women with documented BP ≥ 140/90 mm Hg before pregnancy or before 20 weeks’ gestation, or both accompanied by new-onset proteinuria or other findings.

Proteinuria: At least 30 mg of protein in one deciliter of urine (which is equivalent to at least + 1 proteinuria) using dipsticks [26].

Mid-trimester BP: BP taken at 18-24weeks of gestation [27].

Measurements

Visual urinary dipstick testing: Used to assess the level of protein excretion. We collected fresh morning urine to test for protein. Reporting system of finding: negative, trace, 1+, 2+, 3+, or 4 + which expresses amount of protein excreted in urine [4, 25, 28].

Detection of hypertension: Confirmed by measuring BP with automated or manual measurements in the health facility. We took blood pressure with standardized technique and appropriately sized cuff (i.e., covering 2/3rd of the circumference of mid-upper arm). The fifth Korotkoff sound (marked as disappearance of Korotkoff sounds) was used to designate DBP. If blood pressure is consistently higher in one arm, the arm with the higher values was used for all blood pressure measurements [4].

Height: Taken in standing position unshoed foot and interpreted in centimeters for ambulating mothers and it was taken in supine with leg straight from heal to crown. Height classification was based on estimated percentiles of the study participant and divided into three: ≤164 cm (25th percentiles and lower), 165–170 cm (26th −50th percentile) and > 170 cm (Above 76th percentile) [20].

Body mass index (BMI): Obtained by dividing weight (in kg) (taken before pregnancy within 3 months and until end of first trimester (TM)) divided by height (in meter square) (kg/m²) and was categorized as follows: underweight (BMI < 18.5), normal weight (BMI = 18.5–24.9), overweight (BMI ≥ 25). Broca’s index, Wt (kg) ₌ (Height–100) + (Height–100) x 15%, was taken to estimate the ideal weight of the women who do not remember pre-pregnancy weight, because height is considered as stable. Women who were not booked for early ANC, the pre-pregnancy weight were considered to calculate BMI [20, 29].

Mid-upper arm circumference (MUAC): Measured by using tape meter at mid-point of the distance from the acromion process of the shoulder to the tip of the olecranon process of the mid-elbow with undressed [30].

Monthly income: Amount of money a given family gains per month, which was categorized into ≤ 5000 and > 5000 Ethiopian Birr (ETB) [20].

Consumption of fruit and vegetables: is evaluated using the standardized Food Frequency Questions dietary assessment tool. A list of fruits and vegetables were provided and questioned how often they eat on average during current pregnancy with the eight alternative responses (never, once per month, 2–3/month, once/week, 2–4/week, 5–6/week, once/day and ≥ two/day). For the analysis purpose it was categorized into three groups (never consumers = not consumed at all, occasional consumers = once per month to 2–4 per week and frequent consumers = 5–6 per week to ≥ two per day). For coffee consumption, frequency of drink per day, number of cups in each drink were investigated [24].

Data quality assurance

The data tool was developed in English and translated into local languages Amharic and Wolaitatigna and then back to English by independent language experts for compatibility. On 5% (~ 15 cases, 5 cases 10 controls) of sample of study population the pre-test was performed. This preliminary study was conducted in order to determine feasiblity, cost and time and sample size of main study. In this study, Hawas comprehensive comprehensive specialized hospital (3 cases and 6 controls) and Wachemo University Nigist Eleni Mohammed Memorial Comprehensive Specialized Hospital (2 cases and 4 controls) were chosen prior to data collection in order to evaluate validity and suitability of data collection tools and to ensure quality of data.

Supervisors and data collectors trained on content, administration and demonstrated ahead of collection. During data collection, supervisors and principal investigator checked each questionnaire for its completeness, consistency and correctness.

