Post on 01-Jun-2020
Psychobiological Processes: Pathways Linking Social Factors
with Disease
Andrew SteptoeDepartment of Epidemiology and Public Health
University College Londonhttp://www.ucl.ac.uk/psychobiology/
Department of Epidemiology and Public Health
• Whitehall and Whitehall II epidemiological studies
• English Longitudinal Study of Ageing (ELSA)• Health Survey for England• National Child Development Study (1946 birth
cohort)• Psychosocial factors in Eastern Europe
(HAPIEE)• Ethnic minority psychiatric illness rates
(EMPIRIC) study
Social structure
Social Environment
Work
HealthBehaviors
Pathophysiologicalchanges
Organ impairment
Well-beingMortalityMorbidity
Genes
Early Life
Culture
BrainNeuroendocrine, autonomic and
immune responses
Psychological
Material factors
Issues:
• What are psychobiological processes?
• Why are they relevant?
• What types of study?
• What measures are useful in health care research?
Psychobiological pathways
The pathways through which psychosocial factors stimulate biological systems via central nervous system activation of autonomic, neuroendocrine and immunological responses
Psychobiological responses
Neuroendocrine cortisol, epinephrine, testosterone,norepinephrine
Cardiovascular Blood pressure, heart rateInflammatory C-reactive protein, interleukin (IL-6)
fibrinogenMetabolic Lipids, glucose, insulinHemostatic Platelets, coagulation factorsImmune Lymphocyte counts and activity,
natural killer cells, immunoglobulins
Hypothalamic-Pituitary-Adrenal
Axis (HPA)
Hypothalamus
Pituitary gland
Adrenal cortex
CRF
ACTH
Cortisol
Sympatho-AdrenalAxis
BrainHypothalamus
Adrenal medulla
Epinephrine
Sympathetic nervous systemNorepinephrine
Some effects of cortisol
• Stimulation of glucose production in the liver• Release of free fatty acids from fat stores• Regulation of water balance• Stimulation of anti-inflammatory responses• Immune regulation
Some effects of sympathetic activation
• Increased blood pressure and heart rate• Reduced gut motility and salivation• Stimulation of clotting processes• Acute immune activation• Release of free fatty acids from fat stores
When are psychobiological responses hazardous?
• Heightened reactions or failure of post-stress adaptation
Biological stress responsivity
• Size of the response
• Rate of recovery
0
5
10
15
20
25
0
5
10
15
20
25
Origin of variation in psychobiological responses
• Genetic factors• Early life experience• Adult life factors – adversity, temperament• Appraisal and coping factors
Gianaros,Psychosom Med
2005
When are psychobiological responses hazardous?
• Heightened reactions or failure of post-stress adaptation
• Sustained or repeated exposure to conditions eliciting responses
McEwen,NEJM
1998
Some effects of high cortisol
Potentially damaging effects• Increased lipid (LDL-cholesterol) in the blood• Suppression of immune function• Decalcification of bone• Deposition of abdominal fat• Damage to the hippocampus• Muscle wasting• Impaired reproductive function
Issues:
• What are psychobiological processes?
• Why are they relevant?
• What types of study?
• What measures are useful in health care research?
Psychosocial factors associated with disease incidence or progression in prospective observational cohort studies
• Chronic life stressHigh demand/low control at work; effort-reward imbalance; financial strain; marital conflict; caregiving
• Social environmentSocial isolation; emotional support
• Psychological factorsDepression, anger/hostility, anxiety/distress
Khot et al, 2003JAMA
Work stress and cardiac mortality
0
0.5
1
1.5
2
2.5
3
Low Inter
mediat
e
High
Low
Intermed
iate
High
Haz
ard
Rat
io
25 year follow-up, adjusted for age, sex, smokingPhysical activity, blood pressure, cholesterol, body mass Kivimäki, 2002
high demands /low control
Effort /rewardimbalance
Metabolic syndrome markers (ATPIII)
Three or more of:
Waist circumference > 102 cm (men) or 88 cm (women)Fasting triglyceride ≥ 150 mg/dlFasting HDL-cholesterol < 40 mg/dl (men), < 50 mg/dl (women)Blood pressure ≥ 130 / 85 mmHgFasting glucose ≥ 110 mg/dl
0
1
2
3
4
Satisfied Moderately satisfied Dissatisfied
Odd
s ra
tio
11.5 year follow-up adjusted for baseline MS, age, race, educationSmoking, physical activity, alcohol, depression, anxiety
Marital satisfaction and the metabolic syndrome
Troxel et al, 2005Arch Intern Med
Social support and CHD
0
0.5
1
1.5
Low Inter
mediat
e
High
Low
High
Haz
ard
ratio
15 year follow-up, adjusted for smoking, blood pressure, cholesterol,triglycerides, BMI, waist/hip ratio, diabetes, family history, social class, stress Rosengren, 2004
Social integration Emotional attachment
Psychobiology and Infection
Infectiousagent
Adequateimmune response
Noillness
Infectiousagent
Impairedimmune response
Illness
Psychobiologicalstimulation
1 - 3 4 - 5 6 +
30
40
50
60
# of High-Contact Roles
% w
ith C
olds
OR=4.23*
OR=1.87*
OR=1.00
(Odds ratios are adjusted for control variables.)
