Effect of prenatal alcohol consumption on maternal anemia among pregnant women in Gondar town, Northwest Ethiopia: a retrospective cohort study

Abstract

Background

Alcohol consumption during pregnancy can lead to multiple health, social and behavioral problems for both the mother and the offspring. Anemia is one of the major public health concerns and causes of morbidity and mortality among pregnant women with poor maternal and fetal outcomes in developing countries. This study aimed to identify the link between prenatal alcohol consumption and anemia among pregnant women in Gondar town, Northwest Ethiopia.

Methods

A facility-based retrospective cohort study was conducted among 1669 pregnant women who were booked in for antenatal care in Gondar town. We used a two-stage random sampling technique to recruit and include study participants in the cohort. Data were collected using an interviewer-administered questionnaire. Multivariable analysis was performed to examine the association between reported prenatal alcohol exposure (non-hazardous and hazardous) and anemia in pregnancy using log-binomial regression modeling. The burden of anemia in pregnancy was reported using the adjusted risk ratio (ARR) and population-attributable risk.

Results

The prevalence of anemia during pregnancy was 14.86% (95% CI 13.23 to 16.65%). Hazardous alcohol consumption during pregnancy was significantly associated with anemia in pregnancy (ARR=2.24; 95% CI 1.60, 3.15). The adjusted population-attributable risk of anemia in pregnancy related to hazardous alcohol consumption during pregnancy was 7.68%.

Conclusions

This study revealed that anemia during pregnancy continues to be a public health concern. Regardless of timing, there is a clear association between prenatal alcohol consumption and anemia during pregnancy that suggests a need for targeted prenatal alcohol use screening, and ongoing intervention for alcoholic pregnant women.

alcohol use, anemia, Ethiopia, pregnant women

Introduction

Harmful alcohol use, a pattern of alcohol consumption that can lead to various health problems directly related to alcohol,¹ is a significant contributor to the global burden of disease and risk factors for many premature deaths and disabilities.^2,3 According to a WHO report, alcohol consumption leads to >200 different diseases and injuries.⁴ Negative health impacts and unacceptable social consequences of alcohol consumption are common in women. Likewise, alcohol consumption during pregnancy can lead to multiple health, social and behavioral problems for both the mother and the offspring.⁵

Maternal health is crucial for the overall health and well-being of the infant. Thus, the health of the mother is the health of the infant indirectly. Alcohol crosses the placental barrier and damages the development of an embryo or developing fetus.⁶ There is no well-known threshold of alcohol consumption during pregnancy that is certain to be safe for fetuses and pregnant mothers.⁷ However, many women worldwide are persistently drinking alcohol throughout pregnancy irrespective of the guidelines' recommendations.⁸ In addition to other health risks, alcohol intake during pregnancy results in fetal alcohol spectrum disorders, a group of adverse effects that can occur in an individual whose mother drank alcohol while pregnant.⁹ Moreover, alcohol consumption during pregnancy has a double burden adverse effect. Excessive alcohol drinking (which includes binge drinking: ≥4 drinks on one occasion; heavy drinking: ≥8 drinks per week; drinking while pregnant; and underage drinking: any alcohol use by people aged <21 y)¹⁰ during pregnancy is a risk factor for spontaneous abortion¹¹ that is associated with harmful health outcomes for the mother and developing fetus, from which ensue maternal physical and psychosocial risks.^12,13 Any type of alcohol drinking during pregnancy can have severe consequences for the mother, such as nausea, weakness and damage to her body.¹⁴ Conversely, a limited number of studies conducted in other countries have shown that alcohol consumption during pregnancy is also associated with maternal anemia.^15,16 Specifically, women who are heavily exposed to alcohol may develop iron deficiency anemia, which is critical to address.¹⁷ Furthermore, research has shown that infants whose mothers engage in prenatal binge drinking are at a higher risk of developing iron-deficiency anemia.¹⁸ This highlights the significant impact that maternal alcohol consumption can have, not only on the mother's health, but also on the nutritional status and overall well-being of the newborn.

