Lecture 3 - Research models in psychopathology Flashcards
Longitudinal designs
Retrospective
*Collect a sample of
people with a disorder
*Try to determine what
preceded it
-Self-report. Problems with self-report for it: recall issues, current mood states (memory biases)
-Existing archival data. Problems with archival data: not everybody has records (home videos, school records..)
Follow-up: natural course of a disorder
*Follow people with the
disorder over time
*See what happens to
them
*Already-ill sample
*Difficult to derive
etiological explanations
High Risk
*Variant of follow-up
* Identify people who are
likely to develop a
disorder
-Offspring of people with
a disorder (genetic)
-On the basis of a
biological abnormality
-Behavioral variable
*Follow them over time
High Risk Studies: Cons
*Genetic: Need to find
people who have the
disorder and also have
children
*Biological: associations
not well-proven
*Behaviors: May be a risk
factor, or may be early
manifestation of the
disease
Limitations: not representative sample (ex: most schizophrenic people don’t have kids), see others
Vulnerability Marker
*Should be trait-like, not
state-related
*Has to be correlated with
the disorder, but has to
persist beyond the end of
the episode
-Could be a scar
*Has to be present in a
high-risk population
*Pre-date disorder
Sampling issues
Case Control v. Cohort
*Case Control: compare
one group of people with
disorder to a second
group without the
disorder
*Cohort: a single large
sample of people, some
of whom have the
disorder
Patients vs. Community
*Patient populations not
representative of people
with the disorder in the
community
*Clinical populations (patients) tend to
be more severe, have more
comorbidities, more likely
to be female, chronic
*General population, get a
sense of disorder “in the
wild”
*Very expensive
Why patient populations (clinical) less representative… higher SES, more women, more severe
Community: Milder symptoms, closer to threshold, chance of misdiagnosis
Controls?
*Healthy Controls (HC) or
Psychiatric Controls
(PC)?
-from psychiatric population other than the one you’re studying (if studying depression, could use anxiety or schizo people as control)
-Healthy controls can sometimes be not representative, since most people have met criterias for a psychopathology
*Match on potential
confounds?
-Potential confounds: May less stressors, may higher SES, may more support….
*How do you match on lab
tasks?
Genetic epidemiology: Family Studies
*First step:
* Identify proband: someone who has the disorder
* Assess family members
- Interview (Family Study)
- Informant report (Family
History Study)
*Many disorders do run in
families
*Subthreshold (close to diagnosis, but not enough to meet diagnosis criterias)/ symptoms run in family too
*Coaggregation: aggregate of disorders… higher rates of simultaneous diseases in family than in general population
*Suggest genetic role,
does not prove it (shared environment, experiences,…)
Rates of illness in family must be higher than what we see in the general population or family members of healthy controls
Genetic epidemiology: Adoption studies
*Parent as proband: parent with disorder who gave away child… then see if adopted child also have
*Adoptee as proband: see if adoptive and bio parents of adoptee (with disorder) more likely to have pathology
*Cross-fostering design: kids with low risk adopted into household with pathology + Kids with pathology risk adopted into family with low risk
-Adoption rare event
-Selective placement: Difficult because adoptive parents very carefully screened before adoption (usually can’t adopt if known mental issues)
Genetic epidemiology: Twin studies
*Monozygotic = identical (one egg and one sperm combined and split)
*Dizygotic = fraternal (share 50% of genes)… two eggs fertilized
*Can control for environmental influences
*A= Additive genetic component: how much more alike are monozygotic twins than dizygotic twins… heritability estimate
*Heritability estimate : 0 (no) to 1 (absolutely heritable)
*C= Common environment component= environmental factors makes similar… because siblings, share environment… how much due to common environmental factors makes them similar
*E =Unique environment : how much differ due to unique environmental factors/experiences
*A= 2(rMZ – rDz)
*MZ concordance = 50%
*Dz concordance = 25%
*Difference (D)= 25%
*2D = 50%
*Sample specific
*Higher with less
environmental variance
*Slide 49 = 50% of variance between MZ and Dz explained by A (additive genetic component)
Problems with Twin studies
*MZ twins often share
placenta = means environment (even from gestation) more similar than Dz (unlike what is assumed in twin studies)
*MZ twins treated more
similarly to one another = Assumption that environment of mono same as dizy… but mono usually more similar environment experiences (because often treated same due to same appearance)
*Heritability = estimated
genetic contributions to
observed phenotype
*Not deterministic
*Often don’t model G x E = Genes x Environment interaction
Gene-Environment correlations
*Passive: typically parented by people who gave us our genetic material, does not depend on what a child does
-Can be addressed in adoption studies
*Active (niche-picking): people are not passive receptacles of our envir, gravitate towards/choose environments depending on genetic makeup… selecting into environment based on genetic makeup
*Evocative (reactive): treated differently based on our genetic makeup… can determine the response we receive…. Treated differently depending on ex: height, attractiveness… bring about people’s responses to us
-Active and evocative hard to measure– need better
understanding of how environment shapes traits
*Currently, all rGE attributed to G
*Caution when interpreting genetic contributions
Paradox of Intelligence
* IQ highly heritable: 80%
(approx.): But heritability is very variable (development, environment)…. Not an absolute
* IQ also malleable
-Flynn effect
–Developing countries
*Higher IQ = seek out
“more stimulating” env
– More stimulating
environments available
w/ more development
Heritability varies as a function of environment
* Heritability of IQ increases across
development (Bouchard, 1997; Haworth et al., 2009).
