The Journal of Nutrition

Supplement: The Impact of Climate Change, the Economic Crisis,

and the Increase in Food Prices on Malnutrition

Household Dietary Diversity and Food

Expenditures Are Closely Linked in Rural
Bangladesh, Increasing the Risk of Malnutrition
Due to the Financial Crisis1–3
Andrew L. Thorne-Lyman,4* Natalie Valpiani,5 Kai Sun,6 Richard D. Semba,6 Christine L. Klotz,7
Klaus Kraemer,8 Nasima Akhter,9 Saskia de Pee,5,7 Regina Moench-Pfanner,10 Mayang Sari,11
and Martin W. Bloem5,7,12
4
Department of Nutrition, Harvard School of Public Health, Boston, MA 02115; 5Friedman School of Nutrition Science and Policy,
Tufts University, Boston, MA 02111; 6Department of Ophthalmology, Johns Hopkins University School of Medicine, Baltimore,
MD 21205; 7Nutrition Service, Policy, Strategy and Programme Support Division, World Food Programme, 00148 Rome, Italy;
8
Sight and Life, 4002 Basel, Switzerland; 9Helen Keller International Asia Pacific, Dhaka 1212, Bangladesh; 10Global Alliance for
Improved Nutrition, 1202 Geneva, Switzerland; 11Helen Keller International, New York, NY 10010; and 12Bloomberg School
of Public Health, Johns Hopkins University, Baltimore, MD 21205

Abstract
In Bangladesh, rice prices are known to be positively associated with the prevalence of child underweight and inversely
associated with household nongrain food expenditures, an indicator of dietary quality. The collection of reliable data on
household expenditures is relatively time consuming and requires extensive training. Simple dietary diversity scores are
increasingly used as measures of food security and as proxies for nutrient adequacy. This study examines associations
between a simple dietary diversity score and commonly used indicators of socioeconomic status in Bangladesh. Data
representative of rural Bangladesh was collected from 188,835 households over 18 rounds of bi-monthly data collection from
2003–2005. A simple household dietary diversity score was developed by summing the number of days each household
consumed an item from each of 7 food groups over a 7-d period. The dietary diversity score was associated with per capita
nongrain food expenditures (r = 0.415), total food expenditures (r = 0.327), and total household expenditures (r = 0.332) using
Spearman correlations (all P , 0.0001). The frequency of meat and egg consumption showed greater variation across quintiles
of total monthly expenditure than other items contributing to the dietary diversity score. After controlling for other measures of
socioeconomic status in multiple linear regression models, the dietary diversity score was significantly associated with monthly
per capita food and total expenditures. Low dietary diversity during the period prior to major food price increases indicates
potential risk for worsening of micronutrient deficiencies and child malnutrition in Bangladesh. J. Nutr. 140: 182S–188S, 2010.

Introduction
Food security is defined by a state in which “all people, at all
times, have physical and economic access to sufficient, safe, and
nutritious food to meet their dietary needs and food preference
for an active and healthy life” (1). Historically, food security
in Bangladesh has often been equated with achieving self-
sufficiency in rice production, stability in rice prices, and energy
availability (2,3). Agricultural policies emphasizing production

1
Published in a supplement to The Journal of Nutrition. Presented at the
workshop “The Impact of Climate Change, the Economic Crisis, and the Increase
in Food Prices on Malnutrition,” held in Castel Gandolfo, Italy, January 25, 2009.
The workshop was organized by Martin W. Bloem, United Nations World Food
Programme, Rome, Italy; Klaus Kraemer, Sight and Life, Basel, Switzerland; and
Richard D. Semba, Johns Hopkins University School of Medicine, and with the
support of an educational grant from Sight and Life, Basel, Switzerland.
Supplement contents are solely the responsibility of the authors and do not
necessarily represent the official views of the organization that they are affiliated
with. Publication costs for this supplement were defrayed in part by the payment
of page charges. This publication must therefore be hereby marked "advertise-
ment" in accordance with 18 USC section 1734 solely to indicate this fact.
Supplement Coordinator disclosures: Martin Bloem, Klaus Kraemer, and Richard
Semba have no relationships to disclose. Supplement Guest Editor disclosures:
A. Catharine Ross and Richard Semba have no relationships to disclose. The
opinions expressed in this publication are those of the authors and are not
attributable to the sponsors or the publisher, Editor, or Editorial Board of The
Journal of Nutrition.
2
Supported by a departmental grant from the Department of Nutrition, Harvard
School of Public Health, and a Peipers Fellowship to A.T.L. and by a Lew R.
Wasserman Award from Research to Prevent Blindness to R.D.S.
3
Author disclosures: A. L. Thorne-Lyman, N. Valpiani, K. Sun, R. D. Semba, C. L.
Klotz, K. Kraemer, N. Akhter, S. de Pee, R. Moench-Pfanner, M. Sari, M. W.
Bloem, no conflicts of interest.
* To whom correspondence should be addressed. E-mail: athornel@hsph.
harvard.edu.