Anemia continues to be a significant public health concern and cause of morbidity and mortality among pregnant women, with poor maternal and fetal outcomes in developing countries,^19,20 resulting in low birth weight,^21–23 premature births,^24,25 fetal cognitive impairment and death.²⁶ In Ethiopia, approximately one-third of pregnant women were anemic.²⁷ Abstaining from alcohol use during pregnancy has multiple benefits for maternal and fetal health. To attain Sustainable Development Goals of decreasing the global maternal mortality rate to <70 per 100 000 live births by 2030,²⁸ multifaceted and targeted interventions and collaborative efforts are required. Avoiding alcohol consumption is crucial because it can prevent alcohol-related harm to mothers and newborns.²⁹ As a result, many countries have established guidelines advising women who are pregnant or plan to become pregnant to abstain from alcohol.^30–32 Despite these recommendations, the consumption of both traditionally prepared alcoholic beverages, such as ‘Areke’, ‘Tella’ and ‘Tej’, and commercially produced alcoholic drinks, remains prevalent in Ethiopia, particularly during religious festivals and ceremonies.^33,34 Consequently, many pregnant women in the country continue to consume alcohol, which poses additional health risks.^35–39 Therefore, this study aims to address gaps in the literature by examining the relationship between prenatal alcohol use and anemia throughout pregnancy, ultimately providing insights for program planners and healthcare providers to implement effective interventions.

Materials and methods

Study design, study setting and period

A facility-based retrospective cohort study was conducted in four selected public health facilities of Gondar town: University of Gondar Comprehensive Specialized Hospital, Gondar Polyclinic, Azezo Health Center and Maraki Health Center. The town is located in Northwest Ethiopia, about 727 km from Addis Ababa, the capital city of Ethiopia. In the town, there are eight health centers and one comprehensive specialized hospital.⁴⁰ A variety of locally brewed alcoholic beverages are readily available, including Dashen beer, which is produced in the town. The current study was performed from 29 October 2019 to 7 May 2020.

Study population and participant recruitment

The study was conducted among pregnant women attending antenatal care (ANC) clinics at selected public health facilities. In Gondar town, there are eight public health centers and one referral hospital. We randomly selected three health centers (Azezo Health Center, Gondar Health Center and Marak Health Center) and purposely included one referral hospital, Gondar University Specialized Hospital. The total sample size was distributed proportionally among these facilities based on the number of ANC clients who visited in the previous 3 mo. Participants were enrolled in a prospective cohort study (the details of which are provided separately)⁴¹ if they were in the first week of their third trimester, specifically at 28 wk of gestational age. Each participant was selected through systematic random sampling.

All pregnant women who were residents in Gondar town were the source population. Medical health facility documents or log booklets of mothers were reviewed to collect necessary information relating to medical risk factors and some potential confounders. The study included all pregnant women residing in Gondar town for at least 6 mo and aged ≥18 y. Women who were uncertain about remembering the first day of the last menstrual period, or who had an irregular cycle of menstruation, or who experienced a difference of >7 d in the menstruation cycle, and for whom booking an ultrasound scan estimate was not preferred, were excluded from the study. Excluding these due to women who were uncertain about the date of their last menstrual period could introduce significant variability in estimating gestational age. Accurate gestational age was crucial for the study's objectives, and uncertainty could lead to inconsistent classification of participants, potentially skewing the study results. Moreover, variability in menstrual cycle could also complicate gestational age assessment and increase the risk of misclassification, which could affect the outcomes of the study and limit the generalizability of its findings.

Sample size determination and sampling procedures

We computed the sample size by using EPI INFO version 7.2.1.0 STAT CALC software (Epi Info version 7.2.1.0 was developed by the Centers for Disease Control and Prevention (CDC), which is based in Atlanta, Georgia, United States.) considering two-sided 95% confidence level, power of 80%, the ratio of sample size 2:1 to detect the OR of 2.59 and 10.15% outcome in the unexposed group based on research carried out in Southwest Ethiopia.⁴² The total sample size was 726, including a design-effect of 2 and 10% of the non-response rate. However, because of the availability of data from our prospective cohort study that was published elsewhere,⁴¹ we took a sample of 1669 participants (1113 unexposed and 556 exposed to alcohol). In the first stage, we selected three health centers using a simple random sampling technique, and we took one hospital purposively. In the second stage, we chose pregnant mothers who fulfilled the inclusion criteria using a systematic random sampling technique. We proportionally allocated the sample size to each health facility based on historical client-flow data of the respective health organization that could enhance the validity and relevance of study findings. By doing this, we ensured that our samples were representative of the population under study.

Study variables

Response variable

The response variable of the study was anemia (yes or no).

