Effect of behavior change communication through health development army on optimal nutrition and health practice of pregnant women in Ambo district, Ethiopia: a cluster randomized controlled community trial

Background: Pregnancy is such a critical phase in a woman's life, when mother's nutritional and health status at conception and throughout pregnancy plays a key role in determining her health and well-being, as well as that of her child. Therefore, the objective of the study was to investigate the effect of Behavior Change and Communication (BCC) through women development Army on optimal nutrition and health practices of pregnant women. Methods: A cluster randomized controlled community trial (CRCCT) with baseline and end line measurements using quantitative data collection methods was conducted in Ambo district among 750 pregnant women aged 18-49 years (375 of control and 375 of intervention groups). Of which endpoint data were collected from 372 and 372 pregnant women in the intervention and control clusters, respectively. Pretested semi structured questionnaire was used to collect data. The intervention was started in July, 2018 and data collection for end line was started in October, 2018. Descriptive statistics, chi-square test, bivariate and lastly multivariate binary logistic regressions analysis was used to control confounders. Statistical analyses were conducted primarily on intention-to-treat (ITT) basis and p value <0.05 was used to declare statistical signicance. Result: The overall optimal nutrition and health practice among intervention group were 62.6%, while among control group were 39.0% (p<0.0001). Being in the intervention group, educational status, estimated time to reach health institution, knowledge about nutrition and health were found signicantly associated with optimal nutrition and health practices in multivariable binary logistic regression analysis (p<0.05). Conclusions: Behavior change communication (BCC) through women development army is effective in improving optimal nutrition and health practices of pregnant women. Therefore, to improve optimal nutrition and health practices of pregnant women, BCC through health development Army is recommended. Trial on of on among in


Background
Nutrition throughout life has a major effect on health. Mother's nutritional and health status at conception and throughout pregnancy plays a key role in determining her health and well-being, as well as that of her child (1). According to Essential Nutrition Actions (ENA) framework, optimal nutrition and health practice of pregnant women encompasses optimal quantity and quality of diet; essential micronutrient intake; diseases prevention and treatment practices and supportive life style and care (2)(3)(4)(5).
Behavior change communication (BCC) is widely recognized as one of the main health promotion strategies. It is an interactive process of working with individuals and communities to develop communication strategies to promote positive behaviors, as well as create a supportive environment to enable them to adopt and sustain positive behaviors (6,7).
Suboptimal diet that comprises inadequate intake of calories and other nutrients, combined with a heavy maternal workload, impacts adversely on the health of the mother, the developing foetus and the newborn (8). Pregnant women need at least one extra meal a day than usual, along with plenty of safe water. Consumption of iron-rich foods and iodized salt is also important (9) Naturally, the urge to eat more is experienced by nearly all pregnant women (10). Weight gain is a normal process during pregnancy, in average a woman should gain about 12.5 kg during pregnancy. Many women gain barely half this amount because of poor diets and heavy workloads (4,9).
Malaria prevention and treatment, deworming treatment, sanitation facilities and practices are very important aspects of nutrition and health practices during pregnancy (2,4). Minimizing heavy work load and reducing work hours enables energy-de cient women to reduce her energy expenditure. Nutrients are mobilized from maternal stores to produce breast milk and fetal nutrition, and pregnant woman is vulnerable to depletion of her nutrient stores if she does not decrease work load or not get enough rest during pregnancy (3,5).
Women's status and women's education play pivotal roles in improving nutrition. In addition to bene ting their nutrition and health status, greater education among women paves the way to greater productivity, empowerment, and control of resources allowing them to make better choices that ultimately bene t the health and nutrition of children and families (11,12). In some cultures, because of their status, women are less able to access resources and make decisions to improve their health and nutrition (13).
Studies were done in Ethiopia on dietary knowledge and practice and nutritional status of pregnant women. According to the institution based study done in Mettu Karl Hospital, Southwest Ethiopia, among pregnant women the prevalence of sub-optimal dietary practices was 22% (14). According to the study done in Gedeo zone, southern Ethiopia, in 2018 around one-third (67.8%) of pregnant women had poor dietary practices (15). Different researcher's shows socio demographic factors like Age, wealth index, residence, Size of farmland and illiteracy or low education are common factors affecting maternal nutrition and health status and similarly maternal and health service related factors like years at marriage, ANC visits, level of nutritional knowledge and food practices (consume additional food during pregnancy and variety of food from both animal and plant origins are common factors affecting nutrition and health practices of pregnant women (16)(17)(18)(19).
This paper aims to apply the Health development Army (HDA) as a pivotal role of implementer on improving optimal nutrition and health practices of pregnant women based on ENA framework. To the best of our knowledge, there is no study done on the effect of BCC through women development army on optimal nutrition and health practices among pregnant women based on ENA framework. Thus, this study was aimed to investigate the effect of BCC through women development army on optimal nutrition and health practices among pregnant women in the study area.