182S 0022-3166/08 $8.00 ã 2010 American Society for Nutrition.

First published online November 18, 2009; doi:10.3945/jn.109.110809.
Downloaded from https://academic.oup.com/jn/article-abstract/140/1/182S/4600361
by guest
on 12 March 2018
 of rice, market liberalization, and technological innovations of
the green revolution enabled the country to more than double
rice production from the 1960s to the present and reduce its
dependency on food grain imports from 12% in 1980 to 5.5% in
2004 (4). The focus on households’ access to rice as an
important component of food security makes sense, given that
rice and wheat contribute 74% of total energy intake and 57%
of total protein intake in the country and account for nearly one-
half of all food expenditures (2,5).
Indeed, historical analysis of surveillance data has shown a
strong association between rice prices, expenditures on rice,
and child undernutrition; as rice prices decline, the prevalence
of child underweight also falls (6). Yet, a full understanding of
the association between rice prices and undernutrition must
also explore issues related to the quality of the diet. Research
has shown that as rice prices fall, households in Bangladesh
continue to purchase the same amount of rice as when rice
prices are high (6). A decrease in rice prices enables households
to spend more money on nonrice foods such as meat, fish, fruit,
oil, and vegetables that tend to be more concentrated sources of
nutrients essential for growth and health. A diverse diet, rare
among poor populations in developing countries, proves
especially important for infants and young children who need
essential micronutrients and energy for rapid physical and
mental development (6–8). In rural Bangladesh, per capita
consumption of vegetables, fish, and tubers falls short of the
minimum level recommended by the government for a healthy
life, whereas an even larger deficit exists for pulses, fruits, oils,
and livestock products (2).
Indicators of dietary diversity, derived from the recall of the
number of foods or food groups consumed over a given time
period, have gained increased attention in both the nutrition and
food security communities in recent years (9). Dietary diversity
indicators prove popular in part because the data are fairly easy
to collect and are associated with dietary quality, energy intake,
and food security (8,9). The use of dietary diversity indicators
holds promise as a powerful tool for effective needs assessments
and targeting, as well as efficient program monitoring and
evaluation.
Many studies have linked household dietary diversity
indicators to improved nutrient intake in developed and de-
veloping countries (8,10–13). Though less frequently explored
in peer-reviewed literature, household dietary diversity holds
promise as a food security indicator as well. The underlying
principle is simple; as poor households gain additional income
they are better able to regularly access foods needed for a
healthy life, thus increasing food security. Poor households
often use additional income to purchase additional nonsta-
ple foods, thus increasing household dietary diversity (6,
14,15). Indeed, a recent analysis found income was a sig-
nificant determinant of household dietary diversity in
Bangladesh (3).
Past studies have validated food security measures by
examining associations between the measure and proxies of
household income, as well as markers of increased quality and
quantity of the diet: total food expenditure and expenditure on
food groups consisting of more expensive, nutrient-dense items
(1,16–18). The present work builds on these analyses, identify-
ing associations between a dietary diversity indicator and
monthly per capita total expenditures, food expenditure, and
expenditure on nongrain foods. This analysis adds to the
literature supporting the use of dietary diversity measures as
indicators that may be particularly useful in settings where more
in-depth tools may not be easily applied.
Downloaded from https://academic.oup.com/jn/article-abstract/140/1/182S/4600361
by guest
on 12 March 2018
Participants and Methods
Sampling
Data were collected from 188,835 households that participated in the
Nutritional Surveillance Project (NSP) of Bangladesh in 2003–2005.
The NSP was established in 1989 by Helen Keller International and the
Institute of Public Health Nutrition of the Government of Bangladesh
(19). A stratified multistage cluster design was used and indicators were
chosen based on the United Nations Children’s Fund’s conceptual
framework of the causes of malnutrition (20). Rural households were
selected from 4 subdistricts in each of the 6 divisions of Bangladesh:
Barisal, Chittagong, Dhaka, Khulna, Rajshahi, Sylhet, and the Chitta-
gong Hill Tracts. Sampling was designed to statistically represent each
of the divisions as well as the entire country of Bangladesh (19).
Data collection
Two-person field teams, trained by Helen Keller International, collected
data every 2 mo coinciding with the different seasons of Bangladesh,
resulting in a total of 18 rounds of data collection over the 3 y covered in
this paper. A new cross-sectional sample was drawn each round. Quality
control was ensured by refresher trainings before each round of data
collection and by random supervisor visits with repeated collection
of a subsample of the data within 24 h of the initial data collection.
A structured coded questionnaire was used to collect data. The
household head or another adult household member provided informa-
tion on the household’s composition, weekly expenditures, land own-
ership, parental education, and other socioeconomic and health
indicators. Each household was asked to report expenditures from the
previous 7 d on rice, wheat, eggs, pulses, fish, vegetables, fruits, meat,
poultry, milk, sweet biscuits, snacks, spices, sugar, cooking oil, and other
foods. Data were collected on the following nonfood expenditures for
the previous month: medical care, education, housing, agricultural
inputs and livestock purchases, electricity, fuel, taxes, loan payments,
and other household items. Expenditure variables were collected in
Bangladeshi taka. A 7-d recall was used to collect information on the
number of days each household consumed typical foods commonly
available in rural Bangladesh, including yellow/orange fruits and
vegetables, green leafy vegetables (shak), meat, chicken, fish, eggs, or
pulses (dal).
Data analysis
A simple household dietary diversity score was created from the
nongrain food groups described above. The score was calculated as the
unweighted sum of the number of days in the previous week households
consumed at least one item from each of the 7 food groups described
above. The range of possible scores was 0–49.
In keeping with our interest in the validity of simple measures, the
proxies used in this analysis for food security and household wealth are
straightforward. In the absence of data on energy availability at the
household level, we use household expenditures as our comparison
measures for food security. Household level food consumption and food
expenditure data were collected in the NSP using a 7-d recall. Though we
consider dietary diversity and food security at the household level, we
created per capita expenditure variables for use in the analysis to control
for the effects of household size on spending. Monthly per capita
nongrain food expenditure represents the sum of reported 7-d expen-