Potential confounding variables

Data were collected at the baseline of the cohort study on various risk factors through the pregnancy, comprehensive maternal characteristics and potential confounding variables. Some of the included potential confounding variables were sociodemographic characteristics (maternal age, religion, ethnicity, household wealth status, education of respondent, education of husband, marital status and occupation); obstetrics and some medical factors (parity, history of abortion, antepartum hemorrhage [i.e. any vaginal bleeding after 20 wk gestation], unintended pregnancy, hypertensive disorders in pregnancy⁴³ [chronic hypertension, pre-eclampsia–eclampsia, pre-eclampsia superimposed on chronic hypertension and gestational hypertension], gestational diabetes, mid-upper arm circumference [MUAC] and gender of infant); and behavioral variables (cigarette smoking, coffee intake, khat chewing). Alcohol consumption during pregnancy was the main exposure variable.

Exposure ascertainment and alcohol use measures

We determined prenatal alcohol exposure using maternal self-reporting, which is the most common clinical instrument and standard in detecting.⁴⁴ Detailed information on alcohol intake for specific trimesters was collected via an interview in the ANC clinics. Study participants were interviewed in detail about alcohol use information during pregnancy, including the type and quantity of alcoholic beverages. For each type of alcoholic beverage, pregnant women were asked how often they consumed alcohol and the number of drinks they drank during the specific trimester based on the Alcohol Use Disorders Identification Test—Consumption (AUDIT-C) questionnaire.^45–47

The AUDIT-C questionnaire is the most popular shortened version of the 10-item AUDIT that comprises three items to assess alcohol consumption cross-culturally and identify hazardous drinkers.^48,49 The tool has shown to be a valid instrument for alcohol consumption since pregnancy recognition was based on self-report.⁵⁰ The questionnaire was adjusted by considering the local context of alcoholic beverages of alcohol content and drinking containers. The amount of alcohol content in a standard drink varies from country to country. We used the WHO's standard for this study because Ethiopia has no national alcohol policy defining standard alcohol drinks.⁵¹ Based on this, for a standard drink, 12 g of absolute alcohol was assumed, which was considered as alcohol consumption. A standard drink was determined by converting local drinks to grams of pure alcohol, and then we specified the amount of pure alcohol per local drink and using local units of measurement.

For this study, participants were categorized as abstainers or non-drinkers if women reported that they had never drunk any alcoholic beverages entirely throughout pregnancy (AUDIT-C score=0), as low-risk drinkers or non-hazardous drinkers if they reported scores of 1 or 2 in AUDIT-C for the period of pregnancy)⁵² and as hazardous drinkers if they consumed a pattern or quantity of alcohol with an AUDIT-C score of ≥3.^53–55

Different receptacles were used to measure local drinks, such as ‘tassa’, ‘malekia’ and ‘birille’ for drinks of Tella (traditional Ethiopian beer fermented from mostly barley, but also with wheat, maize and sorghum, and mixed with ‘Gesho’ [Rhamnusprinioides]),⁵⁶  Areke (a whiskey-like drink distilled from fermented barley or maize and mixed with Rhamnusprinioides) and Tej (a honey wine), respectively. The amount of each drink consumed (in mL) was then calculated. This value was converted to grams of absolute alcohol by applying a conversion factor and taking into account the percentage of absolute alcohol present in each drink.

Accordingly, a standard drink equivalent to one bottle of beer (330 mL) at 5% (strength) x 0.79 (conversion factor)=13 g of ethanol; one glass of wine (140 mL) at 12% x 0.79=13.3 g of ethanol; one shot of (malekia) Areki (40 mL) at 40% x 0.79=12.6 g of ethanol, alcoholic content (30–50%); one ‘birille’ Tej (200 mL) at 8% x 0.79=12.64 g of ethanol, alcoholic content (7–11%); and one ‘tassa’ Tella/Korofe (330–500 mL) at 4.5% x 0.79=11.73 g of ethanol, alcoholic content (4–6%).^57–59

Data collection methods and instruments

This data collection tool was similar to previous articles, a nested case-control study and prospective cohort study, which were part of this project published elsewhere.^41,60 The questionnaire was prepared first in English then translated into Amharic (the local language) to suit local applicability and then back to English to ensure its consistency. The tool was developed by reviewing the literature on prenatal alcohol consumption,^{46,47,61–65} and then consultation with experts was sought to ascertain its validity by considering the local situation of the study participants and clinical relevance.

Data were collected using a pretested standardized interviewer-administered questionnaire and by reviewing maternal care logbooks at the health facilities. A detailed interview was conducted for each woman in private with a nurse or a midwife data collector. Data collectors and supervisors were trained on data collection tools, a procedure during data collection, obtaining consent from participants and not missing any questions in the questionnaire. The Amharic version questionnaire was pretested for clarity through a pilot study on 67 respondents in Bahir Dar town, which is 180 km from the actual study area. We preferred Bahir Dar town for pretesting because of its similarity to the study population and other infrastructure related to our study objective and actual study area. The tool was checked for its reliability and validity before actual data collection. To ensure data truthfulness by observing the quality, status and issues in collecting data, we conducted weekly meetings and daily supervision with supervisors and data collectors.