Methods And Materials
Study design, study period and setting A cluster randomized controlled community trial (CRCCT) with baseline and end line measurement was conducted from June 2018 to October 2018 among pregnant women in Ambo district of West Shoa Zone, Ethiopia. Ambo district is located in western part of Ethiopia 114 km from Addis Ababa, the capital of Ethiopia. Based on 2017 district health o ce, it has 37,454 and 6976 reproductive age group and pregnant women respectively(20).

Sample size determination and study population
The sample size was calculated using G power 3.1.9.2 program with a power of 80% for Fisher's exact test and precision of 5%. According to Fekadu B et al., (2016) the proportion of dietary practice among pregnant women (p1) 0.34% was used (21), effect size (h) of 0.3 and with the allocation ratio case to control (N2/N1) of 1, Proportion (p2) 0.65 was obtained. The calculated sample size was multiplied by design effect of two due to cluster sampling. the intervention or control arms). The unit of randomization was cluster and both intervention and control groups were available in the same kebele. Then the sample size was allocated based on proportional to size (PS) sampling technique. The households in the selected clusters with pregnant women were known from baseline information and followed for the study. Eligible households were selected using simple random sampling with computer generated random number among selected clusters of kebeles and proportional allocation was used. A family folder prepared by kebele's Health extension workers (HEWs) was used as sampling frame of household. After baseline data were collected from all 750 pregnant women, then these women were divided into randomized groups using a computerized random-number generator, and drawing lots (control n = 375, intervention = 375, ensuring a recruitment balance of 1:1 throughout the study). Health Extension workers working in selected district randomized the cluster, screened and enrolled the study participants.
Three people from intervention and three from control groups were excluded from the study due to reasons such as withdrawal from regular follow-ups and miscarriage. The study was completed with 744 people (Fig. 1).

Blinding (masking)
Allocation concealment is not possible due to the nature of the intervention. The participants, as well as the HDA, are aware of the intervention. However, both the participants and HDA was blinded to the study objectives. We have used objective and reliable outcomes, such as Knowledge, Attitude and practices on optimal nutrition and health and nutritional status, which are less prone to ascertainment bias. Field supervisors, were blinded to the outcome of interest and nally, we were blinded the data collectors for the intervention.