Data processing and analysis

The questionnaires were checked for completeness then entered into Microsoft Excel. We filled data into Epi data v4.6 and transferred to Statistical Package for the Social Sciences version 25 software for further analysis. Descriptive statistics was used to describe study population by using frequency, percentage, mean, and standard deviation. An independent sample t-test was computed to compare the difference in means of cases and controls. Levene’s test for equality of variance for identification of equal variance assumed or not assumed was used. P-value of < 0.05 was used to define statistically significant difference between means. To compare categorical variables between cases and controls Pearson Chi-squared test was computed. We computed both bivariable and multivariable unconditional logistic regression analysis to Identify variables independently determining the dependent variable. Variables with p-value of below 0.25 in bivariable regression were considered for multivariable logistic regression. Hosmer and Lemeshow goodness of fit test statistics (p-value > 0.05) was carried out to check logistic regression. Variables with p-value < 0.05 in the multivariable logistic regression analysis were supposed as independently important determinants of preeclampsia and the potency of association was measured using AOR at 95% confidence interval. We presented findings in texts, tables, and graphs.

Results

Socio-demographic characteristics of the participants

A total of 270 mothers, 90 (33.3%) cases and 180(66.70%) controls, participated in this study with a response rate of 100%. The mean ages of study participants were 25.73 for cases and 25.26 for controls. Most 234 (86.7%) of the participants were between the ages of 20–34 years. 20 (7.4%) of participants were teenagers and 16 (5.9%) were older than 35, which means seven (7.8%) from cases and nine (5%) from controls were more than 35years. 29 (32.2%) of cases and 49 (27.2%) of controls of study participants were rural residents. In our study, 31 (34.4%) of cases and 55 (64%) of controls had above secondary level of education. Of the study participants involved 13 (14.4%) of case and 28 (15.6%) of control were unemployed in either of governmental or non-governmental organization. The average monthly income of study participants was 5134 ETB (SD: ±3082) for cases and 5570 ETB (SD: ±4050) for controls (Table 1).

Table 1.

Socio-demographic characteristics of pregnant and laboring mothers managed at WSUCSH, ethiopia, 2021

Variables	Categories	Case (n = 90)	Control (n = 180)	X2	t-test	p-value
		N (%)	N (%)
Age	≤ 19	4 (4.4)	16 (8.9)
	20–34	79 (87.8)	155 (86.7)
	≥ 35	7 (7.8)	9 (5.0)
	Mean (SD)	25.73(± 4.7)	25.26(± 4.6)		0.79	0.43
Ethnicity	Wolaita	77 (85.6)	150 (83.3)
	Gamo	7 (7.8)	9 (5.0)
	Amhara	4 (4.4)	5 (2.8)
	Other	2 (2.2)	16 (8.8)
Religion	Protestant	71 (78.9)	131 (72.8)	2.02		0.36
	Orthodox	17 (18.9)	39 (21.7)
	Other	2 (2.2)	10 (5.6)
Educational status	Illiterate	5 (5.6)	11 (6.1)	1.82		0.61
	Primary	18 (20.0)	49 (27.2)
	Secondary	36 (40.0)	65 (36.1)
	≥ Secondary	31 (34.4)	55 (30.6)
Residence	Urban	61 (67.8)	131 (72.8)	0.73		0.39
	Rural	29 (32.2)	49 (27.2)
Occupation	Housewife	45 (50)	102 (56.7)	1.73		0.42
	Employed	32 (35.6)	50 (27.8)
	Unemployed	13 (14.4)	28 (15.6)
Monthly Income (ETB)*	≤ 5000	52(62.2)	106(58.9)
	> 5000	34(37.8)	74(41.1)
	Mean(SD)	5134 (± 3082)	5570(± 4050)		0.9	0.34

Key: *1 USD = 55ETB at the time of data collection (USD: United States Dollar; ETB: Ethiopian Birr)

The reproductive and obstetric related factors

According to antenatal care follow up assessment, 25 (27.8%) case group and 51 (28.3%) of control group had at least four time contacts. Regarding interval between pregnancy 6 (14%) of cases and 14 (15.9%) of controls had duration at least five years. The mean duration for each group was 3.7 (SD: ±3) and 3.5 (SD: ±2.5) respectively. On multifetal gestation analysis, 13 (14%) of cases and seven (3.9%) controls were found to be multiple pregnancy. Of study participants, six (6.7%) of cases and seven (3.9%) of controls changed their partner. Regarding past reproductive status, 13 (13.3%) of case participants and 10 (5.6%) of control participants had previously failed pregnancy. In assessment of mid-gestation blood pressure, 23 (56.1%) and 41 (47.7%) of cases and controls respectively had SBP ≥ 120mmHg; whereas 17 (41.5%) and seven (8.1%) of cases and controls respectively had DBP ≥ 80mmHg. Regarding previous history of preeclampsia/hypertensive disorder of pregnancy (HDP), 15 (16.7%) of cases and two (1.1%) of controls experienced this condition in the past. 19 (39.6%) of case study group and 40 (48.8%) control study group had cohabitation duration of more twelve months. The mean duration for each was 17 months (SD: ±7.8) and 19 months (SD: ±10.7) (Table 2).

