sociodemographic shifts influencing black white disparity in adl and iadl disabilities n.
Download
Skip this Video
Loading SlideShow in 5 Seconds..
Sociodemographic Shifts Influencing Black-White Disparity in ADL and IADL Disabilities PowerPoint Presentation
Download Presentation
Sociodemographic Shifts Influencing Black-White Disparity in ADL and IADL Disabilities

Loading in 2 Seconds...

play fullscreen
1 / 22

Sociodemographic Shifts Influencing Black-White Disparity in ADL and IADL Disabilities - PowerPoint PPT Presentation


  • 101 Views
  • Uploaded on

Sociodemographic Shifts Influencing Black-White Disparity in ADL and IADL Disabilities. Shih-Fan (Sam) Lin slin@projects.sdsu.edu Brian K. Finch bfinch@mail.sdsu.edu Audrey N. Beck abeck@projects.sdsu.edu San Diego State University. Objectives. Black-White Disparity in Disability

loader
I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
capcha
Download Presentation

PowerPoint Slideshow about 'Sociodemographic Shifts Influencing Black-White Disparity in ADL and IADL Disabilities' - todd


An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
sociodemographic shifts influencing black white disparity in adl and iadl disabilities

Sociodemographic Shifts Influencing Black-White Disparity in ADL and IADL Disabilities

Shih-Fan (Sam) Lin

slin@projects.sdsu.edu

Brian K. Finch

bfinch@mail.sdsu.edu

Audrey N. Beck

abeck@projects.sdsu.edu

San Diego State University

objectives
Objectives
  • Black-White Disparity in Disability
    • Examine the black-white disparity trends in late-life disability with Age-Period-Cohort models using the cross-classified random effects model.
      • Socio-demographic and A-P-C adjusted disparity trends.
      • Stratify trends by gender.
    • Elucidate the role of aging and race in disability.
    • Decompose the adjusted and cohort-based disparity trend for ADL and IADL disabilities to assess the relative contribution of each sociodemographic control variable.

INTRODUCTION

age period cohort
Age, Period, & Cohort
  • Time can be captured by three unique temporal dimensions: Age, Period, and Cohort. Each aspect of A-P-C has a unique contribution to the study of population health including disability.
  • Most of previous studies in disparity in disability only focused on period-based or age-based trends.
  • There were very few research studies that examined the disability prevalence on another critical temporal dimension – birth cohort.

INTRODUCTION

slide4
Age
  • Age is a proxy for biological processes that ultimately lead to disease, disability, and/or death.
  • Age may also be associated with changes in status, social roles, and social position (Yang and Land 2007).

INTRODUCTION

period
Period
  • Period or survey year, reflects changes in socio-cultural, economic, technological, and environmental factors that may affect the entire population at a given time simultaneously, but perhaps not equally.
    • For example, a drought may lead to increased food prices, which disproportionately affects those with lower incomes.

INTRODUCTION

cohort
Cohort
  • Cohort describes a unique set of individuals who are both born into a social system during a similar time period and experience similar formative social experiences over their life course.
  • The colloquial concept of “generations” is an attempt to capture the unique characteristics of distinct birth cohorts.

INTRODUCTION

tackling the identification problem
Tackling the Identification Problem
  • Identification problems occur when the predicting variables in a regression are linearly dependent.
  • There is an exact linear dependence between age, period, and birth cohort (Period = Age + Cohort).
    • To break the linear dependence, we group cohorts into 5-year bands.
    • For example, individuals who were born between 1883-1887 were collapsed into the 1885 cohort (mid-point).

METHOD

slide8
Data
  • Integrated Health Interview Series (IHIS), 1982-2009
    • IHIS harmonizes the National Health Interview Survey (NHIS) variables to allow consistent coding across each survey year to facilitate temporal analysis.
    • The NHIS is a repeated cross-sectional survey.
      • Purpose: to investigate and monitor the prevalence of important health outcomes (including disability) of the civilian non-institutionalized U.S. population.
  • Inclusion criteria for our study
    • Older adults aged 70 and over.
      • For the 1982-2009 survey period, the disability items were only asked of those aged 70 and older (71 for 1982).
  • NHIS 1982 is the first survey year that the disability status is available.
  • Finalized sample:
    • Consists of 199,001 respondents
      • 35.7% are white men
      • 4.2% are black men
      • 52.9% are white women
      • 7.2% are black women.

