improving the diversity of complementary diets in western kenya
DESCRIPTION
Community-based educational Intervention improved the diversity of complementary diets in Western Kenya. Community-based educational intervention improved the diversity of complementary diets in Western Kenya: results from a randomized control trial improving the diversity of complementary diets in Western Kenya. Presentation by Lydiah M. Waswa: PhD Student, Justus Liebig University- Giessen Find out more about this research: http://www.bioversityinternational.org/news/detail/improving-nutrition-through-local-agricultural-biodiversity-in-kenya/TRANSCRIPT
Community-based educational intervention improved the diversity of complementary diets in Western Kenya: results from a randomized control trial Lydiah M. Waswa: PhD Student , Justus Liebig University- Giessen 28th August 2014
Background
• Prevalence of malnutrition in Kenya: 35% of children aged under five were stunted, 16%
underweight and 7% wasted
• Lack of diversity is a major problem among infants and young children during the complementary feeding period
• Diets of children are staple-based including few or no animal source foods, vegetables and fruitsOnly 39% of children 6-23 months in Kenya are fed with diets reaching the
minimum required dietary diversity (4 out of 7 food groups) (KDHS, 2008/9)
• Inappropriate feeding practices during the complementary feeding period contribute to inadequate nutrient intake among infants and young children
• Even when food resources are available in the home, caregivers are not able to make the best use of them:
Inadequate knowledge Cultural beliefs and practices Inappropriate advice
• Need for interventions to improve the quality of complementary foods by promoting the consumption of a variety of foods, including local foods
Background…..
Objective
• The study aimed to assess the effect of an education intervention focusing on the utilization of local agro-biodiversity in improving the:
Diversity of complementary diets andNutrition knowledge of caregivers
Hypotheses
It was hypothesised that: Children whose caregivers participated in the nutrition education
intervention would receive more diversified complementary diets
Caregivers who participated in the nutrition education sessions would have improved nutrition knowledge
Methods
Study Setting
Teso SouthLM1, LM2
1550-1800 mm
BondoLM3, LM4,LM51020-1100 mm
8
Baseline survey in households with children 6-23 months and their caregivers Jul/Aug 2012 (n=293)
Restricted cluster randomization (using baseline results)District, AEZs, wealth index, stunting, CDDS, education level of caregivers
Nutrition education sessions on complementary feeding
Feb-Jun 2013
Endline survey in households children aged 6-23 months and their caregivers Jul/Aug 2013 ; Intervention group (n=110),Control group (n=97)
Intervention Group (10 villages) Control Group (10 villages)
Study Flow
Middle survey in same households with children 6-23 months and their caregiversto capture seasonal differences, Nov 2012 (n=218)
9
Baseline survey in households with children 6-23 months and their caregivers Jul/Aug 2012 (n=293)
Restricted cluster randomization (using baseline results)District, AEZs, wealth index, stunting, CDDS, education level of caregivers
Nutrition education sessions on complementary feeding
Feb-Jun 2013
Endline survey in households children aged 6-23 months and their caregivers Jul/Aug 2013 ; Intervention group (n=110),Control group (n=97)
Intervention Group (10 villages) Control Group (10 villages)
Study Flow
Middle survey in same households with children 6-23 months and their caregiversto capture seasonal differences, Nov 2012 (n=218)
Study population
• Baseline survey• Households with caregivers and children 6-23 months• Two stage cluster sampling
15 villages randomly from each district proportional to population size 10 households randomly selected per village
• Intervention 10-15 caregivers with children 6-17 months in each intervention village
• Endline survey • Intervention group: All caregivers who participated in the NE• Control group: 10 households per village randomly selected
207 caregivers interviewed: Control (n=97), Intervention (n=110)
Key Principles
• Community health workers (CHWs) received three days training on the topics of the nutrition education sessions
• Nutrition education (NE) sessions conducted jointly with the CHWs
• NE sessions were participatory: • Group discussions• Cooking demonstrations
Caregivers brought ingredients/ foods Using cooking tools and fuels from participants
Sessions Time Topics Materials
1 February • The importance of complementary feeding Main activity: participatory group discussions
IYCF materials
2 February- March
• Dietary diversity during complementary feeding
• Cooking demonstrationMain activities: participatory group discussions and cooking demonstration
Food circle, Seasonal food availability calendars,Locally available foods
General Elections in March 2013 and Follow-up visits in April 2013
3 May • Making nutritious and diverse meals for children aged 6-23 months
• Cooking demonstrationMain activities: participatory group discussions and cooking demonstration
IYCF materials, Food circle, Seasonal food availability calendars, Locally available foods
4 June • How to obtain and prepare adequate and nutritious meals for children 6-23 months
Main activities: group discussions and presentations
Posters, Brochures: Food circle, Nutritious snacks etc.
