You are on page 1of 11

Socioeconomic Inequalities in Child

Malnutrition in Urban India: A


Systematic Slum Bias?
Sumit Mazumdar (CSSSC, Kolkata) & Papiya G. Mazumdar (FHS, IIHMR,
Kolkata)

8th International Conference on Urban Health


KICC, Nairobi, Kenya
October 21st-24th , 2009

INSTITUTE OF HEALTH
MANAGEMENT
RESEARCH, JAIPUR
Background & Introduction
• India: The ‘Undernutrition’ Burden
• Not a rural phenomenon alone: Growing spread in Urban India
• Increased Urbanization
• Urban India --Livelihood challenges: Growing population pressure
(‘Metro Migration: The City Lights Syndrome’) leading to Slums &
Squatter Colonies with poor civic infrastructure
• Are the Slum population largely responsible for pushing up child
undernutrition levels in Urban India: Looking for a SLUM BIAS

INSTITUTE OF HEALTH
MANAGEMENT
RESEARCH, JAIPUR
Research Questions
• Is the burden of chronic child undernutrition disproportionately on
the population living in the slums of major cities in urban India?
• What is the extent of socioeconomic inequality in chronic child
undernutrition in major cities in India, and how does it relates with
the slum/non-slum differentials in average child undernutrition?
• Is there a definite gradient of socioeconomic inequality in child
undernutrition in the major cities, and is such a gradient more
visible in slums?
• Are the children living in the slums of the major cities in India more
prone to suffer from undernutrition than their counterparts in non-
slum areas?

INSTITUTE OF HEALTH
MANAGEMENT
RESEARCH, JAIPUR
Data & Methods
•National Family Health Survey –III
(NFHS-3): Data for eight major cities
(Delhi City, Meerut, Kolkata, Indore, Cities Slum Non-Slum Total
Mumbai, Nagpur, Hyderabad &
Delhi City 270 (42.1%) 371 (57.9%) 641
Chennai) in India, with separate
Meerut 471 (52.9%) 420 (47.1%) 891
samples of population living in
slums (as defined in Census 2001). Kolkata 296 (61.5%) 185 (38.5%) 481
•Anthropometric data: standardized Indore 356 (53.3%) 312 (46.7%) 668
z-scores for height-for-age (stunting) Mumbai 231 (60%) 154 (40%) 385
primarily used as the major variable Nagpur 312 (50.4%) 307 (49.6%) 619
of interest Hyderabad 374 (49.5%) 382 (50.5%) 756
•SE inequality measured by Chennai 271 (54.1%) 230 (45.9%) 501
Concentration Index. Logistic Total 2581 (52.2%) 2361 (47.8%) 4,942
regressions and Poisson regressions
used for addressing the research
questions

INSTITUTE OF HEALTH
MANAGEMENT
RESEARCH, JAIPUR
Issue 1: Relative burden of Undernutrition in Major Indian
Cities & extent of SE Inequality
Chronic child undernutrition is 60.0% 0.000
significantly higher in the slums
of all the cities 50.0% -0.050

Significant Slum/Non-slum
differentials in Nagpur, N. 40.0% -0.100
Delhi, Kolkata & Indore; More
equal in southern cities 30.0% -0.150

Considerable variation in SE
inequality in chronic 20.0% -0.200

undernutrition in the cities;


Average inequality in the cities 10.0% -0.250

considered substantial (-0.167)


0.0% -0.300
Very high levels of SE

Mumbai

Meerut

Hyderabad
Nagpur

N. Delhi

Kolkata

All Cities

Indore

Chennai
inequality in Meerut (-0.243) &
Chennai (-0.243), less
pronounced in Mumbai (-0.131)
& Nagpur (-0.142) Slums Non-Slums Combined Concentration Index

INSTITUTE OF HEALTH
MANAGEMENT
RESEARCH, JAIPUR
Issue 2: How does SE inequality in chronic undernutrition
associate with intra-city spatial differentials
 No definite association
pattern between extent of
intra-city SE inequality and
slum/non-slum differentials in
chronic undernutrition
(Correlation p = 0.42)

Apparently SE inequality is
greatest in cities with low
slum/non-slum differentials
, at both ends of the
distribution (Chennai &
Meerut): spatial inequality
overridden by SE inequality