Intervention
To continue with intervention design the researchers started by recruiting Health Development Armies (HDA) by the guide of health extension workers. HDA were recruited from the Kebele's that was speci ed as an intervention groups.
Health development armies were women from the same community (and who were already involved in community activities) and known in the government structures.
After recruitment, HDA were received training based on the protocol developed by the principal investigator for a period of one week related to the key optimal nutrition and health practice message based on (ENA) framework, guidance for formative research on maternal nutrition and making a balanced plate for pregnant women in Bangladesh (3,9,22).
The training had both theoretical and practical demonstration. After successful training was given, HAD were used the standard guidelines to ensure the implementation of the intervention. This intervention (i.e. BCC message) was given to pregnant women by making pregnant women in groups (2 Women's of HDA for one cluster (i.e. in average one HDA women for 10 pregnant women) in their speci ed locality once in every two week for 1:00 up to 1:30 hour length on nonworking days and continued for 3 months. On average, one cluster in a kebele includes 20 pregnant women with total of 24 intervention clusters. Similarly 24 clusters were selected for the control group.
To avoid variability among HDA major assessment questions (checklist) were prepared by the researchers (incorporating both theory and skill) and further consensus was reached. In addition, the researchers and supervisors followed HDA closely during pretest and throughout intervention whether they deliver the intended message or not based on the intervention protocol. Those who had defect in delivering health and nutrition BCC message were identi ed and corrective measure was given immediately. HDA were given posters, broachers and pamphlets with appropriate pictures to be displayed and given to pregnant women in addition to theoretical and practical (demonstration skill).
The main components of the intervention associated with optimal nutrition and health practice were message on eating additional foods during pregnancy (increase meal frequency and portion size with gestational age), avoidance of sharing of food with others, eating a variety(diversi ed) of food from vegetable, fruit and animal sources, utilization of iodized salt, appropriate timing and storage of iodized salt, decrease consumption of iron-inhibiting foods, such as tea and/or coffee, with meals, decrease alcohol consumption, not avoiding important food during pregnancy (associated with taboos), importance of taking iron/folate during pregnancy for at least three months, sleep under an insecticide-treated bed net, seek treatment from Health institution if developed illness(priority for malaria and deworming), wash hands with soap during critical moments, drinking only treated water(if out of pipe water), utilization of health care services, reducing heavy workload, taking day rest, improving self-decision making in food and her own health and importance of obtaining support from the family/Community during pregnancy. Knowledge and Attitude towards optimal nutrition and health were assessed during each BCC session as well. Then, BCC was given based on the identi ed gaps and on locally available, acceptable and affordable foods. Each pregnant woman attended six BCC sessions during the intervention period. The intervention was started in July, 2018 for all pregnant women who were identi ed as intervention group. Nutrition and health BCC message was imparted to the pregnant women (intervention group) for a period of 3 months from July to September, 2018.
Pregnant women in the control groups did not receive the intervention but were exposed to the normal standard of the intervention obtained from health care system and from the community as usual. They were followed till the time frame of the intervention and received the same assessments as the intervention group.

Data collection tools and procedures
A semi structured questionnaire prepared in English language was used to collect data. The questionnaire was translated into two languages (Afan Oromo and Amharic) then back to English by language experts to keep its consistency. The questionnaire was pretested in Ginchi town which is nearby to Ambo district on 39(5%) of the total sample size to identify any ambiguity, length, completeness, consistency and acceptability of questionnaire and some skip patterns were corrected before the real data collection.
Eight diploma nurses were recruited to collect data. Training was given to the data collectors on the objective and relevance of the study, con dentiality of information, respondent's right, informed consent and techniques of interview.
The lled questionnaires were checked for consistencies and completeness daily by four supervisors who had BSc degree in Nursing and principal investigators on the spot. The questionnaire includes part one: socio-demographic and economic characteristics, Part two: maternal characteristics, Part three: Knowledge, Attitude and practices on nutrition and health of pregnant women. Questions related to Knowledge, Attitude and practices was adapted from essential nutrition action frame work and from a guide of formative research for promoting maternal nutrition (3,5,17).
Knowledge of pregnant women about nutrition and health practice was assessed by using 14 questions that were used to assess knowledge of pregnant women regarding optimal nutrition and health was adapted based on the recommendation of ENA message. A knowledge score was calculated for each participant based on the number of questions that were correctly answered in the knowledge assessing questions section. Each correct response was scored 1 and incorrect response scored 0. Pregnant woman was considered to have poor knowledge about nutrition and health if she scored below the highest tertile (i.e. in the rst and/or second tertile) and good knowledge about nutrition and health if she scored in the third tertile. Attitude towards nutrition and health practice was assessed by asking attitude eleven questions. When the pregnant women agreed for questions regarding attitude, she has a score of 2 points, for neutral a score of 1 point and if she respond disagree, scored 0 point following the Likert scale. Then, the total attitude score was determined for each pregnant woman by summing up the scores across the eleven attitudes related questions. Pregnant woman was considered to have unfavorable attitudes if she scored below the third tertile and favorable attitude if she scored in the third tertile.
The outcome variable Nutrition and health practice was measured based on ENA frame work and guidance of formative research for promoting maternal nutrition using questions of practice about the dietary quantity, dietary quality, micronutrient intake, disease prevention and treatment and supportive life style and care.
The respondents had asked to choose Yes=1 or No=0 answers by indicating whether each practice of nutrition and health were practiced or not. To determine the practice of the respondents, scores was computed for the practice variables. One point was allocated to a correct response for each questions and zero for incorrect response and was summed together, the sum of the total scores for the practice varied from (0 up to 20 points maximum score) for each respondent's answers for all 20 nutrition and health practice related questions and this score was converted to tertile.
Pregnant woman was considered to have suboptimal nutrition and health practice if she scored below the third tertile and optimal nutrition and health practice if she scored in the third tertile. Baseline data were collected from all pregnant women (n=750) from June 1-21, 2018 and end point data were collected from pregnant women (n=744) in October, 2018.
Data processing and analysis Data were checked manually for completeness and consistency during data collection before data entry. Then it was entered in to EPI data version 3.1 and exported to SPSS for windows version 21 for cleaning and analyses. The effect of intervention was measured at endpoint of follow up.
First, descriptive statistics like mean and Standard Deviation was done for continuous variable and frequency and percentage for categorical data. A chi-square test was performed to compare the baseline characteristics of the intervention and control groups.
Bivariate analysis was performed between nutrition and health practice during pregnancy and associated factors one at a time. Their odds ratios (OR) with 95% con dence intervals (CI) and p-values was obtained. Factors that were signi cantly associated with nutrition and health practice of mothers during pregnancy at p-value <0.25 in bivariable analysis were entered to multivariate binary logistic regression. p-values at <0.05 was used to declare statistical signi cance. The Hosmer -Lemeshow goodness -of -t statistic is used to assess model goodness of t.
Multicollinearity was checked using Variance In ation Factors (VIF) and there was no Multicollinearity between independent variables. Statistical analyses were conducted primarily on intention-to-treat (ITT) basis which is appropriate for a cluster randomized design study data analysis.