The themes and topics for the nutrition education sessions were selected based on the findings from the baseline survey with reference to materials from FAO 2004 and UNICEF 2012
Nutrition education sessions
Data collection
• Semi-structured questionnairesSocio-demographic dataChild feeding practicesNutrition knowledge of caregivers
• 24 hour recalls to assess nutrient intakes and dietary diversity
• Anthropometric measurements: Women: weight, height Children: Length, weight
Bioversity International\ L. Waswa
Bioversity International\ L. Waswa
Data analysis • Data was analysed using SPSS version 22
• Children’s dietary diversity scores (CDDS) calculated from the 24-hour recalls based on seven food groups (WHO 2010)
Score range:0-7
• Infant and young child feeding practices assessed based on WHO 2010 infant and young child feeding indicators:
minimum dietary diversity (MDD)
minimum meal frequency (MMF)
minimum acceptable diet (MAD)
Data analysis….
• Wealth Index: Principal Component Analysis (PCA)Characteristics of household dwelling Household ownership of consumer durables Land ownership
• Nutrition knowledge score was computed based on:Caregivers’ knowledge of vitamin A, iron and vitamin CFood sources of the nutrients Importance of feeding children foods rich in these nutrientsScore range: 0-21
Data analysis• Descriptive analyses:
Chi-square test for nominal variables T-test for continuous variables Mann-Whitney test for ordinal variables
• Treatment effect was assessed using the difference-in-differences (DiD) estimator inside a generalized estimating equations (GEE) framework:
To account for a data structure where 25% of the data are panel data and the remaining 75% are repeated cross sectional data
• CDDS and the knowledge score treated as count variables Count regression with Poisson link function and negative binominal function in the
statistical models Results from Poisson regression are presented as incidence rate ratio (IRR) which
expresses the treatment effect as a percentage
• Binary outcomes (MDD, MMF, MAD) were analysed using logistic regression inside the GEE framework
Results
Characteristics
Baseline (n=198) Endline (n=207)
Control (n=99)
Intervention (n=99)
Control (n=97)
Intervention (n=110)
Age children months (mean ±SD) 14 ±5.15 14 ±4.6 16 ±5.11 17 ±4.30
Age caregivers years (mean ±SD) 25 ±5.04 27 ±7.32 26 ±6.20 26 ±6.49
Household size (mean ±SD) 6 ±2.83 6 ±2.45 6 ±2.20 6 ±2.30Education caregiver (%)
No education 9 3 4.1 3
Some primary educ. 41 46 34 50
Completed primary educ. 31 31 36 23
Some secondary 9 13 16 16
Completed secondary 7 5 6 6
Higher education 2 2 4 3
Selected household characteristics
VariablesBaseline (n=198)
P
Endline (n=207)
PControl (n=99)
Intervention (n=99)
Control(n=97)
Intervention (n=110)
n % n % n % n %
Minimum Dietary Diversity (MDD)
50 51 55 56 ns 54 56 96 87 <0.001*
Minimum Meal Frequency (MMF)
72 74 58 59 ns 70 75 81 77 ns
Minimum Acceptable Diet (MAD)
42 43 45 46 ns 45 46 85 77 <0.001*
*Chi square test, ns=not significant
WHO (2010) Indicators for assessing infant and young child feeding practices: Part 2 Measurement .
