Transcript Racial and Ethnic Variability in the Prevalence of Disorders?

Unit 16:
Use of Race and Ethnicity In
Epidemiological Research
Traditionally used
definitions of race and
ethnicity in the U.S.
Race, Ethnicity, and Culture in
the Sociology of Mental Health
• Definitions:
– Race: a socially constructed category
based on observed phenotypic
manifestations of presumed, underlying
genetic differences.
– Ethnicity: a grouping of persons
according to a shared geographic,
national, or cultural heritage.
Encompasses both biological and nonbiological differences.
Research-based Definitions
• Federal Definitions of Race (5 minimum
categories):
–
–
–
–
–
White/Caucasian
Black or African-American
Asian
Native Hawaiian or Other Pacific Islander
American Indian or Alaska Native
• Federal Definitions of Ethnicity:
– Hispanic or Latino
– Not Hispanic or Latino
• Census 2000 Revision: Persons can report more
than one race
Commentary: White, European, Western
Caucasian, or What? Inappropriate labeling in
research on race, ethnicity, and health
(Bhopal R, Donaldson L. AJPH 1998;88:1303-7).
• Historically, much of the debate has been on
proper labeling of minority populations
• Substantial variation in disease rates occurs
by racial and ethnic groups
• Thus, studying race/ethnicity in epi research
may allow better understanding of causes of
disease, particularly relative contribution of
genetic and environmental factors.
Commentary: White, European, Western
Caucasian, or What? Inappropriate labeling in
research on race, ethnicity, and health
(Bhopal R, Donaldson L. AJPH 1998;88:1303-7)
The terms “White” and Caucasian” are
probably too heterogeneous to be of any
scientific value (e.g. for comparisons)
• Ethnicity is slowly replacing race,
although the terms are still often used
interchangeably.
Racial differences: Psychiatry
• Initial problem with defining
race/ethnicity
• In addition, biases in exposure and
outcome classification by race/ethnicity
-- Problems with raters
-- Problems with diagnostic criteria
-- Problems with assessing cultural
differences
Problems with Diagnostic Criteria:
-- Patterns of referral and treatment of
psychosis vary among white and non-white
patients
-- Black patients are more likely to be
diagnosed
with the following:
• Severe first-rank psychotic symptomatology
• Schizo-affective disorders, specifically
schizophrenia
-- Black patients are more likely to be
hospitalized
Racial differences: Results
Symptomatology Index
First Rank Symptoms
(exhibiting a correlation of .30 or higher only)
• Thought insertion
– r=.32
• Thought withdrawal
– .44
• Being controlled
– .32
• Voices commenting
– .57
• Auditory hallucination
– .48
• Visual hallucination
– .39
• Grandiose incoherence
– .35
• Aggressive behavior
– .30
Problems w/Assessing Cultural Differences:
--For many Western medicine diagnoses, cultural
factors are not considered (exception – psychiatry)
“Susto” (“fright” or “soul loss”): A folk illness
prevalent among some Latinos in the U.S. and
among people in Mexico and Central/South America.
Susto is an illness attributed to a frightening event
that causes the soul to leave the body and results in
unhappiness and sickness. Symptoms may occur
from days to years after the fright is experienced,
and include appetite and sleep disturbances, low
motivation and self-worth, muscle aches and pains,
headache, diarrhea. Ritual healings are focused on
calling the soul back to the body and cleansing the
person to restore bodily and spiritual balance.
Do Races Exist?
Ethnicity in Psychiatric Epidemiology:
Need for Precision
• Do races exist?
– Populations do differ biologically in the
distribution of traits of simple inheritance
such as blood groups, abnormal
hemoglobin, human lymphocyte antigen
system, taste sensitivity.
– However, there are no racial typologies in
grouping such traits.
– It is now accepted that racial classification
based on traits such as skin color are
scientifically invalid.
Singh, SP. Ethnicity in psychiatric epidemiology: need for precision.
Brit J of Psych. 1997;171:305-8.
Scientific American, December 2003
• Do races exist?
---If races are defined as “genetically
discrete groups”, then the answer is
NO.
---However, researchers can use some
genetic information to group individuals
into clusters with “medical relevance”.
Scientific American, December 2003
Outward signs of which definitions of
race are based (i.e. skin color and hair
texture) are dictated by a handful of
genes:
-- “Other” genes of 2 people of the same “race”
can be very different.