The tool also included the Edinburgh Postnatal Depression Scale, which consists of 10 items, each scored on a scale of 0 to 3; therefore, the total score ranges from 0 to 30, and we used a cut-off point of ≥13 on the scale to identify women with depressive symptoms.⁶⁶ The socioeconomic status of the household wealth index, a composite measure of a household's cumulative living standard, was assessed using 16 variables extracted from the 2016 Ethiopia Demographic and Health Survey, and principal component analysis was computed to determine it. The MUAC of the left arm with no clothing was measured in the third trimester using a flexible non-stretchable standard tape measure. Pregnant women with MUAC<22 cm were considered as undernourished and those with MUAC≥22 cm as normal.⁶⁷ We ascertained a woman as anemic if her hemoglobin level was <11.0 g/dL during the first or third trimester.⁶⁸

Statistical analysis

We entered data into EpiData version 3.1. then exported it to the STATA version 14 software package for analysis. We used a generalized linear model with an identity log and binomial link function (log-binomial) to estimate the risk ratio. All variables significantly associated with outcomes of interest in the bivariable log-binomial regression model at p≤0.2 were entered into a multivariable log-binomial regression model to identify the significant variables, as well as control the possible effect of confounders. The strength of association of the regression model was reported using an adjusted risk ratio (ARR) with a corresponding 95% CI, and p<0.05 was set to declare the statistical significance threshold. The interaction effect of the variables was checked by creating the product term, and then a new variable became either statistically significant or not at p<0.05. We used Akaike information criteria to select the best-fitting model for the data from the list of competing models.

The adjusted population-attributable risk (PAR) assesses the proportion of anemia that would not occur in a population if alcohol consumption were eliminated in the entire cohort after adjustment for potential confounders. The proportion of anemia in pregnancy that could be attributed to maternal pregnancy alcohol consumption was estimated using Levin's formula: PAR (%)=P*(ARR–1)/[P*(ARR–1)+1*100, where P is the proportion of women who used alcohol (hazardous or non-hazardous drinkers) during pregnancy and ARR is the adjusted risk ratio of the adverse birth outcomes associated with the alcohol consumption.^69,70

Ethical considerations

We obtained ethical clearance from the Institutional Ethical Review Board of the University of Gondar (R. No.-O/V/P/RCS/05/747/2019), and permission was received from Amhara Public Health Institute and Gondar Town Health Department before starting the study. Before enrolment of the participants, all respondents were informed about the importance of the study, its objective, effects and the significance of participation. We obtained verbaly informed consent because the study involved no more than minimal risk to the participants and did not adversely affect the rights and welfare of the respondents. This was obtained before conducting data collection, and all the information was completed anonymously to maintain confidentiality. After gathering the necessary information, all participants were counseled about the risks of alcohol drinking during pregnancy and were advised not to drink any alcohol during pregnancy or while trying to get pregnant. Also, women who were engaged in hazardous drinking were referred to healthcare providers, and proper linkage was established to obtain possible treatment options in their respective health facilities.

Results

Sociodemographic characteristics of the respondents

Out of 1778 pregnant women approached for the cohort study, a total of 1669 study participants (non-drinkers [1113], non-hazardous drinkers [112] and hazardous drinkers [444]) had adequate information about their hemoglobin level and they were involved in this study. The mean age of the study participants was 26.43 (±4.59) y. The majority of participants (1471; 88.14%) were orthodox Christians, and 1600 (95.87%) were Amhara by ethnicity. Nearly two-fifths (658) of the respondents' spouses had attained tertiary education level (Table 1).

Table 1.