Table 2.

Reproductive and obstetric factors associated with preeclampsia among pregnant and laboring mothers managed at WSUCSH, ethiopia, 2021

Variables	Categories	Case (90)	Control (180)	X2	t-test	P-value
		N (%)	N (%)
Pregnancy status	Unplanned	7 (7.8)	19 (10.6)	0.53		0.46
	Planned	83 (92.2)	161 (89.4)
Gravidity	Primigravida	50 (55.6)	94 (52.2)	0.27		0.60
	Multigravida	40 (44.4)	86 (47.8)
Gestation Number	Singleton	77 (85.6)	173 (96.1)	9.7		0.02
	Multiple	13 (14.4)	7 (3.9)
ANC Visit	< 4	65 (72.2)	129 (71.7)
	≥ 4	25 (27.8)	51 (28.3)
	Mean(SD)	3.8(1.47)	4(1.45)		−1.4	0.15
Interval of Pregnancy(yr.)	< 5	37 (86.0)	74 (84.1)
	≥ 5	6 (14.0)	14 (15.9)
	Mean(SD)	3.7(3)	3.5(2.5)		0.45	0.13
Contraceptive use	No use	47(52.2)	74(41.1)	2.99		0.08
	Hormonal Method	43(47.8)	106(58.9)
Previous PE/HDP	No	75 (83.3)	178 (98.9)	24.6		0.00
	Yes	15 (16.7)	2 (1.1)
Mid-gestation SBP	< 120	18 (43.9)	45 (52.3)
	≥ 120	23 (56.1)	41 (47.7)
	Mean(SD)	132(± 16.6)	120(± 8.7)		4.5	0.00
Mid-gestation DBP	< 80	24 (58.5)	79 (0.9)
	≥ 80	17 (41.5)	7 (8.1)
	Mean(SD)	87(± 14.7)	73(± 5.4)		5.8	0.00
Cohabitation duration	≤ 12 months	29(60.4)	42(51.2)
	> 12 months	19(39.6)	40(48.8)
	Mean(SD)	17(± 7.8)	19.7(10.7)		1.6	0.10

Nutrition related factors

The mean weight before pregnancy for cases was 65.5 Kg with SD (± 12.02) and the mean weight of controls was 60.8 Kg with SD (± 9.55). According to iron and folic acid supplementation, 68 (75.6%) of cases and 155 (86.1%) of controls were provided; while 17.4% of study participants did not take iron and folate. Regarding fruit consumption, 88 (97.8%) of cases and 179 (99.4%) of controls consumed during pregnancy Table 3).

Table 3.

Nutritional, personal and behavioral factors associated with preeclampsia among pregnant and laboring mothers managed at WSUCSH, ethiopia, 2021

Variables	Categories	Case (90)	Control (180)	X2	t-test	P-value
		N (%)	N (%)
Weight(Kg)	< 50	4(4.4)	27(15)
	50–70	62(68.9)	125(69.4)
	> 70	24(26.7)	28(15.6)
	Mean(SD)	65.5(± 12)	60.8(± 9.5)		3.5	0.001
BMI(Kg/m2)	< 18.5)	4(4.4)	19(10.6)
	18.5–24.9)	43(47.8)	117(65)
	≥ 25)	43(47.8)	44(24.4)
	Mean(SD)	24.5(± 4.3)	22.5(± 3.4)		4.26	0.00
MUAC(cm)	< 25.6	42 (46.7)	94 (52.2)
	≥ 25.6	48 (53.3)	86 (47.8)
	Mean(SD)(cm)	25.6(± 3.8)	25.1(± 3.3)		1.25	0.21
Iron and folate intake	No	22 (24.4)	25 (13.9)
	Yes	68 (75.6)	155 (86.1)	4.65		0.03
Fruit consumption	No	2 (2.2)	1 (0.6)
	Yes	88 (97.8)	179 (99.4)
Alcohol consumption	No	87 (96.7)	171 (95.0)
	Yes	3 (3.3)	9 (5.0)
Drinking coffee	No	15(16.7)	19(10.6)	2.01		0.15
	Yes	75(83.3)	161(89.4)
Physical activity (hr/day)	< 6	43(47.8)	109(60.6)
	≥ 6	47(52.2)	71(39.4)
	Mean(SD)	6.24(± 2.6)	5.73(± 2.4)		1.6
Aspirin use	No	83 (92.2)	178 (98.9)
	Yes	7 (7.8)	2 (1.1)
Family HTN	No	70 (77.8)	164 (91.1)	9.2		0.02
	Yes	20 (22.2)	16 (8.9)

