Effectiveness of selling fruits in pieces or in bulk in improving fruit consumption of high-school students: A cluster randomized controlled trial in urban setting of Ethiopia
No Thumbnail Available
Date
2019-10
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Background: Diets high in fnlits and vegetables arc widely recommended for hcahhicr life.
Fruits have historically he ld a place in dietary guida nce because of their concentrations of dietary
fib er, vitamins, minerals, electrolytes; and morc recently phytochcmicals, especially
antioxidants. However, fruit consumption in Ethiopia is very low. The barriers and enablers
affecting consumer behavior towards fruits consumption including price, preference, qual ity,
safety and access are largely undocumented; hence, maki ng it diffi cult to design effective
interventions that improve consumption of fruit s. For example, fnlils arc sold around some
schools, but the minimum amount that one can buy is half a ki logram, This may discourage
students from buying fru its as a snack and may contribute to the purchase of sweets and biscuits
that arc more affordable to thcm.
Objective: The objective of this study was to evaluatc whet her selling fmits in pieces improves
consumption and volume of sale than the current practice of selling fruits in bulk (kilogram).
Methodology: A cluster randomized contfolled trial was conducted from April through June
20 19 in selected schools of Addis Ababa, Ethiopia. A double population proportion for cluster
randomized trial was employed to calcu late thc sample size. After adding 5% non-response rate,
the lolal smnple size was estimated to be 369. Data were collected on socio-demographic
infomlation, fruit consumption pattem, vo lumc of sale per day and other variable that determine
fruit consumption pattem by usi ng semi structured and pre tested tool. EPI data version 4.2.0 was
used to code, enter and clean the collected data. Data WCfe analyzed usi ng SPSS version 20 and
SAS version 9.4. Descriptive statistics were lIsed to describe variables in relation to thc outcome.
Mixed-effect linear regress ion was employed to see the impact of intcrvention. We lIsed
unstructured variance matrix stnteture to estimate the parameters of the random intercept. All
tests were two-sided and P<0.05 was considered stati stically significant. We report Ihe parameter
estimates with 95% CI and standard errors (SE).
Result: Out of 157 partic ipants in the intervention group 80 (51 %) and 77 (5 1 %) were male and
female respectively. In the control group, 65(37.6%) and 108 (62.4%) were male and female
respectively. The mean (SO) age of intervention and control group was 15.9 ( 1.2) and 17 (1.06)
respectively. With regards 10 a one-week recall fruit consumption, Banana consumption was
82% in the intervention group when compared with 65% of control (P <0.001), Water melon and
mango consumption was also higher in intervention group with proportion of 12.7% and 61.1 %
in comparison wi th 4% and 44.5% (P<O.OO 1 and P::0.003) of intervention VS control group
respectively. The mixed effect linier regression output showed that there was a statistically
significant difference in fruit consumption between intervention and control group (13=0.853.
p=O.OO I). But the intervention has the same impact throughout the time of intervention (lJ;;.
0. 138, p=0.37 I ) and (13=·0.04, p=1.00) in round one and two. More over pocket money has no
interaction with the effect of the intervention (/3=0. 145, p=0.686). Volume of sale in int ervention
arm was around 22 kg per day compared to 16 kg in control ann at the tinal observation.
Conclus ion: Selling fruit in pieces has showed a significant increment in overa ll frui t
consumption of high school adolescents. In addition, the study revealed thai consumption of
fru its like banana, orange, mango and strawbelTY has showed improvement in the intervention
ann. The study also showed that there was a change in vo lume of sale in the intervention ann
than the control group. However, fruit consumption based on a 24·hour recall fruit frequency
data didn't bring a statistically significant effcct among adolescents.
Description
Keywords
Fruits, Adolescent. Market intervcntion, Randomized tri al, Addis Ababa