Nutrition Survey Report - Dhamar Governorate, Yemen: Eastern and Western Districts (23 March 2013 to 3 April 2013)

EXECUTIVE SUMMARY

Dhamar governorate is located to the south and southeast of Sana'a Governorate, to the north of Ibb Governorate, to the east of Hodeidah Governorate and to the northwest of Al Bayda Governorate in the central highlands, with an area of 7,586 square kilometers and a population of 1,329,229 people. The governorate contains of two main ecological zones the Eastern (mainly urban) and the Western (purely rural) regions.

Poverty and unemployment in Dhamar and elsewhere in Yemen, exacerbated by the recent political unrest and by rapid population growth. It is anticipated that by 2034, Dhamar governorate’s population will increase to around four million people if the current fertility rate doesn’t shift, translating to an increase of more than two million people when compared to numbers from the 2004 census. The health sector would also require significant economic resources in 2034. It would be extremely difficult to improve the current service level with continued population growth at the current fertility rate. Costs for the health sector are projected to be over $43 million, an increase of $25 million compared with 2009.

The 2012 WFP-CFSS reported that 24.1, 22.0, 46.1 per cents are severely, moderately, and "severe and moderate" food insecure in Dhamar governorate respectively.

Regarding malnutrition, the 2010 IFPRI National Food Security Paper estimated GAM rate based on HBS 2005-06 data in Dhamar to be 11.2 per cent, with SAM at 3.3 per cent, underweight 46.7 per cent, and stunting 70.1 per cent. The 2012 WFP-CFSS, reported the following prevalence: GAM: 9.8 per cent, SAM: 1.8 per cent, underweight: 36.1per per cent, and stunting: 59.2 per cent. According to Nutrition cluster strategy 2012/Yemen, Dhamar fall within the poor zone (GAM from 5 – 9.9%).

Between 23rd March to 3rd April, 2013, MoPHP and UNICEF conducted two inter-agency nutrition surveys using the Standardized Monitoring and Assessment for Relief and Transition (SMART) methodology covering the Eastern and Western districts in Dhamar Governorate. This was a Yemen Nutrition Cluster initiative to establish and monitor the levels of acute malnutrition, stunting and underweight among children aged 6-59 months in the different livelihood/ ecological zones, identify some of the factors associated with malnutrition, and inform on the appropriate responses.

Using a two-stage Probability Proportionate to Population Size (PPS) sampling methodology, 36 clusters in each of the Eastern and Western Dhamar were randomly selected for both anthropometric and mortality assessments. The calculated sample sizes in the Eastern district and Western Dhamar using ENA for SMART software were 681 and 835 households respectively for assessing both the anthropometry and mortality.

In the Western Dhamar the Global Acute Malnutrition (GAM) rate was 9.2 per cent (95% CI: 6.8 - 12.4), with Severe Acute Malnutrition (SAM) 1.0 per cent (95% CI: 0.6 - 1.8).

GAM and SAM rates in the Eastern Dhamar were 4.9 per cent (95% CI: 3.7 - 6.5) and 0.5 per cent (95% CI: 0.2 – 1.5) respectively. According to WHO categorization, these rates indicate that the nutrition situation in Western Dhamar fits with the upper limit of cut-off values for “poor” (which equal to GAM rates 5-9.9 per cent) and in the Eastern Dhamar fits with the upper limit of cut-off values for “acceptable” (which equal to GAM rates <5 per cent).

Stunting rates in the Western and Eastern Dhamar are 69.5 per cent (95% CI: 65.7 – 73.1) and 62.8 per cent (95% CI: 57.4 – 67.8) respectively with severe stunting of 31.8 per cent (95% CI: 28.1 - 35.8) and 29.8 per cent (95% CI: 24.2 - 36.1) respectively.

These rates are much beyond the critical levels of 40 per cent; thus the stunting rates are of great concern.

Underweight rate in the Western Dhamar is 45.5 per cent (95% CI: 41.1 - 50.0), with severe underweight of 12.5 per cent (95% CI: 10.0 - 15.5) while the underweight and severe underweight rates in the Eastern Dhamar are 35.1 per cent (95% CI: 29.7 - 40.9) and 8.9 per cent (95% CI: 6.3– 12.2), respectively. These rates are much more than the critical levels of 30 per cent, as per WHO categorization.

Such pattern of malnutrition–especially in the Western districts- is somewhat similar to the pattern that found in neighbouring Ibb that shows more predominant chronic malnutrition (as reflected by extremely high stunting rates) but less severe acute malnutrition (especially in the Eastern Dhamar). This should be in interpreted as an indication of poor environmental conditions or long-term restriction of a child's growth potential together with chronic food insecurity and poverty.