Infant and young child feeding practices
Consumption of foods from different food groups
Eggs
Vitamin A rich fruits and vegetables
Other fruits and vegetables
Flesh foods (meat, poultry, fish)
Legumes, nuts and seeds
Dairy products (milk)
Grains, roots and tubers
0 10 20 30 40 50 60 70 80 90 100
Endline Intervention
Endline Control
Baseline Intervention
Baseline Control
Percentage (%)
*
*
*
*
*
Effect of intervention on CDDS
• Mean CDDS in the control and intervention groups did not differ significantly at baseline (P=0.510)
• At endline, mean CDDS was significantly higher in the intervention group compared to the control group, P <0.001
• Mean CDDS in the control group dropped significantly at endline, P=0.006
The endline rate of CDDS was at 85.6 % of the baseline values (incidence rate ration (IRR) = 0.856)
Control group Intervention group 0
1
2
3
4
5
6
7
3.78 4.023.4
4.84
Baseline survey (n=198)Endline survey (n=207)
Mea
n CD
DS (0
-7)
Effect of intervention on CDDS
• The treatment effect on CDDS was large, positive and significant (P=0.001)
• The results showed an estimated average treatment effect on the CDDS of plus 27% (IRR=1.27)CDDS rate of the children in the
intervention group was 27% larger than it would have been without the treatment
Control group Intervention group 0
1
2
3
4
5
6
7
3.78 4.023.4
4.84
Baseline survey (n=198)Endline survey (n=207)
Mea
n CD
DS (0
-7)
Effect of intervention on infant and young child feeding indicators
Indicator Odds Ratio (OR)
95% CI P
Minimum dietary diversity (MDD)
4.46 1.84-10.83 0.001
Minimum meal frequency (MMF)
2.21 0.91-5.36 0.080
Minimum acceptable diet (MAD)
3.41 1.50-7.76 0.004
Analysis using the DiD model with logistic regression for binary dependent variables
Effect of intervention on the nutrition knowledge of caregivers
• Mean nutrition knowledge score in the control and intervention groups did not differ significantly at baseline, ( P=0.176)
• At endline, mean nutrition knowledge score was significantly higher among caregivers in the intervention group compared to those in the control group, P<0.001 Control group Intervention group
0
2
4
6
8
10
12
14
2.69 3.143.66
8.21
Baseline survey (n=198)Endline survey (n=207)
Mea
n nu
triti
on k
now
ledg
e sc
ores
(0-2
1)
Effect of intervention on nutrition knowledge of caregivers
• The treatment had a large, positive and significant effect on the nutrition knowledge scores of the caregivers (IRR=2.05), P<0.001
• Nutrition knowledge score did not have a significant or strong effect on CDDS (P=0.731)
Control group Intervention group
0
2
4
6
8
10
12
14
2.69 3.143.66
8.21
Baseline survey (n=198)Endline survey (n=207)
Mea
n nu
triti
on k
now
ledg
e sc
ores
(0-2
1)
Effect of intervention on CDDS and nutrition knowledge of caregivers
Nutrition education
intervention
Children’s dietary
diversity scores
Nutrition knowledge
score of caregivers
P=0.731
P<0.001
P<0.001
Significant effect No significant effect
Conclusions and recommendations
Conclusion
• The nutrition education intervention: Motivated the caregivers to use local food resources to
improve the diversity and quality of complementary diets Low consumption of animal source foods especially flesh meats and
eggs Enhanced nutrition knowledge of caregivers
The increase in nutrition knowledge did not have a direct and significant effect on the CDDS
While increased nutrition knowledge is an important factor, on its own it cannot lead to changes in behaviours
Nutrition education may be more effective when: Combined with other strategies that enhance accessibility to
affordable, culturally acceptable, nutrient dense foods Longer implementation period:
To enable assessment of the long term impact of such interventions on child feeding practices and growth outcomes
Include other members of the family especially fathers and grandmothers:
Have great influence on child feeding and caring practicesProvide a supportive environment that would enable lasting behaviour change among caregivers
Recommendations
• Supervisors:- Prof. Michael Krawinkel (JLU-Giessen)- Dr. Gudrun Keding (Bioversity)- Dr. Irmgard Jordan (JLU-Giessen)
• Funding sources:– BMZ/ GIZ– DAAD/ NCST
Acknowledgements
• Caregivers and their children
• Community health workers
• Local administration
• Enumerators and data entry clerks
• INULA colleagues and Bioversity staff
www.bioversityinternational.org
Thank you