Patterns and predictors of gestational weight gain in Addis Ababa, Central Ethiopia: a prospective cohort study

Introduction: Gaining excessive or inadequate gestational weight is associated with many adverse maternal and fetal outcomes. Inadequate gestational weight gain (GWG) increases the risk of fetal growth restriction, pre-term birth, and low birth weight. It is a public health concern in sub-Saharan Africa. The aim of this study was to assess the patterns and predictors of GWG in Addis Ababa, Ethiopia. Methods: We conducted a prospective cohort study among pregnant women who attended antenatal care in health centres in Addis Ababa, from January to September 2019. Data were collected by a structured questionnaire and checklists and analysed using Stata version-14. Weight at or before 16 weeks gestation was used as a proxy for pre-pregnancy weight. Women’s height and baseline weight were measured by data collectors, and we obtained weight at the end of the 24th and 36th weeks of gestation from women’s medical records. GWG was categorized as inadequate, adequate and excessive based on the United States Institute of Medicine criteria. Predictors of GWG were identified using multinomial logistic regression. Results: A total of 395 pregnant women were enrolled in the study. GWG was assessed for 369 (93%) women. The median GWG was 8.7 kg with inter quartile ranges (25th, 75th percentiles) of 7.0 kg and 11.6 kg. More than two-third of the participants, 248 (67.2% [95% CI: 62.2, 72.0%]), gained inadequate weight; 103 (27.9% [95% CI: 23.4, 32.8%]) gained adequate weight; and 18 (4.9% [95% CI: 2.9%, 7.6%]) gained excessive weight. Three quarters (75%) of underweight women gained inadequate gestational weight, whereas 43% of overweight or obese women gained inadequate gestational weight. Being underweight (AOR = 3.30 [95% CI: 1.32, 8.24]) or normal weight (AOR = 2.68 [95% CI: 1.37, 5.24]) before pregnancy increased the odds of gaining inadequate gestational weight compared to overweight or obese women. Not having paid employment was associated with higher odds of gaining inadequate gestational weight compared to women employed outside the home (AOR = 2.17 [95% CI: 1.16, 4.07]). Conclusions: Most pregnant women in Addis Ababa gain inadequate gestational weight. In particular, three quarters of underweight women gained inadequate gestational weight. Being underweight, normal weight or having no paid employment were associated with higher odds of inadequate GWG. Promoting adequate GWG in Addis Ababa among underweight and normal weight women may be an important public health initiative. This study was conducted in Addis Ababa, which is the capital and largest city in Ethiopia. In the city, there are 42 hospitals (11 government, 6 non-government organisations, and 25 private), 97 Health Centres, and 361 clinics that provide medical care including maternal health care [45, 46]. Around 97% of pregnant women in Addis Ababa receive antenatal care (ANC) from skilled care providers such as doctors, nurses or midwives, at least once [42], of which 90% receive at least four ANC contacts [47]. A prospective cohort study design was employed from January 2019 to September 2019. We calculated the sample size using Open Epi Version 2.3 considering both the single proportion formula (to assess the proportion of GWG) and the double proportion formula (to assess predictors of GWG). The larger sample size was achieved by using the single proportion formula considering the proportion of women with inadequate gestational weight from a study conducted in Harar, Ethiopia (p = 0.69) [5], a half-width of confidence 5%, an alpha value of 0.05, and 20% loss to follow-up. The final sample size was 395. The women were selected from nine health centres. The health centres were selected based on the number of ANC visits and geographic location in the city. Women who met the inclusion criteria were consecutively selected from each health facility until the required sample size was met. We invited all pregnant women in their first trimester (before 16 weeks gestation) who came to the selected health centres for antenatal care (Additional file 1: Table S1). Antenatal clinic staff who provided the antenatal care facilitated the participant selection process. They also assisted in setting appointment dates for the follow-up data collection (at the end of the 24th and 36th weeks of gestation). Women with a twin-pregnancy or with co-morbidities such as diabetes and hypertension were excluded from the study. We collected data through face-to-face interviews and review of health records. We used a range of tools to collect data on socio-demographic characteristics, dietary diversity and food security, intimate partner violence, physical activity and depression related symptoms. Variables such as gestational age (ultrasound result), blood pressure, random blood sugar, anaemia status, and HIV status were obtained from medical records of the women. Principal component analysis was employed to compute a wealth index [48] from a set of household assets questions such as electricity, refrigerator, table, chair, watch, phone, bed with mattress, electric mitad (an Ethiopian oven made up of clay and metal), car, house, improved water, and improved toilet, which were adapted from the Ethiopian demographic and health survey [42]. Gestational age was estimated by the last menstrual period and verified by ultrasound which was a routine practice of the health facilities. The gestational age of our study participants ranged from four to 16 weeks (8.9% were between four and seven weeks of gestation; 41.1% were