Data quality control
Training of data collectors & supervisors was undertaken. Four supervisors who had BSc degree in Nursing and principal investigators supervised the HDA every two week. To reduce dropout and to increase adherence to the intervention and follow-up program, participants were informed of the importance to attend all sessions and act at their home according to the protocol delivered. Adherence to the BCC sessions was controlled by the HDA, and registered in a personal training diary (attendance sheet). The intervention process was pretested before the implementation of the trial. Pre-tested questionnaire was used to collect the data. Cronbatch's alpha value of knowledge and practice > 0.7 for the whole scale of the instrument was obtained which make it t for use in the study area. The questionnaire was also translated in to language spoken to the study area (Afan Oromo & Amharic) to facilitate understanding of the respondents. Supervisors & principal investigator were closely followed the data collection process. Permission was granted from the West Shoa Zonal and Ambo district health o ces to conduct the study in the respective Kebele's. The nature of the study was fully explained to the study participants to obtain their written informed consent prior to participation in the study and all information obtained was kept anonymous. Password protection of soft copy data and use of key and lock for hard copy data was employed to guarantee con dentiality. Personally identi able information will not be used in the presentation of the ndings in any form.

Results
From a total of 770 assessed for eligibility, 20 pregnant women declined to participate (Fig. 1). A total of 750 were randomized to either the intervention (n=375) or the control group (n=375). Three women in the intervention and three in the control group were lost to follow-up because they developed miscarriage and withdraw from the study. A total of 744 participants were included in the study for analysis (372 in the intervention and 372 in the control groups) (Fig. 1).
The total follow up period was three months and the follow-up started from July, 2018 to September, 2018.
At baseline approximately all participants' characteristics were similarly distributed between intervention and control groups. No signi cant difference was observed between the groups in terms of baseline characteristics like age, occupation (both respondents and husbands), educational status (both respondents and husbands), household size, wealth tertial and estimated time to reach health institution for service (P > 0.05) ( Table 1). Knowledge on optimal nutrition and health Participants who had good knowledge on optimal nutrition and health proportion were almost the same between intervention groups (30.9%) and control group (35.7%) at baseline measurements. However, at endpoint of study period, participants who had god knowledge on optimal nutrition and health among intervention group were 53.2%, while among control group were 39.0% (p < 0.0001) (Fig. 2).