Both stunting and underweight significantly differ between the rural and urban areas and significantly higher among children aged 36 and above than younger age group.

Comparable to the findings from previous nutritional surveys (e.g. Hodeidah, Hajjah), all types of malnutrition found to be higher among males than females, however such difference is only statistically significant for underweight.

The two main sources of drinking water in the Eastern Dhamar were house-connected piped water (41 per cent) and water from protected open well (32 per cent) while in the Western Dhamar they were water from protected open well (31 per cent) and unprotected surface water (17 per cent). The majority of Eastern Dhamar (88%) population are having flush/pour flush latrine compared to only 58% in the Western Dhamar. The main two sources of income for Eastern Dhamar are fixed monthly waged work and casual labour (29% each) compared to casual labour (45%) and fixed monthly waged work (14%) for Western Dhamar. More than 80% of the Eastern and Western districts' population seeks health services from a public health facility when sick.

There was high prevalence of common disease, as recorded during the survey: diarrhea, Acute Respiratory Infection (ARI) and fever prevalence are the reported cases 2 weeks before the survey while measles was one month before the survey, as shown in table 1 below and all are significantly higher in the Western than Eastern Dhamar. Of these morbidities only diarrhea prevalence is significantly higher among children aged less than 36 compared to those who are 36 and above (39% vs. 25%, X2 42.1, P<0.0001, df 1). Among children aged 6 to 24 months although morbidities prevalence does not differ by breastfeeding status or feeding 4 times and above (other than breastfeeds), however diarrhea was higher among children who were given more than one milk feed (other than breast milk) in the previous day to the survey (50% vs. 37%, X2 11.3, P<0.01, df 1) which may be related to unhygienic preparation or administration of artificial milk.

Having diarrhea also found to be significantly associated with GAM: 9% vs. 6%, X2 5.0, P<0.05, df 1, and fever significantly associated with underweight: 50 % vs. 45%, X2 5.7, P<0.05, df 1. However, morbidities does not differ by urban/rural residence, gender, vitamin A supplementation, or vaccination except for measles that is significantly higher among non-vaccinated children to measles (5.2% vs. 1.5%, X2 19.0, P<0.0001, df 1).

It is notable that the vitamin A supplementation 6 months prior to the survey in the Dhamar which is 81% is still lower than the recommended 95 per cent coverage Sphere Standards, 2011. Such coverage significantly higher in Eastern than Western Dhamar (89.0% vs. 79.0%, X2 21.8, P < 0.0001, df 1), and urban than rural residence (93.0% vs. 83.0%, X2 6.5, P < 0.05, df 1). Vitamin A supplementation found to be highly significantly associated with lower prevalence of measles (1.8% vs. 6.4%, X2 23.2, P < 0.0001, df 1) but neither with other types of morbidities nor with gender or malnutrition.

Regarding vaccination, the percentage of children who have been vaccinated with the third dose of polio vaccination and measles found to be significantly higher in the Eastern than in Western Dhamar (76.0% vs. 63.0%, X 2 36.4, P<0.0001, df 1) and (79.0% vs. 71.0%, X 2 17.4, P<0.0001, df 1) respectively. Both vaccines are also higher in urban than rural residence: 84.0% vs. 71.0%, X 2 7.1, P<0.01, df 1 and 87.0% vs. 77.0%, X 2 4.5, P<0.05, df 1 respectively. However, no association was found between vaccination and gender, malnutrition and morbidities except for higher measles prevalence with not receiving polio vaccine (5.0% vs. 1.5%, X 2 19.2, P< 0.0001, df 1) and higher ARI, fever and measles prevalence with not receiving measles vaccine.

Only 3.0 per cent of households in Western compared to 21.0% in Eastern Dhamar are using adequately iodized salt (X 2 122.0, P< 0.0001, df 1) which is also significantly lower in rural than urban residence (29.0% vs. 10.0%, X 2 50.0, P< 0.0001, df 1).

Among children aged 6 to 24 months in Dhamar, only 61 per cent still breastfeed. This found to be significantly higher in Western than Eastern Dhamar (57 % vs. 65%, X2 4.0, P<0.05, df 1) but does not differ by urban/rural residence. Breastfed also found to be significantly higher among males than females (65 % vs. 57%, X2 5.2, P<0.05, df 1).

