Low Fruit and Vegetable Consumption is Associated with Linear Growth Faltering among Children in Urban Bangladesh

ABSTRACT.

The objective of this study was to examine the association between dietary diversity and child growth among children 6 to 23 months of age in urban slums of Bangladesh. This prospective cohort study enrolled 192 participants 6 to 23 months of age in Dhaka, Bangladesh. To collect dietary intake information for children 6 to 23 months of age, consumption of five or more of the following eight minimum dietary diversity food groups was recorded using 24-hour dietary recall by the child’s caregiver: 1) breast milk; 2) grains, roots, and tubers; 3) legumes and nuts; 4) dairy products; 5) flesh foods; 6) eggs; 7) vitamin A–rich fruits and vegetables; and 8) other fruits and vegetables. Height and weight were measured at baseline and at a 12-month follow-up. Fifty-five percent of children (106 of 192) consumed five or more of the eight minimum dietary diversity food groups during their first visit to assess dietary intake. Eighty-two percent of children (157 of 192) had a report of consuming an animal source of food (e.g., eggs, dairy, flesh foods) in the past 24 hours, 85% (164 of 192) consumed any source of protein (e.g., eggs, dairy, flesh foods, nuts, and legumes), and 65% of children (125 of 192) were reported to consume any fruits and vegetables. Child consumption of fruits and vegetables was positively associated with change in length-for-age z-scores from baseline to the 12-month follow-up (coefficient, 0.86; 95% CI, 0.03–1.68). Inadequate fruit and vegetable consumption was associated with linear growth faltering among the young children residing in urban Dhaka, Bangladesh. These findings demonstrate the need for effective interventions to increase fruit and vegetable consumption in this susceptible pediatric population.

INTRODUCTION

Undernutrition contributes to nearly 45% of deaths among children younger than 5 years of age worldwide, with low- and middle-income countries bearing most of this mortality burden. Based on 2021 global estimates, there are 149 million children younger than 5 years of age who are stunted.¹ Moreover, in low- and middle-income countries, 250 million young children are at risk of failing to achieve their developmental potential.² At least 31% of children younger than 5 years of age in Bangladesh are estimated to be stunted—a key indicator of childhood development.³ This growth failure during infancy and early childhood is often irreversible, leading to short stature during adolescence and adulthood.^4,5 There are also strong associations between childhood stunting and chronic diseases in adulthood.^6–10 Furthermore, stunting is associated with poor cognitive and psychomotor development, increased susceptibility to infections, and an increased risk of mortality during childhood.^5,11 Stunting can be caused by inadequate nutrient intake, which is itself an important cause of other impaired growth measures and adverse child cognitive development.^12–14 Adequate nutrient intake is important throughout life, but is particularly essential during the first 2 years of life, when rapid brain development occurs.¹²

Deficiencies in zinc, vitamin A, and iron—all considered important components of adequate nutrient intake—have been associated with growth faltering among children.¹⁵ Fruits and vegetables are important sources of many macronutrients and micronutrients.¹⁶ Previous studies have found that fruit and vegetable consumption was associated with a lower risk of mortality and chronic diseases.^18,19 Animal source foods (ASF), including meat, fish, milk, and eggs, is another important food group, being rich in macronutrients, especially protein. Low consumption of ASF among children has been associated with impaired growth and adverse cognitive outcomes in young children.^20,21 Dietary diversity, the regular consumption of several different food groups such as ASF or fruits and vegetables, is associated positively with a greater intake of several macro- and micronutrients among young children and improved nutritional status, whereas low dietary diversity has been associated with stunting and impaired cognitive development.^21–29