Sociodemographic characteristics of the study participants attending public health facilities in Gondar town, Northwest Ethiopia, 2020 (n=1669)

	The pattern of alcohol consumption
Variables	Non-drinkers (n=1113)	Non-hazardous (n=444)	Hazardous (n=112)	Total (%)
Age group (y)
15–24	376 (33.78)	141 (31.76)	28 (25.00)	545 (32.65)
25–34	654 (58.76)	277 (62.39)	74 (66.07)	1005 (60.22)
≥35	83 (7.46)	26 (5.86)	10 (8.93)	119 (7.13)
Marital status
Married	1085 (97.48)	430 (96.85)	110 (98.21)	1625 (97.36)
Single/divorced/separated/widowed	28 (2.52)	14 (3.15)	2 (1.79)	44 (2.64)
Family size
1–2	416 (37.38)	179 (40.32)	45 (40.18)	640 (38.35)
3–4	567 (50.94)	217 (48.87)	58 (51.79)	842 (50.45)
≥5	130 (11.68)	48 (10.81)	9 (8.04)	187 (11.20)
Education level of respondents
No formal education	231 (20.75)	100 (22.52)	30 (26.79)	361 (21.63)
Primary education (grade 1–8)	164 (14.73)	63 (14.19)	17 (15.18)	244 (14.62)
Secondary education (grade 9–12)	382 (34.32)	140 (31.53)	36 (31.14)	558 (33.43)
Tertiary education (above grade 12)	336 (30.19)	141 (31.76)	29 (25.89)	506 (30.32)
Occupation
Housewife	521 (46.81)	187 (42.12)	63 (55.17)	771 (46.20)
Merchant	246 (22.10)	121 (27.25)	13 (11.61)	380 (22.77)
Employed in any organization	232 (20.84)	115 (25.90)	30 (26.79)	377 (22.59)
Others	114 (10.24)	21 (4.73)	6 (5.36)	141 (8.45)
Household wealth index
Poorest	258 (23.18)	80 (18.02)	9 (8.04)	347 (20.79)
Poor	225 (20.22)	84 (18.92)	32 (28.57)	341 (20.43)
Middle	196 (17.61)	97 (21.85)	19 (16.96)	312(18.69)
Rich	223 (20.04)	89 (20.05)	24 (21.43)	336(20.13)
Richest	211 (18.96)	94 (21.17)	28 (25.00)	333(19.95)

Obstetrics and health-related characteristics of the respondents

Almost two-thirds, 1006 (60.28%), of the study participants were multiparous. The majority (1450; 86.88%) of pregnant women had experienced a planned pregnancy pattern. Forty-eight (4.77%) of the respondents had experienced a history of preterm birth. Among the study participants who were screened for antenatal depression (a mood disorder experienced by childbearing women during pregnancy), 308 (18.45%) pregnant women had depression. Only 19 (1.14%) of the study participants had a known history of diabetes mellitus. Approximately one-quarter (24.21%) of the pregnant women had received advice on the risk of alcohol consumption throughout ANC visits (Table 2).

Table 2.

Obstetric and health-related characteristics of the respondents attending ANC at public health facilities in Gondar town, Northwest Ethiopia, (n=1669)

	The pattern of alcohol consumption
Variables	Non-drinkers (n =1113)	Non-hazardous (n =444)	Hazardous (n=112)	Total (%)
Birth interval (n=992)
<24 mo	144 (21.24)	27 (10.93)	8 (11.94)	179 (18.04)
≥24 mo	534 (78.76)	220 (89.7)	59 (88.06)	813 (81.96)
Number of children
No child yet	445 (39.98)	188 (42.34)	44 (38.29)	677 (40.56)
1–2 children	536 (48.16)	220 (49.55)	60 (53.57)	816 (48.89)
≥3	132 (11.86)	36 (8.11)	8 (7.14)	176 (10.55)
History of abortion
Yes	124 (11.14)	65 (14.64)	16 (14.29)	205 (12.28)
No	989 (88.86)	379 (85.36)	96 (85.71)	1464 (87.72)
MUAC
<22 cm	142 (12.76)	83 (18.69)	25 (22.32)	250 (14.98)
≥22 cm	971 (87.24)	361 (81.31)	87 (77.68)	1419 (85.02)
Anemia
Yes	147 (13.21)	62 (13.96)	39 (34.82)	248 (14.86)
No	966 (86.79)	382 (86.04)	73 (65.18)	1421 (85.14)
Hypertensive disorders in pregnancy
Yes	87 (7.82)	53 (11.94)	19 (16.96)	159 (9.53)
No	1026 (92.18)	391 (88.06)	93 (83.04)	1510 (90.47)
Drinking coffee
Yes	833 (74.84)	327 (73.65)	82 (73.21)	1242 (74.42)
No	280 (25.16)	117 (26.79)	30 (26.79)	427 (25.58)
Smoking
Yes	2 (0.18)	2 (0.45)	2 (1.79)	6 (0.36)
No	1111 (99.82)	442 (99.55)	110 (98.21)	1663 (99.64)
Khat chewing
Yes	15 (1.35)	8 (1.80)	2 (1.79)	25 (1.50)
No	1098 (98.65)	436 (98.20)	110 (98.21)	1644 (98.50)
Antepartum hemorrhage
Yes	26 (2.34)	11 (2.48)	7 (6.25)	44 (2.64)
No	1087 (97.66)	433 (97.52)	105 (93.75)	1625 (97.36)

ANC, antenatal care; MUAC: mid-upper arm circumference.