Regarding BMI investigation, 43 (47.8%) of case participants and 117 (65%) of control participants had normal range BMI; while 47.8% of cases and 24.4% controls were grouped in overweight. The mean BMI for respective study groups were 22.5 (SD: ±3.4) and 22.5 (SD: ±3.4) (Fig. 1).

Fig. 1.

BMI of study participants at WSUCSH, 2021

Medical characteristics

In evaluation of chronic hypertension, nine (10%) of case and no control participants had Hypertension detected, which makes 3.3% prevalence of HTN. From all participants include in the study, 1.5% of pre-existing diabetes was Identified. Urinary tract infection was found to be 20 (22.2%) and 33 (18.3%) case and control participants respectively (Fig. 2).

Fig. 2.

Proportion of medical complications of participants at WSUCSH, 2021

The personal/behavioral and family characteristics

Experience of drinking coffee, 75 (83.3%) of cases and 161 (89.4%) of controls drinking coffee during pregnancy. 47 (52.2%) and 71 (39.4%) of women during pregnancy were engaged in a physical activity at least 6 h per day respectively in case and control groups. Chronic Hypertension within the family was found in 20 (22.2%) and 16 (8.9%) of cases and controls, respectively. Aspirin was initiated in seven (7.8%) cases and two (1.1%) of controls during pregnancy in this study (Table 3).

Determinants of preeclampsia among pregnant women

Variables that had an association at p-value ≤ 0.25 were candidate for multivariable binary logistic regression analysis. These include iron and folate supplementation, drinking coffee, maternal height, pre-pregnancy weight, pre-pregnancy BMI, contraceptive use, multifetal gestation, previously failed pregnancy, family history of hypertension, previous history of PE/HDP, mid-gestation systolic blood pressure (SBP), mid-gestation DBP and aspirin initiation.

In multivariable logistic regression analysis, variables that had strong association with preeclampsia at p-value less than 0.05 were pre-pregnancy overweight, mid-gestation DBP, multifetal gestational, family history of hypertension, iron and folate supplementation, and drinking coffee.

The odds of occurring preeclampsia in pre-pregnancy overweight women was about seven times higher compared to normal pre-pregnancy BMI (OR = 6.7, 95%CI :1.1–41.3, p = 0.0035). Women with multiple pregnancy were more than five times high likely to develop preeclampsia compared to women with singleton pregnancy (OR = 5.4, 95% CI: 1.1–26, p = 003). Occurrence of preeclampsia in women with family history of hypertension was about three times greater compared to women with no family history of HTN (OR = 3.1, 95%CI: 1.4–6.8, p = 004). Those mothers whose mid-gestation DPB ≥ 80 mmgh had nearly eight times higher odds of developing preeclampsia subsequently (OR = 7.6, 95%CI: 2.54–22.7, p = 0.002). Drinking coffee during pregnancy was Identified as 73% reduced the odds of developing preeclampsia compared to those nondrinking coffee (OR = 0.27, 95%CI: 0.07–0.94, p = 0.015). Iron and Folic acid supplementation during pregnancy was detected as 52% risk reducing factor of preeclampsia occurrence compared to non-supplemented controls (OR = 0.48, 95%CI: 0.25–0.96, p = 0.041) (Table 4).

Table 4.