Regarding malnutrition and still breastfed, there is only significantly higher prevalence of underweight among those who are still breastfed compared to those who have ceased breastfeeding (41 % vs. 31%, X2 7.7, P<0.01, df 1) may be due overreliance on breastfeeding with poor complementary feeding. One of four children does not receive the recommended number of meals (4 meals and above), as per UN-FAO recommendations. However, this does not differ between the Eastern and Western Dhamar, urban/rural residence, or by gender. Furthermore, the recommended number of meals shows no effect on levels of wasting, stunting, and underweight. Among children aged 6 to 24 months, less than half received more than one milk feed in the last 24 hours. The number of feeds (other than breastfeeds) does not differ between Eastern and Western Dhamar, urban and rural, or gender. Although the children who have received more than one milk feed in the last 24 hours shows lower prevalence of wasting, stunting, and underweight, the difference was not statistically significant.

Regarding WASH, overall 29% of Dhamar households drink water from unclean water container (i.e. algae seen) that found to be significantly higher in the Western than Eastern Dhamar (41% vs. 17%, X2 109.2, P<0.0001, df 1), and in rural than urban residence (53% vs. 4%, X2 32, P<0.0001, df 1). Diarrhea prevalence found to be significantly higher among households using unclean water container (40% vs. 27%, X2 16.0, P<0.0001, df 1). Although wasting, stunting, and underweight are higher among those who are drinking from unclean water container, this was only statistically significant for underweight (46% vs. 35%, X2 11.2, P<0.01, df 1) and SAM (1.6% vs. 0.1%, Fisher exact test X 2 8.1, P<0.05, df 1). Whether these are direct effects or due to other confounding e.g. diarrhea or socioeconomic status is not certain. Around three quarters of households in Dhamar are having flush or pour flush latrine. This was significantly higher in Eastern than Western Dhamar (88% vs. 58%, X2 174.1, P<0.0001, df 1) and in Urban than rural Dhamar (99% vs. 70%, X2 62.1, P<0.0001, df 1). However, having no flush/pour latrine found not to be significantly associated with diarrhea or malnutrition except for stunting (73% vs. 64%, X2 8.4, P<0.01, df 1). Whether this is a direct effect or due to other confounding (e.g. poorer socio economic status etc.) is behind the scope of this survey. Overall, the Western Dhamar have significantly poorer handwashing practices than eastern Dhamar e.g. for washing hand after toilet (47% vs. 62%, X2 38.1, P<0.0001, df 1), washing hand before meal (33% vs. 41%, X2 9.6, P<0.01, df 1), washing hand before cooking (25% vs. 35%, X2 19.3, P<0.0001, df 1). Similarly, overall household caretaker handwashing practices are significantly poorer in rural than urban areas. The availability of soap at handwashing facility was also significantly higher at Eastern than Western Dhamar (94% vs. 75%, X2 109.4, P<0.0001, df 1) and urban than rural Dhamar (97% vs. 83%, X2 21.6, P<0.0001, df 1) however it does not significantly associated with diarrhea. Of those practices only not handwashing after disposal of child faces and not handwashing before cooking were significantly associated with diarrhea: 33% vs. 26%, X2 5.0, P<0.05, df 1and 33% vs. 27%, X2 3.8, P=0.05, df 1 respectively. Furthermore, only not washing hand before cooking (41% vs. 32%, X2 6.9, P<0.01, df 1) shows significant association with underweight and not washing hand before child feeding and not washing hand after child faces disposal were significantly associated with stunting: 68% vs. 58%, X2 5.3, P<0.05, df 1 and 69% vs. 60%, X2 6.5, P<0.05, df 1. Whether such associations are direct associations or confounded by other factors e.g. diarrhea, socio economic status is not ascertained.

Regarding food insecurity, all indicators found to be significantly much higher in Western than Eastern Dhamar e.g. reduced meal size (20% vs. 4%, X2 88.2, P<0.0001, df 1), or reduced meal number (18% vs. 3%, X2 89.8, P<0.0001, df 1), and member go the bed hungry because of not enough food (17% vs. 4%, X2 69.8, P<0.0001, df 1). Coping strategies are also more predominant at Western than Eastern Dhamar e.g. borrowing food/money to purchase food or purchase food in credit or mortgage if the reason that HH has not money (36% vs. 17%, X2 71.2, P<0.0001, df 1) and reduced expenditure on health/education (24% vs. 7%, X2 87.2, P<0.0001, df 1). This confirms the finding of 2012 WFP-CFSS that food insecurity is a problem among Dhamar population. Similarly, food insecurity indicators also much severer in rural than urban Dhamar e.g. reduced meal number (11% vs. 4%, X2 7.4, P<0.01, df 1), and member go the bed hungry because of not enough food (11% vs. 4%, X2 4.2, P<0.01, df 1). This support previous findings that rural-urban inequalities are high in Yemen and the number of food insecure people living in rural areas is more than five times higher than in urban areas.