diture on the food categories given in the expanded list above, except
rice and wheat, multiplied by the mean number of weeks per month
(4.34 wk/mo).
Because few of the poorest households purchase large amounts of
rice, a monetary value for in-kind rice was included in total food
expenditure calculations (21). Rice produced on a household’s land,
received as in-kind payment for labor, and received as a gift was reported
in kilograms and assigned a monetary value determined by the daily
market price of rice recorded at the time of the survey. This monetary
value was summed with cash expenditure on grain and nongrain foods
and divided by household size to calculate monthly per capita total food
expenditure. Extreme outliers of total food expenditure were identified
and removed during data cleaning. Nonfood expenditures, collected
using a 1-mo recall period, were summed and divided by household size.
Dietary diversity in rural Bangladesh 183S
 The result was added to per capita total food expenditure to arrive at an TABLE 1 Characteristics of sample (n = 188,835)
estimate of monthly per capita total expenditure, a proxy for household
income. Characteristic % 95% CI
The amount of cultivable land owned by the household, another
important socioeconomic indicator in Bangladesh, was reported in Female headship 4.8 4.7, 4.9
decimals. For analysis, we first calculated the hectare (ha) equivalent of Education
the decimals owned and then created a categorical land ownership Mother has $1 y formal education 56.8 56.5, 57.1
variable using cutoff values consistent with previous research in Father has $1 y formal education 55.9 55.7, 56.2
Bangladesh (2,22). Households owning 0 ha were categorized as Ownership of cultivable land (ha)
landless. Land ownership fell into the following categories: marginal, Landless (0) 59.8 59.6, 60.1
which included any land up to 0.2 ha; small, from 0.201 to 1.0 ha;
Marginally landless (0.01–0.2) 12.1 11.9, 12.3
medium, from 1.01 to 2.0 ha; and large, which included any land .2 ha.
Small landholder (0.201–1.0) 20.7 20.5, 20.9
The area of the main dwelling, in square meters, was calculated by
Medium landholder (1.01–2.0) 4.8 4.7, 4.9
multiplying the reported length and width. For both maternal and
paternal education, values .12 were recoded as 12 y of education Large landholder (.2.0) 2.6 2.5, 2.7
because of the small sample size of individuals with more than a high Occupation category of main earner
school level education. Short-term employed 62.4 62.2, 62.7
Statistical analyses were performed using SAS Survey (SAS Institute). Full-time employed 19.1 18.9, 19.3
To represent the rural population of Bangladesh, analyses were weighted Day laborer 14.4 14.2, 14.5
according to the population of each of the 7 divisions of Bangladesh and Self-employed 3.3 3.2, 3.4
adjusted for the multi-stage cluster design of the NSP. The unit of Unemployed/housewife/retired 0.4 0.4, 0.5
analysis was the household and those with multiple children were
Other 0.4 0.3, 0.4
counted only once. Spearman correlation tests were used to examine
Proportion of monthly expenditure spent on food 59.9 59.8, 60.0
associations between dietary diversity score and expenditure variables,
parental education variables, amount of cultivable land owned by
household, the area of the main dwelling, sex of household head, and
number of members in household.
Differences in mean dietary diversity score between quintiles of total Significant positive correlations were found between the
expenditure and food expenditure were examined using 1-way ANOVA dietary diversity score and the 3 per capita expenditure variables
with multiple comparison tests undertaken using the Tukey-Kramer
(Table 3). Not surprisingly, the strongest correlation was
adjustment (P , 0.05). ANOVA models were also used to compare mean
weekly household food expenditures, nongrain food expenditures, and
between the dietary diversity score and per capita nongrain
expenditures on each individual food group by quintile of dietary expenditures (r = 0.415; P , 0.0001). The magnitude of the
diversity score. Multiple linear regression models were used to test correlations between the dietary diversity score and per capita
associations between each expenditure variable and the score, control- total expenditure and the per capita total food expenditure was
ling for land ownership, parent education, number of household similar (r = 0.332 and 0.327, respectively; P , 0.0001).
members, and area of main dwelling. Values in the text are means 6 SD. Significant associations were also observed between dietary
diversity and other indicators of socioeconomic status, including
Ethics parental education, area of main dwelling, and amount of
The study protocol complied with Helsinki Declaration principles. The cultivable land owned by the household (P , 0.0001). Sex of
field teams were trained to explain the purpose of the NSP and data
household head was weakly associated with dietary diversity
collection to each child’s mother or caretaker and, if present, the father
score (r = 0.038; P , 0.0001). A linear positive relationship was
and/or household head. Data collection ensued only after participants
gave written informed consent and all participants were free to end their observed across quintiles of per capita total household expen-
voluntary participation at any point in the interview. The NSP was diture (Fig. 1). Similar relationships were observed for per capita
approved by the ethical review committee of the Bangladesh Medical household food expenditure and per capita household nongrain
Research Council. Secondary data analysis was approved by the food expenditure (data not shown).
institutional review board of the Johns Hopkins University School of Per capita total food expenditure and per capita nongrain
Medicine. food expenditure differed significantly between all 5 dietary
diversity quintiles and an increasing trend in these expenditures
was observed with increasing dietary diversity score (Table 4).
Results Among the specific food expenditure categories, expenditures
Basic socioeconomic characteristics of the 188,835 households on meat and fruit had the largest proportional increase between
surveyed are presented in Table 1. The household size was 5.7 6 the lowest and highest quintile of dietary diversity score
2.4. More than 90% of households had ,1 ha of land and the (P , 0.0001). Differences in the means between each quintile
majority of households were landless (Table 1). Most house- were significant for all food groups except for rice (P , 0.0001).
holds relied on short-term employment for income; ,15% of
main earners reported being employed full time, while the
majority reported being employed on a short-term basis. Slightly TABLE 2 Characteristics of the diet diversity score1
more than one-half of the fathers and mothers in each household
had received some formal education and, excluding those with Quintile
no formal education, the number of years of education received
First 4.45 + 1.50
by men was 4.5 6 7.4 y and by women was 3.8 6 6.0 y. The total
Second 8.01 + 0.82
household monthly expenditure was 696 6 583 taka (equivalent
Third 10.46 + 0.50
to 11.6 US$ at the time of the study). About 60% of household
Fourth 12.88 + 0.81
total expenditures were spent on food. Household dietary
Fifth 18.32 + 3.45
diversity scores ranged from 0 to 43, the mean score was 10.3
(4.9), and quintiles for the scores are presented in Table 2. 1
Values are mean 6 SD, n = 188,832.