Bivariable and multivariable analysis of determinants of preeclampsia among pregnant and laboring mothers managed at WSUCSH, ethiopia, 2021

Variable	Cases (n = 90)	Control (n = 180)	COR (95% CI)	OR (95% CI), P-value
	Number (%)	Number (%)
Maternal Height
≤164	43(47.8)	81(45.0)	0.27(0.07, 0.95)	0.23(0.03, 1.8), 0.45
165–170	44(48.9)	79(43.9)	1	1, 1
>170	3(3.3)	20(11.1)	0.28(0.08, 1.0)	0.44((0.05, 3.5), 0.52
BMI(PP)
<18.5	4(4.4)	19(10.6)	0.57(0.18, 1.18)	3.1(0.9, 10.4), 0.33
18.5–24.9	43(47.8)	117(65)	1	1, 1
≥25	43(47.8)	44(24.4)	2.66(1.54, 4.6)	2.4(1.3, 4.4)**, 0.035
PP weight
<50	4(4.4)	27(15)	1	1, 1
50–70	62(68.9)	125(69.4)	1.72(0.9, 3.2)	6.2(0.66,58), 0.08
>70	24(26.7)	28(15.6)	5.8(1.7, 18.8)	1.8(0.23, 13.6), 0.055
Gestation Multiplicity
Singleton	7(85.6)	173(96.0)	1	1, 1
Multiple	13(14.4)	7(3.9)	4.2(1.6, 10.1)	5.4(1.1, 26)**, 0.003
Failed pregnancy or abortion
No	78(86.7)	170(94.4)	1	1, 1
Yes	12(7.3)	10(5.6)	0.4(0.16, 0.92)	0.6(0.17, 2.01, 0.06
Previous PE/HDP
No	75(83.3)	178(98.9)	1	1, 1
Yes	15(16.7)	2(1.1)	17(3.9, 79.7)	13 (0.9, 20), 0.075
Family history of HTN
No	70(77.8)	164(91.1)	1	1,1
Yes	20(22.2)	16(8.9)	2.9(1.4, 5.9)	3.1(1.4, 6.8)**, 0.004
Mid-gestation DBP
<80	24(58.5)	79(91.9)	1	1, 1
≥80	17(41.5)	7(8.1)	7.9(2.9, 21.5)	7.6(2.54, 22.7)**, 0.002
Aspirin initiation
No	83(92.2)	178(98.9)	1	1, 1
Yes	7(7.8)	2(1.1)	0.13(0.3, 0.65)	0.3(0.03, 2.5), 0.1
Drinking coffee
No	15(16.7)	19(10.6)	1	1, 1
Yes	75(83.3)	161(89.4)	0.6(0.3, 1.2)	0.27(0.07, 0.94)**, 0.015
Iron and Folate
No	22(24.4)	25(13.9)	1	1, 1
Yes	68(75.6)	155(86.1)	0.5(0.26, 0.94)	0.48(0.25, 0.96)**, 0.041

Key: **Denotes variables which formed a significant association at p-value < 0.05 when strength of association was measured using an OR with corresponding 95% CI

Discussion

We carried out an unmatched case-control study with the aim of assessing socio-demographic, reproductive and obstetric, nutritional and medical, family and personal/behavioral factors associated with preeclampsia occurrence. Through this study, we found pre-pregnancy overweight, mid-gestation DBP, family history of hypertension and multifetal gestation as independent risk factors, whereas drinking coffee, iron, and folic acid supplementation as protective factor for preeclampsia.

Pre-pregnancy overweight was accompanied with more than six times increased risk of preeclampsia compared to those who had normal BMI before pregnancy (OR = 6.7; 95%CI: 1.1–41.3 p = 0.0035) in our nutritional assessment. This finding was consistent with findings from studies conducted in different countries [20, 31, 32]. The consistency of these findings might be similar in study design, study setting, same target group, and the efficient sample size used. A meta-analysis of cohort studies conducted on prepregnancy body mass index and the risk of preeclampsia concluded that overweight and obese women had increased risk to the development of preeclampsia [33]. Similarly, a registry-based study from Tanzania on prepregnancy maternal overweight and obesity were associated with an increased risk of preeclampsia. The exact mechanism how increased prepregnancy BMI leads to preeclampsia is not clearly stated. But, higher BMI result in impaired lipid metabolism associated with elevated free fatty acids, triglycerides and cholesterol which result in insulin resistance and endothelial dysfunction leading to hypertension [33–35]. Because overweight and obesity are potentially modifiable risk factors, we encourage weight loss prior to pregnancy in order to decrease the risk of preeclampsia.

