Study Area and Period
A health facility-based cross-sectional study was conducted from January 12 to February 12, 2017 in Dessie Town located 401 Kms far from the capital city, Addis Ababa and 480 kms from the regional capital city, Bahir Dar, Ethiopia. Total population in Dessie Town was 216,384, among which 7,292 were expected to be pregnant women. There are 12,17and 4 governmental, private and nongovernmental health facilities (total of 33 health facilities) in Dessie Town respectively. Among the total pregnant women, 4397 pregnant women have got ANC in 2016 in the governmental, private and nongovernmental health facilities of the Town.
Maternal and child health care services are given in most of government, some private and nongovernmental health facilities like immunization, family planning, antenatal care, delivery care and postnatal care. In Dessie Town available common food products include cereals, grains, legumes, dairy products, egg, fish, meat, fruits (Mango, papaya, Avocado, Orange) and vegetables (Spinach, Lettuce, Green pepper, carrot, pumpkin, cabbage).
Study population were, all pregnant women who were attending ANC in 12 randomly selected health facilities (government, private and nongovernmental health facilities) of Dessie town. All pregnant women who were attending in the randomly selected government, private and nongovernmental health facilities of Dessie Town were included. Pregnant women found in a fasting day and have a dietary restriction were excluded from the study.
Sample size and Sampling Technique
The sample size was determined using single proportion population formula by considering the following assumptions of proportion, 50% of adequacy of vitamin A, confidence level of 95%, margin of error 5% and 10% as non-response rate, the total sample size was 422.
InDessie Town, 17 health facilities provide ANC service, among which12 health facilities were randomly selected (07 governmental, 04 private and 01 non-governmental). The sample size was allocated proportionally to the client flow of each health facility. Those pregnant women who fulfilled the inclusion criteria and visited the ANC clinic during the study period were systematically selected as study participants after obtaining an identification card of each pregnant woman
- Data collection Techniques and Data Measurements
An Interviewer-administered pre-tested questionnaire was used for dietary data from known sources (7, 16–18). The questionnaire was prepared by local language (Amharic) and then their response translated to English for the purpose of this study. Eight data collectors and two supervisors were hired and participated based on their prior exposure of research data collection.
The socio-demographic features (such as age, educational level, occupation, marital status, income level, family size), obstetrics related characteristics (birth interval, parity, months of pregnancy) and dietary intake and habits of pregnant women (food taboos, nutrition education, daily intake of vitamin A rich sources, reasons for not daily taking of vitamin A rich vegetables, poultry production, home gardening of vegetables) were assessed to investigate the research objective.
Adequacy of vitamin A and Women dietary diversity score (WDDS) were measured by using the nine food groups from FAO dietary diversity guideline based on interactive 24 hours dietary recall. Details of the food groups were probed or interviewed from each participant. Dietary information was assembled and estimated using portion size and gram amounts to know the Carotenoids equivalent and retinol equivalent (RE). And then to determine the level of adequacy, individual RE from each food group was summed. The corresponding nutrient adequacy ratio (NAR) was calculated. As recommended by joint FAO/WHO consultation, mean requirement of vitamin A in mcg RE/day for pregnant women is 370 and safe intake in mcg is 600 RE/day. Adequacy of vitamin A was defined as a daily intake of the vitamin from plant and/or animal sources that is equal to the estimated average requirement of vitamin A, 370mcg RE/day.
So the adequacy of Vitamin A categorized using a cutoff value of 1 as adequate and <1 as inadequate (12, 17).
The WDDS was classified as low, medium and high, according to the FAO’s classification of WDDS as “low” if WDDS <3, as “medium” if WDDS ranged 4–5 and as “high” if WDDS ≥6 (17).
Dietary intake of vitamin A of each study participant was converted to gram amounts then gram amounts calculated to Carotenoids equivalent as per 100 gram edible portion in the food composition table of Ethiopia.
Carotenoids equivalents were converted to retinol equivalent (RE) by a conversion factor of six. Foods that have retinol values were simply recorded and summed to values obtained by conversion of the Carotenoids to retinol (RE). To determine nutrient adequacy ratio (NAR) of vitamin A, the ratio of observed intake was calculated in relation to recommended intake based on estimated average requirement (EAR) of 370 mcg RE/day (12).
Training was given for data collectors on the objectives of the study and items of the data collection tools. Information about the essential technical skills on how to collect a 24 hour dietary recall method was also part of the training. Frequent supervision and checking of the data for its consistency and completeness was done. To reduce recall bias on dietary data, corresponding portion size of dietary intake was estimated by local measurement utensils by allowing study participants to choose from a variety of commonly used local measurement utensils offered during data collection and show which variety of local measurements were used for the portion size of their diet. Gram amounts for the portion sizes were determined as per the food composition table of Ethiopia (per 100 gram edible portion). Some consumed foods were estimated by their raw portion size or gram amounts.
Data management and analysis
Data entry was done using Epi-Info 7.2 software program and data cleaning was also done by using SPSS version 20 software program. The food composition table of Ethiopia version III and IV were used to convert raw dietary data to Carotenoids equivalent and retinol equivalent of vitamin A rich foods consumed in the past 24 hours.
Descriptive statistics such as percentage, mean and standard deviation were done to express the findings. After dichotomizing the dependent variable, binary logistic regression and multiple logistic regression analysis were done to see the association of independent variables on the outcome variable and to control the possible confounding effects. The Odds ratio was used to assess the presence and degree of association between dependent and independent variables. A P-value less than 0.05 used to decide whether the differences that occur would be statistically significant or not.