* Among affluent families, heritability of IQ
estimated at 0.72.
- Among less affluent families, little observed
additive genetic influence (heritability of 0.10;
Turkheimer et al., 2003)
* Heritability of alcohol use for those residing
in a neighborhood with ten or more alcohol
outlets was 74%, compared with 16% for
those in a neighborhood with zero outlets
(Slutske et al., 2018)
* Eating disorder symptoms show minimal
heritability before puberty but significant
genetic effects (i.e., greater than 50%)
during and after puberty (Klump et al., 2003;
Klump, Perkins, et al., 2006).
Mode of Transmission (see if this and the following slides were actually in the lecture… i don’t remember them)
Quantitative Genetics
Single-Gene
Transmission
Problems with Single-Gene:
*Single dominant gene,
would expect 50% of
relatives to have disorder
*No psychiatric disorders
show these rates of
familial transmission
*Possible single-gene
transmission
-Expressed differently in
different relatives
*Mendelian disorders v.
rare
-Prevalence about 1 in
10,000
*Most psychiatric
disorders have
prevalence rates at least
.5%
*Monogenic disorders–
distinct from “normal”
-E.g., Parkinson’s
*Psychiatric disorders
dimensional,
continuously distributed
Polygenic transmission
*Psychological
phenotypes likely
controlled by more than
1 gene
-Skin color– at least 3
genes
-Can be modified by GxG
interactions
-GXE interactions
*Action of multiple genes
-Additive or interactive
effects
Missing Heritabilities:
*Big Five personality traits have heritability estimates of
0.40 to 0.60 (e.g., Bouchard, 2004),
*Autism spectrum disorder currently estimated at 0.38
(e.g., Hallmayer et al., 2011),
*Schizophrenia at 0.64 (e.g., Lichtenstein et al., 2009)
*Estimates so high, explanations provided in quantitative
genetics explains very little
*Where do we get the remaining genetic variation?
Mixed Transmission
Gene Environment Interactions
*Notion that adverse effects of
genes on mental health only
expressed under certain
environmental conditions
*Influence of life stress on
depression: moderation by a
polymorphism in the 5-HTT
gene
Epigenetics:
*Regulation and expression of
genes
*DNA
*Action of genes can be
regulated
*Some genes “turn on” at
certain developmental
periods
*Under certain environmental
circumstances
*Alterations heritable
Michael Meaney
*Good rat mothering associated
with better functioning of
neuroendocrine stress response
*Bad rat mothering= high levels
of stress and cortisol
*Cross-fostering
*Changes in glucocorticoid
receptor gene
*Only evident when switch
occurred early
Genotypes-Endophenotypes-Phenotypes
- Genoptype → Phenotype
- Phenotypes very complex phenomena, multiply-determined, often poorly
defined - Endophenotype: intermediate step between microscopic genes and nerve
cells and the experiential and psychological phenotype
-must segregate with illness in the population. - must be heritable.
- must not be state-dependent (i.e., manifests whether illness is active or in
remission). - must co-segregate with illness within families.
- must be present at a higher rate within affected families than in the population.
- must be amenable to reliable measurement, and be specific to the illness of
interest.