184S Supplement
Downloaded from https://academic.oup.com/jn/article-abstract/140/1/182S/4600361
by guest
on 12 March 2018
 TABLE 3 Spearman correlations between dietary diversity
score and selected socioeconomic and food security
indicators (n = 188,835)
Characteristic R P-value
Monthly per capita nongrain food expenditure 0.415
Monthly per capita expenditure 0.332
Monthly per capita total food expenditure 0.327
Maternal education, y 0.285
Paternal education, y 0.274
Area of main dwelling, m2 0.227
Cultivable land owned by household head1 0.184
Household size, n 0.175
Sex of household head 0.038
1
Cultivable land treated as a categorical variable using same categories in Table 1.
Rice expenditures did not differ between the 3rd and 4th or
between the 4th and 5th dietary diversity quintiles.
There was little variation in frequency of vegetable con-
sumption over the previous week across quintiles of total
monthly expenditures (Table 5). Whereas meat was consumed
relatively infrequently by households in all quintiles, there was
more than a 4-fold difference in consumption between the first
and 5th quintiles (P , 0.05). Similarly, egg consumption differed
markedly between the first and 5th quintiles (P , 0.05).
In multivariate models, after adjusting for the presence of
other covariates, each unit increase in dietary diversity score was
associated with an increase of 29.2 taka of per capita monthly
total expenditure, 16 taka of monthly total food expenditure,
and 15.6 taka of monthly nongrain food expenditure (Table 6).
Comparisons of regression models showed that the dietary
diversity score was better able to account for the variance in
monthly nongrain expenditures than for per capital monthly
total expenditure or per capita monthly total food expenditure.
Discussion
Increasingly, tools based on simple food frequency recalls are
being used to assess both food security of populations in
developing country settings and the nutrient adequacy of the
diets they consume. The analyses presented here suggest that a
simple household dietary diversity score composed of 7 food
FIGURE 1 Mean dietary diversity score by quintiles of total
expenditure. Error bars represent 95% CI.
Downloaded from https://academic.oup.com/jn/article-abstract/140/1/182S/4600361
by guest
on 12 March 2018
groups is fairly well correlated with often-used indicators of
food security and socioeconomic status in rural Bangladesh,
including nongrain food expenditures, total household expen-
ditures, total food expenditures, parental education, house size,
and the amount of cultivable land owned by households (Table
,0.0001 3). The dietary diversity score consistently reflected associations
,0.0001 with food security as measured by total food expenditure and
,0.0001 nongrain food expenditure.
,0.0001 This analysis shows that even prior to the recent food price
,0.0001 crisis and global economic downturn, there was relatively little
,0.0001 variety in the diet of rural Bangladeshi households. Even
,0.0001 households in the upper quintile of total household expenditure
,0.0001 consumed meat, legumes, fruits, and eggs ,2 times in the
,0.0001 previous 7 d (Table 5). Per capita spending on rice differed little
between households in the different quintiles of dietary diversity
score. However, those households in the highest dietary diversity
quintile as measured by the score spent more than 7 times more
on meat, more than 6 times more on fruit, and nearly 6 times
more on eggs than those households in the lowest dietary
diversity quintile. Our findings suggest that nearly all house-
holds, even those in the poorest expenditure quintile, prioritize
the purchase of staple foods. Yet, those with lower dietary
diversity scores are likely to be much more deficient in their
intake of micronutrient-rich foods, especially those of animal
origin. These findings are consistent with a study validating
another food security indicator in which households classified by
the indicator as moderately and severely food insecure con-
sumed significantly less meat, eggs, and fruit than households
classified as food secure (16).
The results also corroborate findings from a multi-country
study that showed significant positive associations between
dietary diversity and per capita purchases of food and nonfood
goods as well as between dietary diversity and 7-d food intake
(17). Similar to our finding that the simple dietary diversity score
was positively associated with mean expenditures on nongrain
food groups, but not as strongly associated with food expendi-
tures that included rice, the multi-country study found the
positive association between increased dietary diversity and per
capita energy availability was stronger for nonstaple foods than
for staples (17).
The dietary diversity score used in this analysis suffers from
some limitations that are largely related to its virtues as a simple
score for gauging household socioeconomic status and food
security. First, no data were collected about the serving sizes or
quantity of food that households consumed from a particular
food group. The score also cannot account for differences in
quality of foods from within the same food group. For example,
a large fish may cost a household so much as to limit con-
sumption of that fish to just 1 time per week, yet some of the
nutrients consumed might exceed those consumed by a house-
hold eating small fish several times in the same week. Thus, the
score may indicate that 2 households have similar access to food
groups when in fact they consume diets of very different nutrient
quality. Recent research suggests that while a simple dietary
diversity score may be a fairly accurate predictor of nutrient
adequacy, a score based on minimum portion size requirements
strengthens correlations between the score and nutrient ade-
quacy (23). Similarly, administering the 7-d recall to households
more than once would be expected to reduce potential misclas-
sifications associated with one-time events (such as weddings or
funerals) that may have been captured in a one-time recall.
The dietary diversity score used in the present analysis may
also lack comparability with scores validated in other contexts.
The 7 food groups included in our score are fewer than the 10
Dietary diversity in rural Bangladesh 185S
 TABLE 4 Monthly food expenditures within each quintile of dietary diversity score1,2