The likelihood of developing preeclampsia in mothers who had family history hypertension (first degree relatives) was about three times increased compared to those mothers with no family history of hypertension (OR = 3.1, 95%CI: 1.4–6.8, p = 004). This finding is comparable with study done at different areas in Ethiopia [9, 10]. The development of severe preeclampsia within the same family suggests a genetic component of the disease occurrence [36]. To achieve optimal health before and during pregnancy, we recommend assessment of family history of HTN for every pregnant woman during her initial contacts.

The occurrence of preeclampsia in women with multifetal gestation was about five times increased compared to singleton pregnancy (OR = 5.4, 95% CI: 1.1–26, p = 003). This finding is comparable with findings from Addis Ababa, Ethiopia and Nepal [10, 37]. The proposed mechanism for elevated likelihood of preeclampsia development in multifetal gestation might be due to the large placental mass and increased plasma concentration of soluble fms-like tyrosine kinase-1(sFlt-1) [37]. This implies that the primary maternal health care providers should be alert when encountering women with multiple gestation. Moreover, referral of such women to specialized facility is paramount.

The odds of developing preeclampsia in mothers whose mid-gestation DBP measures ≥ 80 mmHg was more than seven times increased compared to those with mid-gestation DBP < 80 mmHg (OR = 7.6, 95%CI: 2.54–22.7, p = 0.002). A similar findings were reported by International Federation of Gynecology and Obstetrics, 2017 and study done in Sweden [4, 38]. Both studies explain its association as follows: Hemodynamic adaptation during pregnancy including dropping of blood pressure until mid-gestation might have beneficial effect on placental development, blood flow, and function. These reflect later development of placental abnormality that could be associated with mid-gestation rise in diastolic BP. Most of preeclampsia were diagnosed in women with stage I HTN in early gestation and elevated diastolic BP until mid-gestation [38]. Based on these findings, we suggest second trimester BP measurement for prevention and detection of preeclampsia.

In this study, women who drank coffee during pregnancy had 73% lower odds of developing preeclampsia compared to those who did not (OR = 0.27, 95%CI: 0.07, 0.94, p = 0.015). The current finding is in line with research study done in Japan and Netherlands [12, 39]. The possible justification could be chlorogenic and quinic acids present in coffee have antioxidant and antihypertensive effects [12]. Other finding consistent with our finding is from a prospective population-based cohort study done on maternal caffeine intake, blood pressure, and risk of hypertensive complications. In this study, women who consumed 180-360 mg (approximately 2–4 cups of coffee) of caffeine per day have reduced risk of pre-eclampsia [40]. Inconsistent finding in the same study and against of our finding is the presence of positive associations between caffeine intake and systolic blood pressure levels in first and third trimester of pregnancy. This effect could be explained by the mechanisms by which caffeine exposure affect heart rate and blood pressure through increase in serum catecholamine level, which might subsequently lead to vasoconstriction.

The other possible explanation could be increase in serum homocysteine level and caffeine could interrupt placental angiogenesis and increase the risk of HDP [40, 41]. Additional case control study conducted on caffeine consumption and incidence of hypertension in pregnancy discovered absence of any relationship between the risk of preeclampsia and different types of caffeine consumption. As a result, holistic and broader study is recommended studies in this area [11]. These controversial findings between caffeine and hpertentensive disorders in pregnancy could be due to types of caffeine contents, amount of caffeine, frequency of consumption and its preparation. Because of this, to examine the consistency of our finding, further research is needed to fully understand the relationship between caffeine intake and preeclampsia.