METHOD

v ariables
Variables
  • Dependent variables
    • ADLdisability (yes/no)
    • IADL disability (yes/no)
  • Independent variables
    • Age – cohort median centered; entered linearly and as a square term
    • Period – year of interview
    • Cohort – birth year; 5-year grouping
      • Period and cohort variables were not modeled directly, but estimates were obtained via a post-estimation strategy describe in the analysis section.
    • Sociodemographic controls:
      • Dummies: race (black/white), region of residence, marital status, employment status, & BMI
      • Continuous: education (cohort-median centered), & CPI adjusted family income (grand-mean centered).

METHOD

analysis m odel specification
Analysis – Model Specification
  • We fit a logistic- cross-classified random effects model for each disability outcome (ADL and IADL disabilities) and each was stratified by gender.
    • For each disability outcome, 2 sets (men and women) of logistic regression were performed.
  • For each regression:
    • The ADL or IADL disability (yes/no) was regressed on age (cohort-median centered) in linear and squared terms, race (black/white), and a host of sociodemographic control variables in the fixed effect portion.
    • We also specified the random intercept for each period and cohort. Cohort was crossed with each period and vice versa.

METHOD

analysis predicted probability
Analysis – Predicted Probability
  • For each of the four racial-gender subgroups (white men, white women, black men, and black women), the predicted probabilities of ADL/IADL disabilities were calculated separately for each A-P-C dimension while holding the rest of predicting variables at the intercept.
    • For example: The predicted probability of ADL disability for each cohort among black men was calculated while specifying the race variable at 1 (0=white, 1=black), sociodemographic dummy variables at the reference category (married, retired, residence at the northeast region, and normal BMI) and holding years of education at the median value of each cohort and income at the grand mean value.
  • The difference (black minuses white) in predicted probabilities for each age, period, and cohort and for men and women were plotted for both ADL and IADL disabilities.
  • To make more interpretable results of the predicted probability for age, we used the non-cohort-median-centered age and age squared model while other variables remained the same form.

METHOD

analysis decomposition
Analysis – Decomposition
  • We also decomposed the age-period -adjusted cohort-based disparity trend for ADL and IADL disabilities to assess the relative contribution of each sociodemographic control variable using Fairlie’s decomposition method for nonlinear outcomes.

METHOD

key findings for a p c trends
Key Findings for A/P/C Trends
  • Blacks have higher predicted probability of having ADL/IADL disability compared to whites and the racial gaps tend to be larger among women than men for both outcomes.
    • These patterns were observed for age, period, and cohort trends.
  • The age trends show a persistent increase of disparity across age for both disability outcomes (less apparent for men).
    • This supports the “double jeopardy” hypothesis – both aging and minority status contribute to the poor health outcomes (i.e. ADL and IADL disabilities) for blacks at later age.
  • The period-based disparity trends for ADL disparity were flat and the IADL disparity trends were less consistent as there was a small drop of IADL disparity between 1996 and 1999.
    • These trends are the same for men and women.
  • The cohort-based ADL and IADL disparities declined continually across each successive cohort for both genders.
    • Gender differences in ADL and IADL disparities also substantially decrease with each successive cohort.
    • The persistent black-white disparities for both types of disability seem to almost disappear in the most recent cohort (1940).

RESULTS

sociodemographic contributions to the cohort based adl disparity trends among women
Sociodemographic Contributions to the Cohort-based ADL Disparity Trends among Women

RESULTS

key findings for decomposition
Key Findings for Decomposition
  • Education and income persistently contributed to the ADL and IADL disparities for men and women; however, education had a more pronounced impact.
    • For example, in the1900 cohort, the ADL racial disparity would have been 53% smaller if black men were to have the same education profile as white men; whereas, the reduction of the disparity would have only been 36% if black men’s income profile were similar to that of white men’s.
  • Racial differences in employment status only explained a very little portion of the ADL or IADL disparity for men and women.
  • Marital status gradually became a more important contributor to the racial disparity, however, this was more salient for IADL disparity.
  • There was a remarkable decline in the importance of region of residence across the cohorts for the IADL disparity among men and women. For the ADL disparity among men and women, there seems to be an initial decline of importance of the region of residence and the trend reversed.

RESULTS

key findings for decomposition1
Key Findings for Decomposition
  • BMI does not play an important role for men regardless of the type of disability gap.
  • Among women, BMI increasingly contributed to the ADL disparity between the 1920-1940 cohorts.
    • These contributions were negative: the ADL disparity between black and white women would have been larger if black women were to have the same BMI distribution as white women.
  • The contribution of BMI for the IADL disparity among women remained quite constant.

RESULTS