Dietary diversity score quintiles

Expenditure category Q1 Q2 Q3 Q4 Q5 P-value3

n 41219 51721 31108 31466 33318

taka
Rice 154.3a (152.6, 156.1) 164.5b (163.0, 166.1) 169.0c (167.1, 177.0) 169.9d (169.0, 171.8) 172.8e (171.0, 174.6) ,0.0001
Vegetable 30.1a (29.8, 30.4) 33.3b (33.0, 33.5) 36.2c (35.8, 36.5) 38.6d (38.3, 39.0) 44.3e (43.9, 44.6) ,0.0001
Fish 22.9a (22.3, 23.5) 37.1b (36.6, 37.6) 46.7c (46.0, 47.3) 55.1d (54.5, 55.8) 77.3e (76.6, 77.9) ,0.0001
Meat 7.7a (6.9, 8.5) 14.6b (13.9, 15.3) 21.8c (20.8, 22.7) 31.2d (30.3, 32.1) 57.7e (56.8, 58.5) ,0.0001
Legume 4.1a (3.9, 4.3) 6.3b (6.2, 6.5) 8.8c (8.6, 9.0) 11.2d (11.0, 11.4) 16.6e (16.5, 16.8) ,0.0001
Fruit 3.8a (3.5, 4.1) 6.6b (6.4, 6.9) 9.8c (9.5, 10.1) 13.6d (13.3, 14.0) 25.1e (24.8, 25.4) ,0.0001
Eggs 1.6a (1.5, 1.7) 2.5b (2.3, 2.6) 3.7c (3.6, 3.9) 5.0d (4.8, 5.1) 8.5e (8.4, 8.7) ,0.0001
Total food expenditure 287.2a (284.8, 289.5) 342.3b (340.2, 344.4) 382.9c (380.2, 385.6) 419.7d (417.1, 422.3) 517.9e (515.4, 520.4) ,0.0001
Nongrain food expenditure 132.6a (130.9, 134.4) 178.0b (176.5, 179.6) 215.8c (213.8, 217.8) 253.1d (251.2, 255.1) 352.7e (350.8, 354.5) ,0.0001
1
Values are means (95% CI). Means in a row without a common letter differ, P , 0.05 (Tukey-Kramer test).
2
1 US$ was equivalent to a mean of 60 taka during the study period.
3
P-values are from 1-way ANOVA tests used to test group differences.