Attitude on optimal nutrition and health
Similarly, participants who had positive attitude towards optimal nutrition and health proportion were almost the same between intervention group (32.5%) and control group (30.1%) at baseline. However, at endpoint of study period, participants who had positive attitude towards optimal nutrition and health among intervention group were 43.3%, while among control group were 28.8% (p < 0.0001) (Fig. 3).
Speci c nutrition and health practices

Quantity of food
There was no signi cance difference at base line measurement between control and intervention group in quantity of food related issues during this pregnancy (p>0.05). However, there was a signi cant difference in the consumption of additional food during pregnancy (44.1% VS 68.5%, p<0.0001) and absence of sharing of food (8.3% VS 27.2%, p<0.0001) at end point of the study as compared to intervention and control group respectively (Table 2). Generally, this study revealed that there was signi cance difference between control and intervention group in the overall optimal nutrition and health practices at the end point of the study(38.7% VS 62.6%, p<0.0001) ( Table 2). Factors associated with optimal nutrition and health practices at endpoint of the study during pregnancy in Ambo district Study group (being in the intervention group), residence, age of the respondent, occupation of the respondents, occupation of the husband, respondent educational status, husband educational status, household size, gravida, number of Antenatal care visit, gestational age, estimated time to reach health institution, knowledge about nutrition and health and attitude towards nutrition and health were found signi cant at p-value <=0.25 in bi-variate analysis. However, Study group (being in the intervention group), respondent educational status(being 5-8 grades), husband educational status (being secondary and above education), estimated time to reach health institution and knowledge about nutrition and health were found signi cantly associated with optimal nutrition and health practices in multivariable binary logistic regression analysis (p<0.05)( Table 3).