Furthermore, women who were supplemented with iron and folic acid during pregnancy had 52% lower odds of developing preeclampsia compared to those who were not supplemented (OR = 0.48, 95%CI: 0.25–0.96, p = 0.041). This finding is in line with research done in Ethiopia [14]. The consistency of the results are same in study setting, study design and target group; but focused on folic acid supplementation, which affects to draw conclusion. A randomized clinical trial conducted on micronutrient supplementation during pregnancy and the risk of pregnancy-induced hypertension showed that the reduced odds of pregnancy induced hypertension among younger women consuming iron folic acid compared to folic acid [42]. Even if this research is large randomized, double-blind controlled trial and assessment of supplementation was made through home visit, it is difficult to generalized because the participants were normal weight women with no or mild anemia. Another randomized controlled trial performed on multivitamin supplementation of HIV-Positive women during pregnancy neither support nor refutes our finding. In this study, supplementation of iron and folate alone did not shown effect on preeclampsia despite all participants received iron and folate. Investigators failed to generalize the results because the data collection limitation and participants involved were HIV-positives with compromised nutritional status [43]. Therefore, it is cumbersome to conclude the beneficial effect of iron and folate supplementation on preeclampsia because of inexistent evidences and it needs further research.

Strengths and limitations of study

The strength of this study was conducting case-control study design; focusing on the specific category of pregnancy induced hypertension (preeclampsia) and performing study at comprehensive hospital where variety of cases were managed. In addition, requiring fewer time and being inexpensive makes this design suitable for investigation of potential risk factors in low-setting countries like Ethiopia.

Problem of showing cause-effect relationship and existence of selection and recall biases were main limitations of our study. The small sample size in our study is another challenge, which may increase margin of error and less precision of making real association between risk factors and preeclampsia occurrence.

Besides this, late ANC initiation, few visits, gaps in evaluation and documentation and lack of transfer of documents during referral were difficulties for women from other facilities. Finally, generalization of finding may not be possible because institutional study could not be representative of pregnant women population in Wolaita Zone.

Conclusion and recommendation

In this study, different risk and protective factors of pre-eclampsia were assessed. Factors positively linked with pre-eclampsia were pre-pregnancy overweight, family history of hypertension, gestational multiplicity, and high mid-gestation DBP whereas drinking coffee and supplementation of iron and folic acid were associated with lowering risk of pre-eclampsia. Therefore, health care providers should emphasis on preconception care and thorough clinical evaluation during pregnancy, which may provide opportunity to detect women likely to develop preeclampsia.

Besides this, nutritional intervention with supplementation of iron and folic acid for all pregnant women is recommended to reduce the risk of preeclampsia occurrence. Additional research is required to understand more about protective factors of preeclampsia particularly on dietary habits like coffee intake during pregnancy.

Finally, our recommendations go to university hospital and zonal health bureaus in facilitating, capacity building, and paving way in struggle to prevent preeclampsia related maternal and perinatal morbidity and mortality.

Abbreviations

ANC

Antenatal Care

BMI

Body Mass Index

CI

Confidence interval

DM

Diabetes Mellitus

DBP

Diastolic Blood Pressure

ETB

Ethiopian Birr

MUAC

Mid-Upper Arm Circumference

PE

Preeclampsia

SBP

Systolic blood pressure

SD

Standard deviation

WSUCSH

Wolaita Sodo University Comprehensive Specialized Hospital

Authors’ contributions

B.S was the main author, involved in the interpretation, design, procurement, and analysis of the data and developed the manuscript. A.A was academic adviser, participated in project, analysis, and conceptualization of the data, and carefully revised the manuscript. W.W was academic and clinical adviser, contributed methodology, acquisition, analysis, and interpretation of the data and manuscript revision. W.B was co-author, contributed organization, analysis, and interpretation of the data and seriously revised the manuscript. B.Z was a co-author contributed analysis and interpretation of data and preparation of tables and figures. All of them provided final permission of the version for publication; and agreed to be responsible regarding all prospect of the work.

Funding

This research is funded by College of Health Sciences and Medicine, Wolaita Sodo University.

Data availability

Data is provided within the manuscript or supplementary information files.

Declarations

Ethics approval and consent to participate

Wolaita Sodo University, College of Health Sciences and Medicine Institutional Ethical Review Committee granted ethical approval. In addition, WSUCSH provided permission to carry out the research. We requested all subjects for voluntary participation and reassured them that confidentiality of information they provide. We took verbal informed consent for disclosure, understanding and voluntariness. Investigators respected the autonomy of study participants through decision-making. For the minority participants (age ≤ 18 years), we took verbal consent from their parents or caregivers. We used anonymity and omitting personal identity information to guarantee confidentiality of all study participants.

Consent for publication

Not relevant.

Competing interests

The authors declare no competing interests.