Although the prevalence of some types of malnutrition (e.g. underweight) found to be slightly higher among food insecure households, such difference was not statistically significant. This may indicates that the malnutrition in Dhamar is a health related problem (e.g. associated with diarrhea) or a practice problem (e.g. IYCF) rather being a purely food insecurity in origin.

Finally, about 80% of the caretakers in Dhamar have no formal education. Although, the prevalence of underweight and stunting was higher among illiterate mothers' children: 40% vs. 33% and 68% vs. 61% respectively such differences were not statistically significant. Similarly some morbidities e.g. diarrhea is higher among illiterate mothers' children (32% vs. 27%) however, such difference was not statistically significant.

Nevertheless, illiterate mothers' children have significantly lower polio vaccination (70% vs. 81%, X2 9.5, P<0.01, df 1) and lower measles vaccination (76% vs. 86%, X2 8.7, P<0.01, df 1).

Finally, overall mortality rates are low and within the acceptable levels according to WHO categorisation, hence not raising major concern.

Recommendations

In a view of very high chronic malnutrition found in this survey with less severer acute malnutrition –especially in Eastern Dhamar-, but with the risks to health may be similar to many acute emergencies, short-term emergency responses alone are not enough to address such long-term problems and there is a need to concentrate on both immediate as well as medium term Interventions. Therefore, there is a need to deliver an integrated package of services to mothers and their children not only to address the acute malnutrition but also to address the high level of stunting and underweight as well as other development indicators.

Immediate Interventions

• Develop detailed integrated response micro-plan articulating district level humanitarian needs, delivering response package, coverage and gaps to document the progress, advocacy and lessons learnt.

• Priority should be given to pockets of vulnerability in both zones through mobilizing outreach services to rapidly address the high GAM/SAM rates.

• Development of CMAM protocol, strengthen and expanding CMAM services to reach all the existed health facilities and outreach services.

• CMAM services should adhere to the CMAM protocol (ensuring systematic treatment and full consideration of moderate acute malnutrition management) that should be integrated with infant feeding, hygiene promotion and food security interventions.

• To promote appropriate IYCF practices (early initiation of breastfeeding, exclusive and sustained breastfeeding for 2 years and promotion of appropriate complementary feeding practices for children aged 6 to 24 months) along with micronutrient supplementations and deworming.

• Accelerate the integration of IYCF counseling into all CMAM services delivered by both fixed and mobile clinics.

• Intensive social mobilisation campaigns on IYCF feeding and caring practices through behavior change / communication interventions mainly in the following areas; timely introduction of complementary food and continue breastfeeding up to two years, along with Vit. A supplementation, micronutrient supplements, and promotion of safe sanitation and hygienic practices including hand washing with soap as well as safe disposal of children’s excreta, diarrhoea prevention measures and appropriate management of ARI among young children.

Medium Term Interventions

• From the development point of view, there is an essential need for Yemen to be an active member in the global SUN movement.

• High level advocacy with the GOY and politicians to mobilise their commitment to fight undernutrition among U5 Yemeni children.

• Scaling up implementation of the national nutrition strategy and related action to address the high level of malnutrition in line with the lifecycle approach along with promotion of maternal nutrition.

• Continued support for longer term water development and sanitation programmes throughout the governorate, with community mobilization activities to promote safe sanitation and hygienic practices.

• Follow up SMART nutrition survey and coverage survey in 2014 to track the progress on implementation of the response plan.

• Exploring new initiatives to promote small scale income generating projects, draw lessons learnt and replicate the successful projects.

• Promote improved latrine use and other hygiene services like Community Led Total Sanitation (CLTS) strategy.

Other Recommendations

• Further investigation is needed to understand the causality tree behind high level of acute malnutrition among boys compared to girls found earlier surveys conducted in Hodeidah, Taiz, Hajja governorates as well as in this survey (especially for underweight).

• Undertaking full scale national nutrition and mortality survey.

• In a view of high malnutrition among illiterate mothers' children as well as poor child feeding practices and health indicators (e.g. vaccination, Vitamin A supplementation) found in this survey as well as previous surveys in Yemen, a focus on girls’ education is necessary in the long term battle against malnutrition as well as for broader development.