between eight and 12 weeks of gestation; and 50% were between 13 and 16 weeks of gestation). The height of the women was measured when barefoot using a height measuring board in a standing position and recorded to the nearest 0.1 cm. The maternal weight was measured by a digital weight scale with minimum clothing and the reading was recorded to the nearest 100 g. We asked women if they knew their pre-pregnancy weight, however only 172 (43.5%) of the participants were aware of their pre-pregnancy weight. Therefore, we used weight at or before 16 weeks as a proxy for pre-pregnancy weight in all women. Women’s height and baseline weight were measured by data collectors, while weight at the end of 24th and 36th weeks of gestation was collected from women’s medical records. Body Mass Index (BMI) was calculated by dividing weight by height, squared. The women’s BMI at or before 16 weeks of gestation (for those whose ages were ≥ 20 years old) was categorized into four categories based on the World Health Organization BMI cut-off points as underweight (BMI ≤ 18.5 kg/m2); normal weight (18.6 to 24.9 kg/m2); overweight (25.0 to 29.9 kg/m2); and obese (≥ 30.0 kg/m2). BMI-for-age (at or before 16 weeks of gestation) was calculated for adolescent women (women aged 18 and 19 years old); and BMI was categorized using WHO reference cut-off points as thin (Z-score < -2 standard deviation (SD)), normal (-2 SD ≤ Z-score ≤  + 1SD), overweight (+ 1SD < Z-score ≤  + 2SD) and obese (Z-score >  + 2SD). Total weight gain was calculated by subtracting the pre-pregnancy weight from their weight at the 4th antenatal care visit (at the end of 36 weeks of gestation). It was categorized as inadequate, adequate and excessive according to the IOM classification. Mid upper arm circumference (MUAC) was measured using an adult MUAC non-stretchable measuring tape and the reading was taken to the nearest 0.1 cm. A MUAC measurement below 23 cm was categorised as low (or wasting) and above 23 cm was categorized as normal. Dietary diversity of the women was assessed using a minimum dietary diversity-women (MDD-W) set from the Food and Agricultural Organisation (FAO) and USAID’s Food and Nutrition Technical Assistance III Project (FANTA) [49]. The food groups assessed in MDD-W include: grains, white roots, tubers and plantains; pulses; nuts and seeds; dairy; meat, poultry and fish; eggs; vegetables; other vitamin A-rich fruits and vegetables; other vegetables; and other fruits. The MDD-W is a dichotomous indicator of whether or not women have consumed at least five out of ten defined food groups the previous day or night. The proportion of women who reach this minimum can be used as a proxy indicator for higher micronutrient adequacy. Household food insecurity was assessed using the Household Food Insecurity Access Scale (HFIAS). In each domain of the HFIAS questions ask about anxiety and uncertainty; insufficient quality; and insufficient food intake and any physical consequences, with a recall period of four weeks (30 days)[50]. Women’s physical activity level was measured using the International Physical Activity Questionnaire (IPAQ-long form). The IPAQ assesses physical activity across a range of different domains including recreation-time, housework, being employed and transportation related physical activities. Each domain assesses walking, moderate and vigorous physical activities over a seven day period. Women were asked if they had completed these activities continuously for at least 10 min. Responses to IPAQ questions on the frequency and duration of physical activity were converted to the metabolic equivalent task per minute (MET-minutes) [51]. A MET is the ratio of specific physical activity metabolic rates to the resting metabolic rate, with one MET defined as the energy needed by an individual while at complete rest, which is equivalent to l kilocalorie per kilogram per hour [52]. The level of physical activity for each woman was categorized as; Perinatal depression symptoms were measured using the Edinburgh postnatal depression scale (EPDS) [53], which is a ten-item questionnaire. It has been validated and used by many studies for detecting perinatal depression in Ethiopia [54–57]. Intimate partner violence was measured with a questionnaire used by the WHO multi-country study on women’s health and domestic violence [58]. It includes physical violence, sexual violence and emotional abuse by intimate partners. This questionnaire has also been used in the Ethiopian Demographic and Health Survey (EDHS) 2016 [42], making the survey suitable to use in the current study setting. Data were entered into CSPro version 7.1, and exported to STATA (V.14, Stata Corp, 2015) for analysis. Frequencies and proportions were estimated to describe the variables. BMI-for-age was calculated for adolescent pregnant women using WHO AnthroPlus software. We conducted bivariable and multivariable analyses using a multinomial logistic regression model, because the outcome variable (i.e., GWG) consisted of three categories (inadequate, adequate and excessive GWG). Pregnant women with inadequate or excessive GWG were compared to women with adequate GWG (reference category). Variables with P-value < 0.25 in the bivariate analysis were included in the multivariable analyses. The variables in the multivariable analyses included educational status, wealth index, occupational status, BMI, MUAC, perinatal depression and asking permission from partners for health care seeking. Crude odds ratios (COR) and adjusted odds ratios (AOR) were calculated to determine the association between the explanatory variables and GWG.