-- Conversely, 2 people of different “races” can
share more genetic similarity than 2
individuals of the same “race”.
Scientific American, December 2003
The medical implications of racial genetic
differences are still under debate:
-- The FDA (as well as NIH) advocate collection of
race and ethnicity data in all clinical trials.
-- Some investigators assert that genomic data,
rather than self-reported race, should be
collected on each study participant.
-- The extent to which different “races” respond
differently to treatment regimens, and vary in
disease susceptibility remains a matter of
scientific debate.
“Commentary on Why National
Epidemiological Estimates of Substance
Abuse by Race Should Not be Used”…
Kip KE, Peters RH, Morrison-Rodriguez B.
Am J Drug & Alcohol Abuse, 2002; 28(3): 545-56
BACKGROUND
• In the U.S., 3 large scale epidemiological studies
have estimated prevalence of substance use, abuse,
and/or dependence:
--Epidemiologic Catchment Area (ECA): 1980-84:
5 community-based sites (n>20,000): structured DIS.
--National Comorbidity Survey (NCS): 1990-92:
National sample of 8,098 non-institutionalized
persons ages 15 to 54: structured CIDI.
--National Longitudinal Alcohol Epidemiologic
Survey (NLAES): 1992: National sample of 42,862
non-institutionalized adults: structured AUDADIS.
BACKGROUND
Epidemiologic Catchment Area (ECA):
•Prevalence rates of alcoholism similar between
Blacks and Whites, except in age group 18
to 29
(50% lower rates in Blacks).
•Lifetime prevalence of drug dependence 50%
lower in Black women ages 18-29 vs. White
women
•Indicators of SES (education, income) generally
inversely associated with prevalence of
alcoholism and drug abuse/dependence.
BACKGROUND
National Comorbidity Survey (NCS):
•Blacks estimated to have 65% lower odds of
substance use disorder in their lifetime vs.
Whites, and 53% lower odds in the past 12
months.
•Among persons with lifetime drug dependence,
Blacks 3 times more likely to be dependent in the
past 12 months vs. Whites.
•Education and income inversely associated with
1-year prevalence of substance use disorder, and
in particular, lifetime drug dependence.
BACKGROUND
National Longitudinal Alcohol Epidemiologic
Survey (NLEAS):
•Odds of drinking >12 drinks on >12 occasions in
the previous year were 53% lower in Blacks vs.
Whites.
•Among persons with lifetime alcohol
dependence, Blacks 61% more likely than Whites
to be dependent in the past 12 months.
•Education and income inversely associated with
1-year prevalence of alcohol and drug
dependence.
Discussion Question
What are some of the possible
explanations for epidemiological
differences in substance abuse
disorders by race
(i.e. other than valid, true differences)?
Explanation #1
Social and Family Support Systems:
• Religiosity is inversely associated
with range of substance
use/abuse/dependence.
Explanation #2
Inadequate Sampling in Non-Civilian
Settings:
• NCS and NLAES conducted among
civilian populations only; possible bias
if Blacks are over-represented in
institutionalized settings (prisons,
nursing homes, mental hospitals).
Explanation #3
Ethnic and Diagnostic Instrument Bias:
• Perhaps the structured instruments used
have differential reliability and validity
among Black vs. White individuals.
Explanation #4
Possible Reporting Bias:
• Perhaps the instruments per se are not
invalid, but self-reporting of substance
use is biased.
Explanation #4
Possible Reporting Bias:
• Several studies have noted that Blacks
tend to be more likely than other racial
groups to underreport drug use (e.g.
heightened suspicion or fear of possible
adverse consequence from admitting
illicit drug use).
Explanation #4
Possible Reporting Bias:
• In NCS, no difference by race for anxiety
of affective disorders – perhaps greater
stigma and social disapproval for
reporting substance use problems
(particularly among Black women).
Explanation #4
Possible Reporting Bias:
• Indicators of SES are generally
inversely associated with substance
use disorders – given overall lower SES
of Blacks in U.S., results are counterintuitive.
Explanation #4
Possible Reporting Bias:
• In all 3 studies, no apparent attempt to
match race/ethnicity between
interviewer and interviewee.
Explanation #4
Possible Reporting Bias:
• Unclear (not intuitive) as to why Blacks
would be at lower risk of developing a
substance use disorder, but once
dependent, become more persistently
dependent.