For children 6 to 23 months of age, the minimum dietary diversity (MDD) score is a population-level indicator used by the WHO to assess child dietary quality and adoption of complementary feeding practices.³⁰ This indicator is defined as a child having consumed five of eight identified food groups during the past 24 hours: 1) breast milk; 2) grains, roots, and tubers; 3) legumes and nuts; 4) dairy products; 5) flesh foods; 6) eggs; 7) vitamin A–rich fruits and vegetables; and 8) other fruits and vegetables. Achievement of MDD in children is a helpful tool for assessing infant and young child diet quality and appropriate complementary feeding practices, because child MDD is associated positively with mean dietary micronutrient adequacy.³¹ Low MDD was found to be associated with a greater risk of stunting in a study conducted using Demographic and Health Survey (DHS) data from Afghanistan, Bangladesh, India, Nepal, and Pakistan.²⁶ Consistent with this finding, studies using DHS data from 21 countries in Africa, Asia, the Americas, and Europe found that MDD was linked with greater length attainment.^32,33 In our study setting in Bangladesh, MDD has been found to be associated with a reduced risk of stunting, underweight, and wasting—three of the most important child growth indicators.^34,35 These findings suggest that diversity in the diets of young children is important to ensure proper growth for this susceptible population.

In our prospective cohort study, we sought to expand on this evidence base by investigating the association between child dietary intake and subsequent growth in urban Bangladesh. We hypothesized that inadequate consumption of fruit and vegetables, ASF, protein, and meeting of MDD would all be associated with impaired growth outcomes among young children.

METHODS

Ethical approval.

Informed written consent was obtained from a parent or guardian for all study participants. The study procedures were approved by the Research and Ethical Review Committee of International Center for Diarrheal Disease Research, Bangladesh (icddr,b), (PR-15133) and the institutional review board of the Johns Hopkins Bloomberg School of Public Health (6785).

Study design.

This prospective cohort study of 192 children 6 to 23 months of age was conducted in slum areas of Dhaka, Bangladesh, as part of the Cholera Hospital-Based Intervention for 7 Days (CHoBI7) mobile health (mHealth) program randomized controlled trial (RCT). The study was conducted from December 2016 to April 2019, and recruited patients with diarrhea presenting with three or more loose stools over a 24-hour period from two tertiary hospitals: icddr,b Dhaka Hospital (private hospital) and Mugda General Hospital (government hospital). Enrolled household members were followed prospectively for 12 months. The eligibility criteria for patients with diarrhea were the following: 1) patients had to have three or more loose stools over the past 24 hours; 2) families planned on residing in Dhaka for the next 12 months; 3) not have a basin for running water in their home (mostly those residing in slum areas of Dhaka); 4) had a child under 5 years of age in their household; and 5) there was at least one working mobile phone in the household. After the recruitment of patients with diarrhea, their corresponding household members were also enrolled. The complete methodology of the CHoBI7 mHealth program and trial is published elsewhere.^36,37 The 6- to 23-month group was the focus of our analysis because this is the age group used for the MDD assessment. This is the weaning period for infants and window in which children are most susceptible to growth faltering.^30,38

Child growth measurements.

Children were measured for anthropometry at baseline and at a 12-month follow-up. Research assistants with training in standardized anthropometry measured the child’s weight once and length three times. The mean value of the three measures of length was used. These measurements were used to calculate z-scores according to the WHO child growth standards.³⁹ Length-for-age z-scores (LAZ), weight-for-age z-scores (WAZ), and weight-for-height/length z-scores were calculated.

Assessment of dietary diversity.

To collect the dietary information of children 6 to 23 months of age, research assistants conducted a dietary assessment during a clinical surveillance visit after the initial baseline assessment through a structured interview administered to caregivers of children younger than 5 years that included whether the following eight dietary diversity food groups were consumed in the past 24 hours based on previously published methods³⁰: 1) breast milk; 2) grains, roots, and tubers; 3) legumes and nuts; 4) dairy products; 5) flesh foods; 6) eggs; 7) vitamin A–rich fruits and vegetables; and 8) other fruits and vegetables. These food groups were used to assess whether children consumed ASF, protein (any source), fruits and vegetables, and to determine their MDD score. MDD is defined by children consuming five out of eight dietary diversity food groups, and is a population-level indicator used by the WHO to assess child dietary quality and adoption of complementary feeding practices.³⁰ Previous studies found MDD to be associated positively with dietary micronutrient adequacy, and with growth.^40–42 We also investigated the association between child growth and consumption of ASF (meat, fish, eggs, milk, or dairy products), any protein source foods (flesh, eggs, milk and dairy foods, legumes, and pulses), any vegetable and fruit source foods (green leafy vegetables, vitamin A–rich fruits and vegetables, and other fruits and vegetables), individual ASF (Supplemental Table S1), and the dietary score as a continuous variable separately (Supplemental Table S2). ASF, any protein sources of foods, and any fruit and vegetable consumption in the past 24 hours was evaluated individually in addition to the MDD, given the literature demonstrating the association between these dietary measures and child health outcomes.^{18–21,27,43}

Statistical analysis.