Alcohol consumption during pregnancy

Summarized details of alcohol consumption during pregnancy are provided in Table 3. Out of 556 pregnant women who were exposed to alcohol, 112 (20.14%) drank alcohol hazardously. Tella was used by the majority of pregnant women (41.61%), followed by beer/draft (32.43%). The majority of study participants (343; 61.69%) drank alcohol once a month or less frequently. One hundred and forty-eight (26.62%) respondents consumed alcohol 2–4 times per month, while 63 (11.33%) pregnant women drank 2–3 times per week. On the other hand, 29 (5.22%) pregnant women admitted to drinking ≥6 drinks on one occasion during their current pregnancy.

Table 3.

Varieties of alcoholic beverages consumed by pregnant women attending ANC at public health facilities in Gondar town, Northwest Ethiopia, (n=1669)

	First trimester	Second trimester	Third trimester	Total
Types of alcohol used	Number (%)	Number (%)	Number (%)	%
Tella	312 (58.9)	178 (32.96)	175 (32.96)	41.61
Areke	12 (2.26)	7 (1.30)	3 (0.56)	1.37
Tej	2 (0.38)	3 (0.56)	2 (0.38)	0.44
Korofe	5 (0.94)	3 (0.56)	3 (0.56)	0.69
Beer/draft	106 (20)	218 (40.37)	196 (36.91)	32.43
Wine	11 (2.08)	20 (3.70)	19 (3.58)	3.12
≥2 types	82 (15.47)	111 (20.56)	133 (25.05)	20.36

ANC, antenatal care.

The association of alcohol consumption during pregnancy with anemia

The study showed that 14.86% (95% CI 13.23 to 16.65%) of study participants were affected by anemia. The study also indicated that 147 (13.21%) non-drinkers, 62 (13.96%) non-hazardous drinkers and 39 (34.82%) hazardous drinkers (χ²=37.96, p<0.001) had anemia.

The study revealed that women who consumed hazardous levels of alcohol during pregnancy had a 2.24 times (ARR=2.24; 95% CI 1.60 to 3.15) increased risk of anemia compared with women who abstained from consuming alcohol throughout pregnancy. On the other hand, mothers who were non-hazardous alcohol drinkers had no significant association with anemia during pregnancy (ARR=1.03; 95% CI 0.79 to 1.36). Participants who had primary level education had a decreased risk of anemia (ARR=0.54; 95% CI 0.36 to 0.81) relative to those who had tertiary level education. The risk of having anemia in pregnancy was 1.34 times (ARR=1.34; 95% CI 1.01 to 1.78) higher in participants who had unplanned pregnancy status compared with those who had planned pregnancy status. Similarly, women who used to consume coffee had an increased risk of anemia (ARR=1.41; 95% CI 1.06 to 1.88) relative to those who did not consume coffee. Finally, study participants who had a history of antepartum hemorrhage had a 2.68 times (ARR=2.68; 95% CI 1.84 to 3.92) greater risk of having anemia compared with their counterparts (Table 4). The adjusted PAR of anemia in pregnancy related to hazardous alcohol consumption during pregnancy was 7.68%.

Table 4.

Relationship between maternal alcohol consumption during pregnancy and anemia in pregnancy at public health facilities in Gondar town, Northwest Ethiopia, 2020