groups proposed in recent guidelines by the FAO of the United
Nations to standardize dietary diversity score validation (24).
While consistency would enhance comparison across settings,
recent work examining 5 dietary diversity scores of varied
construction found that none proved better than the others at
predicting energy consumption (25). Food group choices for the
score used in our study were based on the availability of food
consumption information in the NSP dataset. However, given
the relative lack of variety in the rural Bangladeshi diet, it is
likely that the inclusion of fewer items may still have enabled us
to capture much of the variety present in the diet compared with
settings where variation might be greater.
Despite these limitations, these findings suggest that the
present dietary diversity score can serve as a useful tool for
assessing food security status in rural Bangladesh, particularly in
situations where rapid assessments are undertaken following
disasters or where it is impossible to administer lengthy
questionnaires. Future research exploring the benefits and
practicalities of collecting information on portion size as part
of such a method is needed, which could potentially enhance its
usefulness as an indicator of nutrient adequacy. Future work
could look into relationships between the dietary diversity score
and indicators of nutritional status. Considering the seasonal
nature of food availability in Bangladesh and the impact of
season on dietary diversity shown in prior research, the score
should be analyzed for consistent utility across seasons (3).
Finally, although a recent study was unable to identify an
effective threshold value that might be used across settings to
consistently identify households falling below a given level of
energy consumption, it would be useful to look into potential
cutoff values of the score that might be used to identify
households that would benefit from food and nutrition assis-
tance programs in Bangladesh (25).
In conclusion, the significant differences in nongrain food
spending and consumption between poorer and wealthier
households in the present analysis draw attention to the need
to consider a broad definition of food security, particularly when
developing policy responses to food crises. Indeed, analysis of
data from Bangladesh has consistently shown that the percent-
age of underweight children was lower in households who spent
more on nonrice foods (6,26). While the findings presented here
reinforce the idea that lowering the price of rice will enable
households to diversify their diets, the fact that even those in the
upper quintile of expenditures consume animal source foods
infrequently suggests the need for interventions that specifically
address micronutrient deficiencies to be part of the response to
food crises (27).
Other articles in this supplement include (28–42).

TABLE 5 Number of times in the previous 7 d that household consumed from food groups within
monthly per capita expenditure quintiles1

Per capita total monthly expenditure quintiles

Food group Q1 Q2 Q3 Q4 Q5 P-value2

n 37730 37735 37735 37731 37732

n/7 d
Vegetables 1.98a (1.96, 2.00) 2.04b (2.02, 2.06) 2.05b (2.04, 2.07) 2.07b (2.05, 2.08) 2.11c (2.10, 2.13) ,0.0001
Fish 3.46a (3.43, 3.48) 3.89b (3.87, 3.92) 4.19c (4.16, 4.21) 4.44d (4.42, 4.46) 4.79e (4.76, 4.81) ,0.0001
Meat 0.34a (0.33, 0.35) 0.46b (0.44, 0.47) 0.61c (0.60, 0.62) 0.87d (0.86, 0.88) 1.40e (1.39, 1.41) ,0.0001
Legumes 1.03a (1.01, 1.04) 1.19b (1.17, 1.21) 1.35c (1.34, 1.37) 1.55d (1.53, 1.56) 1.87e (1.86, 1.89) ,0.0001
Fruit 0.74a (0.72, 0.76) 0.89b (0.88, 0.91) 0.97c (0.95, 0.98) 1.05d (1.04, 1.07) 1.17e (1.16, 1.19) ,0.0001
Eggs 0.62a (0.60, 0.63) 0.82b (0.80, 0.83) 1.02c (1.00, 1.03) 1.25d (1.23, 1.26) 1.62e (1.60, 1.63) ,0.0001
1
Values are means (95% CI). Means in a row without a common letter differ, P , 0.05 (Tukey-Kramer test).
2
P-values are from 1-way ANOVA tests used to test group differences.

186S Supplement
Downloaded from https://academic.oup.com/jn/article-abstract/140/1/182S/4600361
by guest
on 12 March 2018
 TABLE 6 Multivariate linear regression results with per capita expenditure variables as dependent variables1

Per capita monthly total Per capita monthly total food Per capita monthly nongrain
expenditure expenditure food expenditure
Predictor Coefficient SE P Coefficient SE P Coefficient SE P

Dietary diversity score 29.20 0.40 ,0.0001 15.98 0.17 ,0.0001 15.59 0.14 ,0.0001
Household size, n 232.20 0.86 ,0.0001 216.63 0.31 ,0.0001 216.79 0.24 ,0.0001
Mother's education, y 6.16 0.33 ,0.0001 2.36 0.13 ,0.0001 2.37 0.10 ,0.0001
Father's education, y 4.09 0.23 ,0.0001 1.52 0.09 ,0.0001 1.45 0.07 ,0.0001
Cultivable land owned by household, ha 115.82 5.71 ,0.0001 12.57 0.95 ,0.0001 1.97 0.69 ,0.0001
Area of main dwelling, m2 3.41 0.13 ,0.0001 1.58 0.05 ,0.0001 0.164 0.04 ,0.0001
1
US$ was equivalent to a mean of 60 taka during the study period.