Our objective was to determine the relationship between child dietary diversity and growth outcomes in urban Bangladesh. Linear growth faltering was defined as a decline in LAZ between baseline and the 12-month follow-up. Growth faltering was also assessed for the change in weight-for-length z-scores and WAZ between baseline and the 12-month follow-up. To examine the association between dietary diversity and changes in child growth, linear regression models were fit using generalized estimating equations to account for clustering at the household level. Dietary measures (ASF, MDD, any protein, and fruit and vegetable consumption) were used as the predictors based on reports from the first visit after baseline enrollment for each child. All regression models were adjusted for baseline age, number of sleeping rooms, roof type, wall type, refrigerator ownership, whether the child was an index patient with diarrhea, and whether the child was breastfed. Regression models were also adjusted for baseline anthropometric z-scores when the outcome was a growth measurement. STATA 13.0 software was used for the analysis (Stata Corp. LLC, College Station, TX).

RESULTS

Household characteristics.

Overall, 192 children 6 to 23 months of age had baseline anthropometric measurements and dietary diversity assessments. Eighty-six percent of children (170 of 192) had follow-up anthropometric measurements and were included in our analysis. At baseline, the median age for children was 8 months, and 35% of children (67 of 192) were female (Table 1). Undernutrition was prevalent, 20% of children (39 of 192) were stunted at baseline, 22% (42 of 192) wasted, and 27% (51 of 192) underweight. Eighty-four percent of children (161 of 192) resided in households with at least one household member who could read and write, and 79% of children (151 of 192) had a caregiver that reported being able to read and write. The median number of individuals living in a household was 4 ± 2 (SD) (range, 3–10), with 70% of children (135 of 192) residing in households with only one room for sleeping. Fifty-one percent of children (98 of 192) resided in households that owned a refrigerator, and 21% of children (40 of 192) resided in a household that owned an animal. Thirty-three percent of children (64 of 192) resided in a household with a concrete roof, 74% (142 of 192) with a concrete wall, and 97% (186 of 192) with a concrete floor. Improved latrine coverage was high, with only 6% of children (11 of 192) residing in households that used an unimproved latrine (flush or pour-flush latrine to other than a sewer system or septic tank, a pit latrine without slab or with broken slab, or a hanging latrine).

Table 1.

Baseline population characteristics (N = 192)

Characteristic	n	%
Female	67	35
Average age of children at baseline, median ± SD (minimum–maximum)	8 ± 3 (1–15)	—
Average no. of individuals living in a household, median ± SD (minimum–maximum)	4 ± 2 (3–10)	—
Literacy of household members (at least one household member who could read and write)	161	84
Caregiver literacy	151	79
Unimproved latrine*	11	6
Household refrigerator ownership	98	51
Household animal ownership (any type)	40	21
Sleeping rooms in household
1 room	135	70
> 1 rooms	57	30
Household with concrete roof	64	33
Household with concrete floor	186	97
Household with concrete wall	142	74
Stunting	39	20
Wasting	42	22
Underweight	51	27
Median height-for-age z-score at baseline, median (SD)	186	–0.86 (1.64)
Median weight-for-age z-score at baseline, median (SD)	190	–1.20 (1.61)
Median weight-for-height/length z-score at baseline, median (SD)]	189	–1.09 (1.42)
Minimum dietary diversity food groups in the past 24 hours	106	55
Breast milk	155	81
Grains, roots, and tubers	176	92
Legumes and nuts	88	46
Dairy products	60	31
Flesh foods	120	63
Eggs	98	51
Vitamin A-rich fruits and vegetables	91	47
Other fruits and vegetables	100	52
Animal sources of food in the past 24 hours	157	82
Protein sources of food in the past 24 hours	164	85
Any vegetable and fruit source of food in the past 24 hours	125	65

^*Unimproved latrine: flush or pour-flush latrine to other than a sewer system or septic tank, a pit latrine without slab or with broken slab, or a hanging latrine.