	Anemia
Variable	Yes	No	CRR (95% CI)	ARR (95% CI)
Age of the mother (y)
15–24	79 (14.50)	466 (85.85)	1	1
25–34	142 (14.13)	863 (85.87)	0.97 (0.76, 1.26)	0.85 (0.65, 1.10)
≥35	27 (22.69)	92 (77.31)	1.57 (1.06, 2.31)	1.24 (0.83, 1.84)
Education level
No formal education	59 (16.34)	302 (83.66)	0.94 (0.70, 1.27)	0.82 (0.60, 1.12)
Primary education	27 (11.07)	217 (88.93)	0.64 (0.43, 0.95)	0.54 (0.36, 0.81)
Secondary education	74 (13.26)	484 (86.74)	0.76 (0.06, 1.01)	0.76 (0.57, 1.02)
Tertiary education	88 (17.39)	418 (82.61)	1	1
Household wealth status
Poorest	39 (11.26)	308 (88.76)	1	1
Poor	58 (17.01)	283 (82.99)	1.51 (1.04, 2.21)	1.26 (0.86, 1.84)
Middle	49 (15.71)	263 (84.29)	1.40 (0.94, 2.07)	1.22 (0.83, 1.80)
Rich	58 (17.26)	278 (82.74)	1.54 (1.05, 2.24)	1.26 (0.86, 1.85)
Richest	44 (13.21)	289 (86.79)	1.18 (0.78, 1.76)	1.05 (0.69, 1.60)
Status of pregnancy
Planned	199 (13.72)	1251 (86.28)	1	1
Unplanned	49 (22.37)	170 (77.63)	1.63 (1.23,2.15)	1.34 (1.01, 1.78)
Coffee consumption
Yes	197 (15.86)	1045 (84.14)	1.33 (0.99, 1.77)	1.41 (1.06, 1.88)
No	51 (11.94)	376 (88.06)	1	1
Alcohol consumption
Hazardous	39 (34.82)	73 (65.18)	2.64 (1.96, 3.54)	2.24 (1.60, 3.15)
Non-hazardous	62 (13.96)	382 (86.04)	1.06 (0.80, 1.39)	1.03 (0.79, 1.36)
Non-drinker	147 (13.21)	966 (86.79)	1	1
Depression
Yes	58 (18.83)	250 (81.17)	1.35 (1.03, 1.76)	0.86 (0.66, 1.13)
No	190 (13.96)	1171 (86.04)	1	1
Prepregnancy alcohol use
Yes	143 (16.34)	732 (83.66)	1.24 (0.98, 1.56)	0.93 (0.74, 1.18)
No	105 (13.22)	689 (86.78)	1	1
Antepartum hemorrhage
Yes	18 (40.91)	26 (59.09)	2.89 (1.99, 4.20)	2.68 (1.84, 3.92)
No	230 (14.15)	1395 (85.85)	1	1

Bold values are the strength of association the variable with anemia (outcome variable)

ARR: adjusted risk ratio; CRR: Crude Relative Risk.

Discussion

To the best of our knowledge, this study is the first to retrospectively link prenatal alcohol exposure to anemia in pregnant mothers in Ethiopia. The overall proportion of pregnant women who were with anemia in both exposed and unexposed was 14.86%. Focusing on the level of alcohol consumption, the study revealed that there was no significant difference in anemia between pregnant women who were categorized as non-drinkers and non-hazardous drinkers. On the contrary, there were considerable differences in anemia proportions between abstainers (13.21%) and hazardous (34.82%) drinkers. Factors that were identified as risk factors for anemia in pregnancy were the educational level of the respondents, the status of pregnancy (pregnancy planning), antepartum hemorrhage, coffee consumption and hazardous alcohol use. We found no remarkable differences in non-hazardous alcohol consumption during pregnancy compared with abstainers for prenatal anemia. However, hazardous maternal drinking was significantly associated with anemia. These findings demonstrated that there was dose-dependency between alcohol use during pregnancy and anemia in pregnancy.

The evidence from the literature in different regions and parts of Ethiopia reported that the prevalence of anemia during pregnancy ranged from 9.7 to 56.8%.^71,72 Our finding was also within the range of previous studies' reports. The current study result could indirectly imply that the risk of several maternal and fetal complications was significant in the study area because anemia contributes about one-quarter of indirect causes of maternal deaths in developing countries⁷³ and is positively associated with low birth weight, preterm birth and other negative impacts.^74,75 The high magnitude of anemia might be linked to sociodemographic and economic characteristics (age, level of education, marital status, residence and nutritional status),^42,76 obstetric (gravidity, parity, pregnancy planning, history of excessive menstrual bleeding),^77,78 medical conditions (parasitic infections)^79,80 and behavioral factors (smoking, coffee/tea drinking, alcohol drinking, utilization of prenatal care services).^81–83 The risk factors of anemia and other adverse pregnancy outcomes could be due to a lack of adequate counseling during preconception and antenatal visits.⁸⁴ For instance, about 61% of pregnant women were not informed about the harmful effect of alcohol consumption during pregnancy among Ethiopian women.³⁶ In this study, 76% of pregnant women lack access to health information regarding the negative outcomes of drinking alcohol during pregnancy.