INSTITUTE OF HEALTH
MANAGEMENT
RESEARCH, JAIPUR
Issue 3: Examining the Gradient of SE Inequality in
Undernutrition– A Slum Bias?
However, there is no explicit evidence of a
Along expected lines, a higher higher burden of chronic undernutrition
proportion of households in slums among the poor SES from slums (except in
belong to poor SES classes, averaged Meerut & Chennai), as seen from a
over all the major cities. comparison of inter-SES class differentials
between slum & non-slum households
Slum/Non-slum comparison of chronic
Non- undernutrition by interquintile range for SES
26.8 30.8 42.5
Slum 50
40
30
20
10
Slum 44.1 37.3 18.6
0

Poor Middle Rich


Slum Non-slum

INSTITUTE OF HEALTH
MANAGEMENT
RESEARCH, JAIPUR
Issue 3: Examining the Gradient of SE Inequality in
Undernutrition– A Slum Bias?
Odds ratios from Logistic Regression showing relative likelihood of
Significantly lower likelihood of chronic undernutrition according to SES & place of residence
Suffering from chronic
undernutrition in higher SES Cities Model 1 Model 2 Full Model
classes; Intensification of SES Middle SES -0.49** -0.47** -0.27**
gradient on controlling for other Rich SES -0.98** -0.99** -0.37**
model covariates Middle SES*Non-Slum - -0.19 -0.14

No clear evidence of a pro-slum Rich SES* Non-Slum - -0.18 -0.08


bias: The poor SES in the slums do Non-Slum -0.005 -0.003
not have (statistically significant)
higher risks of suffering Predicted probabilities of likelihood of chronic
child undernutrition according to SES, Slum &
undernutrition Non-Slum areas
No significant differences in the
predicted probabilities of suffering 47.4%
46.9% 36.0%
from chronic undernutrition
31.6% 25.7%
between poor SES in slums or non- 22.6%
slum areas
The SES gradient in
Poor Middle Rich
undernutrition is evident, without Slum Non-slum
a discernible pro-slum bias
INSTITUTE OF HEALTH
MANAGEMENT
RESEARCH, JAIPUR
Issue 4: Do staying in slums per se increase the risks of
being chronically undernourished?
Incremental Relative Risk Ratios (IRR) from Poisson Regression showing relative likelihood of
chronic undernutrition according to SES & place of residence (Slum/Non-Slum)

Slum-Non Slum SES Differentials SES Differentials


Model Description Differences (Poor-Middle) (Poor-Rich)
With only slum/non-slum parameter 0.733**
With SES confounder 0.898* 0.821** 0.545**
With confounders for SES, mother's
education, mothers BMI, safe water
availability in hh (community proxy) 0.935 0.870** 0.757**
With full model (confounders for
demographic, socioeconomic,
health care seeking behaviour and
community characteristics) 0.966 0.856** 0.834**

Relative Risk Ratios justify higher risks of being chronically undernourished for children residing in
slums, which, however withers away on gradual introduction of potential confounders with a visible reduction
in risks alongwith loss of statistical significance. Notably, the SES gradient persists even controlling for all
model covariates indicating its overarching influence in shaping risks towards chronic undernutrition in Indian
cities.

INSTITUTE OF HEALTH
MANAGEMENT
RESEARCH, JAIPUR
Conclusion & Policy Implications
• A visible spatial (slum/non-slum) differential coupled with significant SE inequality
in child malnutrition marks the scenario for chronic child undernutrition in major
cities of India

• Absence of any clear-cut evidence suggesting higher burden of chronic


undernutrition in slums, or among the poorer SES households in slums compared to
the non-slum counterparts, suggesting SE inequality in child undernutrition to be a
pan-urban phenomenon in Indian cities, without an identifiable slum bias

• The apparent disproportionate burden of chronic undernutrition in slums, is not


singularly due to slum residence, or even due to non-linear impact of SES in slums
and non-slum localities, but most possibly due to other potential factors like access
to improved health & sanitation facilities, education among mothers, and mother’s
own nutritional status

• Future research needs to specifically focus on typical frameworks to further


understand the dynamics of SES and locality-effect on chronic undernutrition in
major cities across the developing world.
INSTITUTE OF HEALTH
MANAGEMENT
RESEARCH, JAIPUR
Thanks for a patient hearing
Correspondence:
Sumit Mazumdar, Ph.D.
Research Consultant,
Centre for Studies in Social Sciences, Calcutta
R-1, Baishnabghata-Patuli Township, Kolkata
PIN-700094
WEB: www.cssscal.org
email: sumit_majumdar@cssscal.org

INSTITUTE OF HEALTH
MANAGEMENT
RESEARCH, JAIPUR

You might also like