Acknowledgments
A.T.L., N.V., R.D.S., C.L.K., K.K., N.A., S.d.P., R.M.P., M.S.,
and M.W.B designed the research; N.A. conducted the research;
K.S., A.T.L., N.V., and C.L.K. analyzed the data; A.T.L. and
N.V. wrote the paper; and A.T.L. had primary responsibility
for final content. All authors read and approved the final
manuscript.
Literature Cited
1. FAO of the United Nations. Food: a fundamental human right. Rome:
FAO; 1996.
2. Hossain M, Naher F, Shahabuddin Q. Food security and nutrition in
Bangladesh: progress and determinants. J Agr Dev Econ. 2005;2:
103–32.
3. Rashid DA, Smith L, Rahman T. Determinants of dietary quality:
evidence from Bangladesh. American Agricultural Economics Associa-
tion Annual Meeting; 2006 July 23–26; Long Beach, CA. Available
from: http://ageconsearch.umn.edu/bitstream/21326/1/sp06ra11.pdf.
4. Chowdhury N, Farid N, Roy D. Food policy liberalization in
Bangladesh: how the government and markets delivered. IFPRI MTID
Discussion paper No. 92. Washington, DC: International Food Policy
Research Institute; 2006.
5. Bangladesh Bureau of Statistics. Household expenditure survey 1995–6.
Dhaka (Bangladesh): Bangladesh Bureau of Statistics; 1998.
6. Torlesse H, Kiess L, Bloem MW. Association of household rice
expenditure with child nutritional status indicates a role for macroeco-
nomic food policy in combating malnutrition. J Nutr. 2003;133:
1320–5.
7. Pan-American Health Organization and WHO. Guiding principles for
complementary feeding of the breastfed child. Washington, DC: PAHO/
WHO; 2003.
8. Arimond M, Ruel M. Dietary diversity is associated with child
nutritional status: evidence from 11 demographic and health surveys.
J Nutr. 2004;134:2579–85.
9. Ruel M. Is dietary diversity an indicator of food security or dietary
quality? A review of measurement issues and research needs. FCND
Discussion Paper No. 140. Washington, DC: International Food Policy
Research Institute; 2002.
10. Kant AK. Indexes of overall diet quality: a review. J Am Diet Assoc.
1996;96:785–91.
11. Savy M, Martin-Prével Y, Sawadogo P, Kameli Y, Delpeuch F. Use of
variety/diversity scores for diet quality measurement: relation with
nutritional status of women in a rural area in Burkina Faso. Eur J Clin
Nutr. 2005;59:703–16.
12. Steyn NP, Nel JH, Nantel G, Kennedy G, Labadarios D. Food variety
and dietary diversity scores in children: are they good indicators of
dietary adequacy? Public Health Nutr. 2006;9:644–50.
13. Kennedy GL, Pedro MR, Seghieri C, Nantel G, Brouwer I. Dietary
diversity score is a useful indicator of micronutrient intake in non-
breast-feeding Filipino children. J Nutr. 2007;137:472–7.
14. Behrman J, Deolalikar A. Is variety the spice of life? Implications for
calorie intake. Rev Econ Stat. 1989;71:666–72.
15. Ruel M, Minot N, Smith L. Patterns and determinants of fruit and
vegetable demand in developing countries: a multi-country comparison.
Prepared for the Joint WHO/FAO Workshop on Fruit and Vegetables
for Health; 2004 Sept 1–3; Kobe, Japan.
16. Melgar-Quinonez HR, Zubieta AC, MkNelly B, Nteziyaremye A,
Gerardo MF, Dunford, C. Household food insecurity and food
expenditure in Bolivia, Burkina Faso, and the Philippines. J Nutr.
2006;136:S1431–7.
17. Hoddinott J, Yohannes Y. Dietary diversity as a food security indicator.
Food and Nutrition Technical Assistance Project Report. Washington,
DC: FANTA; 2002.
18. Hatloy A, Hallund J, Diarra MM, Oshaug A. Food variety, socioeco-
nomic status and nutritional status in urban and rural areas in Koutiala
(Mali). Public Health Nutr. 2000;3:57–65.
19. Bloem MW, Moench-Pfanner R, Panagides D, editors. Health and
nutritional surveillance for development. Singapore: Helen Keller
Worldwide; 2003.
20. de Pee S, Bloem MW. Assessing and communicating the impact of
nutrition and health programs. In: Semba RD, Bloem MW, editors.
Nutrition and health in developing countries. Totowa (NJ): Humana
Press; 2008.
21. Best CM, Sun K, de Pee S, Bloem MW, Stallkamp G, Semba RD.
Parental tobacco use is associated with increased risk of child
malnutrition in Bangladesh. Nutrition. 2007;23:731–8.
22. Bloem M, Hye A, Gorstein J, Wijnroks M, Hall G, Matzger H,
Sommer A. Nutrition surveillance Bangladesh: a useful tool for policy
planning at the local and national levels. Food Nutr Bull. 1995;16:
131–8.
23. Daniels MC, Adair LS, Popkin BM, Truong YK. Dietary diversity scores
can be improved through the use of portion requirements: an analysis in
young Filipino children. Eur J Clin Nutr. 2009;63:199–208.
24. Kennedy G, Nantel G. Basic guidelines for validation of a simple dietary
diversity score as an indicator of dietary nutrient adequacy for non-
breastfeeding children 2–6 years. Rome: FAO; 2006.
25. Coates J, Rogers BL, Webb P, Maxwell D, Houser R, McDonald C.
Dietary diversity study. Prepared for the Strengthening Emergency Needs
Assessment Capacity Project. Rome: World Food Programme; 2007.
26. Campbell AA, de Pee S, Sun K, Kraemer K, Thorne-Lyman A, Moench-
Pfanner R, Sari M, Akhter N, Bloem MW, Semba RD. Household rice
expenditure and maternal and child nutritional status in Bangladesh.
J Nutr. 2010;140:189–94.
27. Valpiani N, Thorne-Lyman AL, de Pee S, Godfrey S, Gentilini U, Bloem
MW. Proposed strategies for mitigating the impact of high food prices
on nutrition and health in Latin America and the Caribbean. CAB
Reviews: Perspectives in Agriculture, Veterinary Science Nutrition and
Natural Resources. 2008;3:1–12.
28. Badham J. World food crisis: a wake-up call to save a generation of
children. participants statement: Castel Gondolfo, Italy, 25 January,
2009. J Nutr. 2010;140:130–1.
29. Bloem MW, Semba RD, Kraemer K. Castel Gandolfo Workshop: An
introduction to the impact of climate change, the economic crisis, and
the increase in the food prices on malnutrition. J Nutr. 2010;140:
132–5.
30. Nabarro D. Editorial comment: helping the vulnerable. J Nutr. 2010;
140:136–7.

