Date

June 2009

Document Type

Dissertation

Degree Name

Ph.D.

Department

Dept. of Public Health and Preventive Medicine

Institution

Oregon Health & Science University

Abstract

Background Food insecurity is a problem that affects millions of households in the United States every year, and is associated with poor health outcomes. Despite its high prevalence, food insecurity is a phenomenon that is not routinely screened for by most health care providers. Children with special health care needs (CSHCN) are a population that is defined by the presence of, or risk for, a chronic condition, and an elevated need and use of health care services. The increased medical costs for families of CSHCN contribute to the risk for food insecurity; this population represents unique opportunities for screening and outreach. While the financial burden of having a CSHCN has been documented in the scientific literature, the association between food insecurity and having a CSHCN has not yet been studied. This study seeks to: 1) estimate prevalence of food insecurity among mothers in Oregon, 2) estimate prevalence of children with special health care needs (CSHCN) among two-year-olds in Oregon and 3) test the hypothesis that mothers who have two-year-old children with special health care needs are at higher risk for food insecurity than mothers whose children do not have special health care needs. This hypothesis will be examined using both a cross-sectional and a longitudinal analysis approach. Methods Oregon Pregnancy Risk Assessment Monitoring System (PRAMS) surveys mothers who delivered live births in Oregon. At two to six months postpartum, mothers are questioned about perinatal health events and birth outcomes. Questionnaires are mailed to mothers identified by monthly stratified sampling, oversampling for race/ethnicity categories and low birth weight babies. Oregon PRAMS-2 was developed as a two-year follow-up survey of mothers who responded to PRAMS, and was designed to assess early childhood health issues. PRAMS and PRAMS-2 data are weighted to provide a population-based sample and to allow for population-based estimates. Data from the 2006 Oregon Pregnancy Risk Assessment Monitoring System: Two-Year Old Survey (PRAMS-2), a two-year follow-up study of women initially surveyed in 2004 Oregon PRAMS, were used for this study. Responses to PRAMS and PRAMS-2 were considered to occur at Time 1 and Time 2, respectively. Individual responses were also linked to birth certificate data. Of the 1,968 women who responded to 2004 Oregon PRAMS, 865 responded to PRAMS-2. The weighted response rate to PRAMS-2 was 51.1%. The unweighted response rate was 44%. Food insecurity at both time periods was assessed with a question on both surveys by asking “During the 12 months before your new baby was born, did you ever eat less than you felt you should because there was not enough money to buy food?” (Time 1) and “In the past 12 months, did you ever eat less than you felt you should because there was not enough money to buy food?” (Time 2). Mothers who answered “yes” to these questions were classified as food insecure for the respective time periods. Having a CSHCN at Time 2 was assessed with a 10-item question about ongoing needs lasting 6 months or more for specific health services: specialty health care, behavioral health or mental health services, physical therapy, occupational therapy, speech services, medication, home health services, special diet, use of assistive devices, or durable medical equipment. Mothers who responded “yes” to any 1 or more of the 10 items were classified as having a CSHCN; this variable was further categorized to represent the number of health services needed: one ongoing need, and two or more ongoing needs. Two logistic regression analytic approaches were applied to study the hypotheses. The first was a cross-sectional analysis using data from the PRAMS-2 survey to examine whether having a CSHCN was associated with self-reported food insecurity. The second was a longitudinal examination of the cohort of women who were food secure at Time 1 to assess whether having a CSHCN was predictive of a shift to food insecurity at Time 2, compared to mothers whose children do not have special health care needs. PRAMS and PRAMS-2 datasets contain weighted data accounting for complex sampling design. STATA 10 was used for analysis of weighted data in this study; all percentages reported are weighted. Results In this sample of mothers of two-year-olds in Oregon, 11.9% were food insecure at Time 1, while 12.8% were food insecure at Time 2. 62 (6.6%) women experienced a shift from food security to food insecurity in the 2-year follow-up period. 38 (5.5%) reported a shift from food insecurity to food security in the follow-up period. Of the two-year-olds in this sample, 11.7% were classified as CSHCN. Cross-Sectional Model: In the multivariate cross-sectional model, having a CSHCN was not significantly associated with food insecurity (OR for one ongoing needs= 1.23, 95% CI: 0.31 – 4.82; OR for two or more ongoing needs = 1.86, 95% CI: 0.49 – 7.06). Low annual household income (OR for income less than $15,000 = 28.98, 95% CI: 4.07 – 206.54; OR for income $15,000 to $24,999 = 19.7, 95% CI: 3.29 – 118.03; OR for income $25,000 to $34,999 = 13.73, 95% CI: 3.47 – 54.32) being American Indian/Alaska Native (OR = 2.32, 95% CI: .099 – 5.47), and maternal education (OR for 12 years of education/GED = 2.45 (1.04 – 5.79) were significantly associated with food insecurity in this model. Longitudinal Model: In the longitudinal model, having a child with ongoing needs for two or more health services was significantly predictive of a shift to food insecurity in the two-year follow-up period (OR = 6.50, 95% CI: 1.71 – 24.74; p = 0.006) after adjusting for covariates. Other factors in this model which were significantly predictive of a shift to food insecurity include: never being married (OR = 3.8, 95% CI: 1.20 = 12.02), being unemployed and looking for work (OR = 6.32, 95% CI: 1.43 – 27.89), being American Indian/Alaska Native (OR = 3.81, 95% CI: 1.07 – 13.52), and low household income (OR for income less than $15,000 = 137.26, 95% CI: 6.61 – 2849.47; OR for income $15,000 to $24,999 = 47.63, 95% CI: 3.57 – 635.02; OR for income $25,000 to $34,999 = 99.04 (8.11 – 1209.82). Discussion While the associations between food insecurity and having a CSHCN were not statistically significant in the cross-sectional analysis, a trend of increasing odds of food insecurity was observed as the number of ongoing health service needs increased. The longitudinal model in this study provides support that having a child with ongoing need for two or more health services at Time 2 is predictive of a shift to food insecurity from Time 1 to Time 2. This study also identifies other risk factors for food insecurity, including marital status, low household income, young maternal age, being unemployed and being of American Indian/Alaska Native descent. This information will help guide clinical programs and public health interventions aimed at preventing food insecurity. This preliminary evidence identifies a unique and particularly vulnerable population for screening and intervention, and provides support for the importance of implementing routine food security screening by health care providers. Future longitudinal research is needed to further identify risk factors that are predictive of a shift to food insecurity over time. A major strength of this study was in the use of both a cross-sectional and a longitudinal analytic approach to study the association between maternal food insecurity and having a 2-year-old CSHCN. The PRAMS-2 survey question used to identify CSHCN in this study was a potential limitation.

Identifier

doi:10.6083/M49G5JS8

School

School of Medicine

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.