Dietary diversity, food consumption, and child growth.

Fifty-five percent of children (106 of 192) consumed five or more of the eight MDD food groups at their first dietary assessment visit (Table 1) at baseline. Only 6% of children (11 of 192) consumed all eight food groups. Five percent of children (10 of 192) consumed only one food group. Eighty-two percent of children (157 of 192) reported consuming ASF in the past 24 hours, 85% of children (164 of 192) had a report of eating any source of protein, and 65% (125 of 192) had a report of consuming fruits and vegetables. Eighty one percent of children (155 of 192) consumed breast milk; 92% (176 of 192) consumed grains, roots, and tubers; 46% (88 of 192) consumed legumes and nuts; 31% (60 of 192) consumed dairy products; 63% (120 of 192) consumed flesh foods; 51% (98 of 192) consumed eggs; 47% (91 of 192) consumed vitamin A–rich fruits and vegetables; and 52% (100 of 192) consumed other fruits and vegetables.

At the 12-month follow-up, 41% of children (69 of 170) were stunted, 12% (20 of 170) were wasted, and 31% (52 of 170) were underweight. Consuming fruits and vegetables was associated with a positive change in LAZ from baseline to the 12-month follow-up (coefficient, 0.86; 95% CI, 0.03–1.68), after adjusting for refrigerator ownership, roof type, baseline, status of patient with diarrhea, and breastfeeding (Table 2). Consumption of milk and dairy was associated positively with a change in WAZ from baseline to the 12-month follow-up (coefficient, 0.64; 95% CI, 0.08–1.20) (Supplemental Table S1). There were no other significant associations found between dietary measures and child growth (Supplemental Table S2).

Table 2.

Associations (crude and adjusted) between dietary measures from 24-hour recall and change in growth from baseline to 12 months for children 6 to 24 months

Outcome (change from baseline to 12 months)	Minimum dietary diversity,* coefficient (95% CI)	P value	Animal source food consumption, coefficient (95% CI)	P value	Protein consumption, coefficient (95% CI)	P value	Fruits and vegetables consumption, coefficient (95% CI)	P value
Change in length-for-age z-score
Crude	0.16 (–0.65 to 0.98)	0.69	0.67 (–0.34 to 1.68)	0.19	0.90 (–0.24 to 2.05)	0.12	0.62 (–0.23 to 1.48)	0.15
Adjusted	0.38 (–0.43 to 1.20)	0.36	0.42 (–0.59 to l.44)	0.41	0.45 (–0.69 to 1.59)	0.44	0.86 (0.03 to 1.68)	0.04
Change in weight-for-age z score
Crude	–0.00 (–0.59 to 0.59)	1.00	0.48 (–0.24 to 1.19)	0.19	0.68 (–0.12 to 1.49)	0.10	0.42 (–0.19 to 1.03)	0.18
Adjusted	–0.05 (–0.59 to 0.48)	0.84	0.08 (–0.57 to 0.74)	0.80	0.16 (–0.57 to 0.89)	0.67	0.32 (–0.22 to 0.86)	0.24
Change in weight-for-length z-score
Crude	–0.21 (–0.78 to 0.37)	0.48	–0.07 (–0.77 to 0.64)	0.85	–0.03 (–0.82 to 0.76)	0.94	–0.06 (–0.67 to 0.54)	0.84
Adjusted	–0.47 (–1.03 to 0.08)	0.09	–0.37 (–1.05 to 0.31)	0.29	–0.36 (–1.12 to 0.40)	0.36	–0.39 (–0.96 to 0.18)	0.18

Adjusted models included refrigerator ownership, roof type, baseline, status of patient with diarrhea, and breastfeeding. CIs were estimated with generalized estimating equations accounting for clustering within households. Significant findings are in bold type. Adjusted models include refrigerator ownership, sleeping room, roof type, wall type, baseline age, whether the child was a patient with diarrhea, and whether the child was breastfed.

