Risk of Low Birth Weight Associated with Family

Risk of Low Birth Weight Associated with Family Poverty in Korea Bong Joo Lee Se Hee Lim Department of Social Welfare, Seoul National University. A Paper submitted ISCI International Conference, November 4 -5, 2009 1

Contents 1 Introduction 2 Previous Research 3 Data and Methods 4 Findings 5 Discussion and Conclusion 2

I. Introduction • The rate of low birth weight increased about 60% from 2. 56% in 1993 to 4. 13% in 2004 in Korea • There has been little empirical research on causes for such a trend in Korea • Previous literatures in Korea, with exceptions of Son(2004) and Lee & Hong(2003), have focused on medical factors in understanding the causes of low birth weight 3

• Son(2004) and Lee & Hong (2003) showed low birth weight is not only a medical concern but also a social problem related to socioeconomic factors in Korea. • But, those study relied on cross-sectional, aggregatelevel data. • Such a strategy is limited in their ability to identify the independent effects of various demographic and socioeconomic factors on low birth weight. • Aggregate-level analyses often fail to disentangle the data into its components, and cross-sectional data are also limited in identifying time order of events of interest. 4

• The primary purpose of our study is to: – examine the effects of family poverty on low birth weight using individual-level data – from a longitudinal survey – based on nationally representative sample in Korea • we also aim to examine what factors mediate the effects of poverty on low birth weight. 5

II. Previous Research • WHO uses low birth weight rate as a primary indicator of public health • WHO specifies low birth weight as weight at birth less than 2. 5 kg. – The definition is based on a study result that showed the risk of infant death is 20 times greater for babies born with weight less than 2. 5 kg compared to those with birth weight 2. 5 kg or higher 6

Consequences of LBW • Low birth weight is known to be associated with infant death, retarded growth, cognitive development, and school achievement of a child. It is also associated with risk of chronic disease at later life • These consequences have a lasting impact on individual’s outcomes, net of socioeconomic conditions 7

Risk Factors of LBW • Most studies show that low socioeconomic status is associated with increased risk of low birth weight • This association persists across various measures of socioeconomic status, including occupation of the mother and/or father, income, and education 8

Pathway of the Poverty Effects • Low income can result in economic deprivation such as inadequate nutrition, housing, and access to health services. • Mental health problem such as depression associated with low socioeconomic status can be a pathway. • Unhealthy behaviors found more in the low income population can increase the risk of low birth weight births among low income mothers. 9

III. Data and Methods 1. Data • Korea Welfare Panel Data from 2006 to 2008 • We pulled 3 years worth of data on new born children of 2006 and 2007 and their mother’s characteristics during the pregnancy period from 2005 and 2006. 10

2. Variables • Low birth weight: weight at birth less than 2. 5 kg • Family poverty: below 60 percents of median income • Control Variables: mother’s age, birth order, education, employment status, mother’s selfreported childhood poverty status • The mediating factors of poverty ü ü The economic deprivation Mother’s depression Smoking Drinking 11

3. Analysis Method • to examine the effects of poverty on low birth weight while controlling for other socioeconomic factors and mother’s behavioral characteristics Logit analysis 12

• to examine the mediating factors Chi-Square method: to examine whether the distribution of experiences of economic deprivation, depression, smoking, and drinking are different across poverty status Baron & Kenny (1986) method: to test if they are mediating factors 13

IV. Findings 1. Characteristics of the Study Population ＜Table 1＞ The Socio-demographic Characteristics of the Subjects: 2006 -07 newborns Means(%) Standard deviation low birth weight 5. 91% . 2363 Poverty 11. 71% . 3221 Birth order(the first) 42. 12% . 4946 Mother’s age at birth 15. 02% . 3579 Mother’s education level 84. 80% . 3595 Mother’s employment status 27. 39% . 4468 Mother’s smoking 0. 2% . 0493 Mother’s drinking 23. 12% . 4224 mother’s childhood poverty 37. 80% . 4858 Economic deprivation 12. 55% . 3318 1. 0% . 0973 Mother’s depression 14

2. Bivariate Analysis Results <Table 2> The rate of low birth weight by household income The rate of low birth weight Household Income Poverty Non Poverty 16. 66% 4. 27% Total 5. 91% • Source: Korea Welfare Panel Data from 2006 to 2008 • N=273(weight: personal longitudinal weight of wave 3). 15 p <0. 05

<Table 3> Economic deprivation, Mother’s depression, Mothers’ smoking and Mother’s drinking by Household income Economic deprivation Household Income Poverty Non poverty p Mother’s depression Mothers’ smoking Mother’s drinking 50. 00 % 6. 25% 0. 00% 28. 12% 7. 46 % 0. 41% 0. 42% 22. 36% <. 01 <. 05 Source: Korea Welfare Panel Data from 2006 to 2008 N=273(weight: personal longitudinal weight of wave 3). 16 >. 10

3. Multivariate analysis results *p<. 10, **p<. 05, ***p<. 01 Model 2 Model 3 Exp(B) 4. 807*** 4. 925** 4. 617** Mother’s age at birth(more than 35) 1. 018 1. 017 1. 012 Birth order(the first) . 297* . 296* . 285* Mother’s educational level(more than high school) . 614 . 619 . 473 Mother’s employment status 1. 043 1. 045 1. 009 Mother’s Childhood poverty 1. 303 1. 312 1. 355 Mother’s smoking . 000 Mother’s drinking . 645 . 647 . 533 Variables Poverty Economic deprivation . 940 12. 557* Mother’s depression Constant -2 Log Likelihood 17 . 070*** 107. 673 107. 667 105. 055

<Figure 1>Mediation effect of depression between poverty and low birth weight Poverty =. 043 P<. 05 Depression =. 301 P<. . 05 Low Birth Weight =. 131 P<. 01 =. 117 P<. . 01 is lower than and still significant. So Depression has a partial mediation effect 18

V. Discussion and Conclusion 1. Findings 1) Controlling for other sociodemographic variables family poverty is a strong predictor of a low birth weight birth 2) Mother’s depression partially mediates the effect of poverty on low birth weight 19

• • • We did not find maternal age at birth being a significant risk factor for low birth weight in our study Birth order is a significant risk factor for low birth weight No effects of mother’s education, employment status, mother’s childhood poverty, smoking, and drinking on the likelihood of low birth weight births in our multivariate models 20

2. Discussions 1)More prevention-based prenatal services are needed for low income pregnant women in Korea 2)Various support services to prevent depression by increasing mother’s self esteem and social support might be effective program tool to reduce low birth weight births in Korea 3)By not including all related variables, there is always a possibility of model misspecification 21

22