After adjustment for potential confounding factors including age, educational level of the respondents, status of pregnancy, antepartum hemorrhage, prepregnancy alcohol use, depression and household wealth status, hazardous alcohol drinking (AUDIT-C score≥3) during pregnancy presented a greater than twofold higher risk of anemia in pregnancy, whereas non-hazardous drinking (AUDIT-C score 1–2) tended not to be associated with anemia in pregnancy compared with abstainers.

This finding confirmed that alcohol consumption during pregnancy was an important risk factor for developing anemia with a dose–response relationship. This result is in agreement with previous studies carried out in different areas,^42,85 but studies conducted in Ethiopia and Russia had insufficient evidence for the statistical association of anemia with alcohol consumption during pregnancy.^86,87 The comparisons between the findings of the previous studies and our result may be complicated because there were methodological variations in measurements of alcohol consumption and the categories of alcohol drinking (i.e. in previous reports, these were drinking or not drinking, but in the current study, they were non-drinking, non-hazardous drinking and hazardous drinking). However, the mechanism behind the link between hazardous alcohol drinking and anemia during pregnancy could be that alcohol can be directly toxic to the bone marrow; a proportion of alcoholics have both reduced red blood cell production in the bone marrow and accompanying reduced erythrocyte survival time,⁸⁸ and acute alcohol consumption results in the mild breakdown of red blood cells.⁸⁹ Alternatively, acute and chronic alcohol consumption may impair intestinal absorption and utilization of various nutrients, including iron, which causes a deficiency of folate, and is also attributable to reduced hepatic uptake and increased excretion of nutrients in stool.^90,91

Finally, the adjusted PAR takes into consideration which measures the excess risk in the community associated with hazardous alcohol consumption compared with the non-exposed group. It showed that 7.68% of the anemia in pregnancy could be prevented if hazardous alcohol consumption were eliminated in the entire cohort population. The findings of this study could be generalizable to similar settings, provided they are interpreted with caution and take the selection criteria of the study participants into account.

Strengths and limitations of the study

Inclusion of local beverages by measuring local containers is the strength of this study. However, it has the following limitations. Prenatal alcohol exposure relied on participants' self-reporting of alcohol use, which might have masked the exact exposure and dose of alcohol due to a loss of recall. Participants might have falsely denied their use of alcohol for reasons of social desirability, and this could have resulted in a misclassification of exposure and the association of alcohol and anemia. Because most of the alcoholic beverages used were locally homemade brews, the exact volume held by containers and hence the amount in drinks might not have been well understood by the respondents, therefore it was difficult to obtain the exact standard drink during conversion. Finally, this study might not address all potential confounding factors that may have affected the findings, thus interpretation of the results needs special attention.

Conclusions

This study revealed that anemia during pregnancy continues to be a public health concern in Northwest Ethiopia. Regardless of timing, the clear association between prenatal hazardous alcohol consumption and increased risk of anemia during pregnancy suggests a need for targeted prenatal alcohol use and anemia screening, and ongoing intervention for pregnant women who are alcoholics and have anemia. Further, the authors recommend ongoing counseling and education about the harmful effects of alcohol consumption and anemia for pregnant women and women of reproductive age in general. Furthermore, although this study found that anemia during pregnancy is likely to vary depending on the quantity of alcohol consumed, further future studies shall be conducted to uncover the timing when the effect of drinking is more harmful to enable targeted action.

Authors’ contributions

AEA conceptualized, designed, coordinated data collection, analyzed the data and drafted the manuscript. TA and MY designed the study and critically revised the manuscript. All the authors read and approved the final manuscript.

Acknowledgements

We thank the Gondar town health department, University of Gondar Comprehensive Specialized Hospital and health centers for their provision of necessary information and support during data collection. Our genuine thanks also go to the study participants, data collectors and supervisors who took part in the study.

Funding

This study was funded by the University of Gondar and Wachemo University. The universities are following whether findings are presented and published. The universities have no role in the design, data collection, analysis and interpretation of the findings and in writing the manuscript. All the statements and findings are the responsibility of the investigators.

Competing interests

The authors declare that they have no competing interests.

Ethical approval

Ethical approval was received from the University of Gondar Institutional Ethical Review Board (R. No.-O/V/P/RCS/05/747/2019) and a permission letter was also obtained from the Gondar town health department. All the study participants were aged >18 y. They were notified about the objective and verbally informed consent was obtained before conducting data collection. They were also informed that they had the full right to withdraw or refuse to participate in the study. No financial incentive was given to participants for their participation in the study. Data obtained from study participants were held anonymously and confidentially.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.