^*Minimum dietary diversity was defined as whether five of eight of the following dietary diversity food groups were consumed in the past 24 hours: 1) breast milk; 2) grains, roots, and tubers; 3) legumes and nuts; 4) dairy products; 5) flesh foods; 6) eggs; 7) vitamin A-rich fruits and vegetables; and 8) other fruits and vegetables.

DISCUSSION

In this prospective cohort study conducted in Dhaka, Bangladesh, we investigated the association between child dietary consumption and growth. The majority of children achieved MDD, and consumed fruits, vegetables, and ASF. Inadequate consumption of fruits and vegetables was associated with linear growth faltering among young children. These findings demonstrate the importance of fruit and vegetable consumption in early life on child growth, and the need for interventions to promote diverse diets among young children.

Our finding that 55% of children in urban Bangladesh attained MDD is consistent with studies from Nepal (54%)²¹ and western China (55%),⁴⁴ with children achieving similar MDD. Our finding is less than the 74% of children who achieved MDD in Tanzania,²⁵ but greater than the 18% of children achieving MDD in Cambodia.⁴⁵ Our study finding that 55% of children achieved MDD is greater than the 27% found in Bangladesh using 2017 to 2018 DHS data.³⁴ Eighty-two percent of our study children consumed ASF and 63% consumed fruits and vegetables. These results are greater than the 60% consumption of ASF and 50% consumption of fruits and vegetables found in Nepal, and similar to the 80% of children consuming ASF in Indonesia.⁴⁶

Fruit and vegetable consumption was positively associated with linear growth among young children in our study. This is likely attributed to fruits and vegetables containing macronutrients and micronutrients, such as dietary fiber, which are important for maintaining normal function of the human intestines, and vitamin A and folate, which are important for child development.^17,18,43,47 Consistent with our findings, a previous study among Indian children found that linear growth attainment (LAZ) was significantly less among those who did not consume fruits and vegetables.⁴⁸ In a study of DHS data across 49 countries, fruit consumption was found to be associated with decreased stunting in young children.⁴⁹ Although fruit and vegetable consumption is considered an important aspect of a diverse diet, most studies have focused on ASF consumption and improvements in child growth outcomes.^49,50 Future studies are needed that investigate the association between fruit and vegetable consumption and child growth.

In our study, unexpectedly, we did not find an association between child growth and consumption of ASF, protein, or MDD, and the growth of young children, although milk and dairy food consumption was found to be associated with change in WAZ (Supplemental Table S1). Our findings are in contrast with previous multi-country studies that have found an association between ASF and MDD, and linear growth.^{26,31–33,49} One potential reason for this unexpected finding is that our study population had a relatively high consumption of ASF (82%) and MDD (55%), and perhaps this made fruit and vegetable consumption a more important determinant of child growth than when nutritional adequacy was lower. Additional studies in this population that collect blood samples to assess the micronutrient status of children are needed to elucidate the causal pathway by which this association occurred.

This study has some limitations. First, our analysis focused on dietary diversity and not quantity of foods consumed. Future studies should consider including minimum meal frequency at multiple time points. Second, we did not assess the micronutrient status of study children to determine whether they were deficient in vitamins such as vitamin A. Future studies should assess the micronutrient status of study children. Third, we calculated MDD based on a single day’s diet, which may not be be representative of child's overall diet. Fourth, our cohort study focused on households with patients with diarrhea residing in slum areas in urban Dhaka, Bangladesh, which may limit generalizability to other settings. There are estimated to be 29 million individuals living in slum areas in Bangladesh, and 1.2 billon individuals globally.^51–53 Therefore, the study findings could be potentially generalizable to this large population at high risk of diarrheal diseases.

This study has several strengths. First, the study was prospective, minimizing bias and providing stronger evidence for causality. Second, we collected anthropometric data at baseline and the 12-month follow-up, which allowed us to assess linear growth faltering over time. Third, we were investigating the critical role of fruit and vegetable consumption on child growth, building on previous studies that mostly focus on dietary diversity and ASF.

CONCLUSION

This prospective cohort study has shown that inadequate fruit and vegetable consumption is associated with linear growth faltering among the children residing in urban Dhaka, Bangladesh. These findings indicate the need for effective interventions to increase consumption of fruits and vegetables among this susceptible pediatric population.

Note: Supplemental tables appear at www.ajtmh.org.