Dietary diversity in rural Bangladesh 187S

Downloaded from https://academic.oup.com/jn/article-abstract/140/1/182S/4600361
by guest
on 12 March 2018
 31. de Pee S, Brinkman H-J, Webb P, Godfrey S, Darnton-Hill I, Alderman
H, Semba RD, Piwoz E, Bloem MW. How to ensure nutrition security in
the global economic crisis to protect and enhance development of young
children and our common future. J Nutr. 2010;140:138–42.
32. Webb P. Medium- to long-run implications of high food prices for global
nutrition. J Nutr. 2010;140:143–7.
33. Alderman H. Safety nets can help address the risks to nutrition from
increasing climate variability. J Nutr. 2010;140:148–52.
34. Brinkman H-J, de Pee S, Sanogo I, Subran L, Bloem MW. High food
prices and the global financial crisis have reduced access to nutritious
food and worsened nutritional status and health. J Nutr. 2010;140:
153–61.
35. Darnton-Hill I, Cogill B. Maternal and young child nutrition adversely
affected by external shocks such as increasing global food prices. J Nutr.
2010;140:162–9.
36. Ruel MT, Garrett JL, Hawkes C, Cohen MJ. The food, fuel, and
financial crises affect the urban and rural poor disproportionately: a
review of the evidence. J Nutr. 2010;140:170–6.
188S Supplement
Downloaded from https://academic.oup.com/jn/article-abstract/140/1/182S/4600361
by guest
on 12 March 2018
37. Christian P. Impact of the economic crisis and increase in food prices on
child mortality: exploring nutritional pathways. J Nutr. 2010;140:
177–81.
38. Sari M, de Pee S, Bloem MW, Sun K, Thorne-Lyman A, Moench-
Pfanner R, Akhter N, Kraemer K, Semba RD. Higher household
expenditure on animal-source and nongrain foods lowers the risk of
stunting among children 0–59 months old in Indonesia: implications of
rising food prices. J Nutr. 2010;140:195–200.
39. West KP, Mehra S. Vitamin A intake and status in populations facing
economic stress. J Nutr. 2010;140:201–7.
40. Semba RD, de Pee S, Sun K, Bloem MW, Raju VK. The role of expanded
coverage of the national vitamin a program in preventing morbidity
and mortality among preschool children in India. J Nutr. 2010;140:
208–12.
41. Sztam KA, Fawzi WW, Duggan C. Macronutrient supplementation and
food prices in HIV treatment. J Nutr. 2010;140:213–23.
42. Timmer CP. Preventing food crises using a food policy approach. J Nutr.
2010;140:224–8.