Quality of food
Study group (being in the intervention group) was found to be signi cantly associated with optimal nutrition and health practices. Pregnant women who received intervention were 2 times higher odds of optimal nutrition and health practices compared to pregnant women who were in the control group (AOR=2.051,95%CI: 1.318-3.192). Respondent education (being 5-8 Grades) is signi cantly associated with optimal nutrition and health practices. Those pregnant women who attended 5-8 Grades were 52.6% less likely to practice optimal nutrition and health as compared to pregnant women who had no formal education (AOR=0.474, 95%CI: 0.231-0.974). Similarly, husband education is signi cantly associated with optimal nutrition and health practices. Those pregnant women whose husbands had secondary and higher education were 3.152 times more likely odds of optimal nutrition and health practices compared with no formal education (AOR=3.152, 95%CI: 1.406-7.070). Estimated time to reach health institution was signi cantly associated with optimal nutrition and health practices. The study showed that odds of optimal nutrition and health practices among pregnant women who traveled 30-60 minutes and greater than 60 minutes to get care from health institution were 38.6% and 53.2% less likely as compared to pregnant women who travel less than 30 minutes (AOR=0.614, 95%CI: 0.342-1.104) and (AOR=0.468, 95%CI: 0.288-0.762) respectively. The study revealed that those pregnant women who had good knowledge on optimal nutrition and health were 2.187 times higher odds of optimal nutrition and health practices than their counter parts (AOR=2.187, 95%CI:1.356-3.528) ( Table 3).  (23,24). Similarly, participants who were in the intervention group had positive attitude towards optimal nutrition and health than the control group. This study is in line with the study done among pregnant women in Indonesia that revealed providing nutrition and reproductive health education through small groups with interactive methods improves the attitudes of pregnant women (25).
To the best of our knowledge, this is one of the rst cluster randomized controlled community trial (CRCCT) to evaluate the effect of BCC through health development Army on optimal nutrition and health practices. From this study, BCC had a signi cant effect on optimal nutrition and health practices in the intervention group compared with the control group.
Even if di cult to compare with similar target group and topic, this is in line with cross-sectional study done in the Shabelle zone, Somali region of eastern Ethiopia that BCC bring change on infant and young child feeding (IYCF) practices. Result con rmed that in order to be effective BCC needs to be performed using health developmental armies to bridge the gap between knowledge and practice (26). This suggested that, BCC is an important tool to maintain behavior of pregnant women in improving optimal nutrition and health practices among pregnant women (27).
Similarly, the study done in Indonesia, revealed that providing nutrition and reproductive health education improves the practices of pregnant women (25).
Our intervention also supported by health behavior change communication theory, the 'Integrative Model of Behavioral Prediction' which reported a strong intention of a person to perform a behavior, accompanied by the necessary skills and abilities to perform it, and under a conducive environment results in the expected behavior change(28). This behavior change communication message, other than theoretical aspect it also includes a practical demonstration (visual, interactive and skillful) and since its participatory nature engages the pregnant women involved on preparing a nutritionally diversi ed and balanced diet, motivated to take iron/folete, follow disease prevention and treatment practices and also a need of support from the family and community. Behavior change communication through health development Army is an affordable; does not require any food supplementation and suitable intervention since conducted at the community level to improve optimal nutrition and health practices of pregnant women and more likely to be sustainable in resource-poor settings including Ethiopia.
Moreover, this nding had an implication that following close and supportive supervision for health development armies by health professionals, who were used as an implementer of the intervention, resulted in improvement of optimal nutrition and health practices of pregnant women.
This study revealed that, those pregnant women who attended 5-8 Grades were less likely to practice optimal nutrition and health as compared to pregnant women who had no formal education. This nding is supported by the study done in Ghana that, those pregnant adolescents who did not go to school are more likely to be associated with good eating habits than those who attended school (29). Since it is di cult to give the possible reason there is a need for further studies to ascertain contributory factors.
According to this study, husband education is signi cantly associated with optimal nutrition and health practices. The likelihood of optimal nutrition and health practices increased as the education level of the husbands increased. This study is in congruent with the study done in Dakar, Senegal on factors in uencing nutritional practices among mothers that husband education levels were positively correlated with nutritional practices of mothers (30). Similarly, this nding was consistent with study done in in Gedeo zone, southern Ethiopia on dietary practices and its determinant among pregnant women that revealed respondents husband with formal education were more odds of good dietary practice than those without formal education (15). This study also supported by the study done in Harar, Ethiopia; where the risk of malnutrition doubled among women with illiterate husband compared with those with literate husbands (31). This nding implies that husband educational attainment is clearly related to optimal nutrition and health practices of pregnant women.
This nding revealed that the odds of optimal nutrition and health practices among pregnant women who traveled 30-60 minutes and greater than 60 minutes to get care from health institution were less likely as compared to pregnant women who travel less than 30 minutes. This nding is consistent with the study done in Tigray, Ethiopia, that mothers reported the unavailability of health facilities close to their locality was a key barrier to utilizing available services from health institution (32). This has an implication that pregnant women are forced to either travel long distances to get health services or otherwise don't access the service given at health institution including nutrition and/or health related issues.
Moreover, in this study knowledge had shown signi cant association with optimal nutrition and health practice during pregnancy. This nding is supported by the study done in different parts of Ethiopia like Misha Woreda(33), Addis Ababa (34) and West Gojjam (35). The reason behind the similarity might be due to the fact that when women is knowledgeable on optimal nutrition and health, they may be aware of the consequences of suboptimal nutrition and health practice on their fetus, as well as on themselves, and they will more enforced to practice optimal nutrition and health related behaviors. This implies that, though knowledge is not behavior, it determines the practice through motivating mothers to make changes that could maximize optimal nutrition and health practices.

Limitation
This study acknowledges the following limitations; allocation concealment is not possible due to the nature of the intervention, however, both the participants and HDA was blinded to the study objectives.
Since there were no previous studies on this topic, these ndings cannot be easily discussed on the overall optimal nutrition and health practices.

Conclusions
This study demonstrated that behavior change communication (BCC) through women development army is effective in improving optimal nutrition and health practices of pregnant women. Behavior change communication through health development Army is an affordable and suitable intervention to improve knowledge, Attitude and practices of optimal nutrition and health of pregnant women. Therefore, to improve optimal nutrition and health practices of pregnant This gure shows the ow of the study participants through the trial, Ambo district, Ethiopia, 2018 This gure shows attitude towards optimal nutrition and health at baseline and endpoint among intervention and control groups, Ambo district, Ethiopia, 2018