Cohen et al, JAMA, 1997
Depression and physical illness
• Premature mortality• Coronary heart disease• Diabetes• Disability• Multiple sclerosis• Rheumatic conditions• Metabolic syndrome
Depression and Physical IllnessCUP, in press
Socioeconomic position and psychosocial factors
Lower socioeconomic position associated with• Greater exposure to chronic stress (domestic,
neighborhood)• Greater financial strain• Lower job control, higher job demands, effort-reward
imbalance• Greater social isolation and lower social support• Lower sense of control• Greater hostility• Less use of active coping and planning responses• More depressed mood, lower optimism
PsychosocialFactors
AdverseWork stress
NeighbourhoodFinancialDomestic
ProtectiveSocial networks
Coping responsesSelf-esteem
Psychobiologicalprocesses
Psychologicalwell-being
Physicaldisease risk
Affectivedisorders
Health-compromisingbehaviours
Social positionOccupationEducation
Income
Issues:
• What are psychobiological processes?
• Why are they relevant?
• What types of study?
• What measures are useful in health care research?
Applications of Psychobiological Measures
• Epidemiological studies
• Clinical studies
• Experimental studies
• Naturalistic monitoring studies
Epidemiological surveys
• Large samples• Disease endpoints• Prospective designs
Fibrinogen
Inflammatory and thrombogenic factorIncreases blood viscosity, stimulates
atherogenic cell proliferation, platelet aggregation
Associated with:• Low SES• Low control at work, high effort/reward
imbalance• Social isolation
Epidemiological surveys
• Large samples• Disease endpoints• Prospective designs
Single measures under resting conditionsLimited information about mechanismsConfounding issues in psychosocial studies
Experimental studies
• Stress-inducing tasks:Problem solving, emotional interviews, simulated public speaking
• Measurement of:Blood pressure, heart rate, hormones, cholesterol, blood clotting, inflammation, muscle tension
• Comparison of:Clinical groups;People high or low on psychosocial characteristics;Experimental conditions inducing low or high control, social support, etc
Laboratory mental stress testing
• Sophisticated biological measures• Control of confounders• Experimental manipulation of conditions
C-reactive protein
• Acute phase protein synthesized in liver• Increases in response to inflammatory stimuli
(cytokines), infection and tissue damage• Antimicrobial, clears apoptotic cells, enhances
phagocytosis• Marker of chronic low grade inflammation• Risk marker for CVD, functional significance
debated
Work stress study
• 105 healthy nonsmoking men, mean age 32.1 years. Full-time employment
• Measures of effort-reward imbalance• Responses to simulated public speaking
and mirror tracing tasks• Blood samples for C-reactive protein
before and after tasks
0
0.025
0.05
0.075
0.1
0.125
Low High
Tertile of Effort-Reward Imbalance
Log
CR
P ch
ange
Adjusted for baseline C-reactive protein, age and BMI
Effort-reward imbalance and C-reactive protein stress responses
Psychobiological responses and SEP
Participants• 238 members of the Whitehall II (prospective) cohort aged 47-59
years in full-time employment.Sampled by grade of employment:
Higher Men 49 Women 41 Total 90Intermediate Men 44 Women 37 Total 81Lower Men 36 Women 31 Total 67
Conditions• Cardiovascular, neuroendocrine, cytokine and hemostatic
responses to colour/word and mirror tracing tasks. Blood drawn at baseline, immediately post-task, and 45 minutes later.
Occupational grade and education
0
20
40
60
80
100
Higher Intermediate Lower
BasicHigh schoolDegree
Stress ratings by SEP
1
2
3
4
5
6
7
Base CW MT Recovery
HigherIntermediateLower
Factor VIII, Stress, and SEP
120
122
124
126
128
130
132
134
136
Baseline Stress 45 min
IU /
dl
Lower
Higher
Intermediate
Steptoe et alThromb Haemo, 2003
Laboratory mental stress testing
• Sophisticated biological measures• Control of confounders• Experimental manipulation of conditions
Small acute responsesGeneralisation of acute responses into everyday lifePrediction of future health risk
Blood pressure reactivity and atherosclerosis
Jennings et al2004Circulation
Whitehall Psychobiology Follow-up
• Assessment of 209 participants 3 years after mental stress testing (92% response rate)
• Measurement of resting blood pressurefasting lipid profilesBMI, waist and hip circumferenceAmbulatory BP in a subset (153)
3 year systolic BP increase ≥ 5 mmHg
0
1
2
3
4
5
SBP recovery DBP recovery
Good recoveryMediumPoor recovery
Adjusted for baseline BP, age, gender, grade of employment, hypertensive medication, BMI, and smoking
Adj
uste
d od
ds ra
tio
Steptoe & MarmotJ Hypertension, 2005
5.2
5.4
5.6
5.8
Base Stress
mm
ol /
l
Adjusted for socioeconomic status, age, body weight, smoking, alcohol, and haematocrit
Cholesterol response to acute stress
0
0.25
0.5
0.75
1
Low High
Tertile of cholesterol stress response
Cha
nge
in c
hole
ster
ol (m
mol
/l)
Adjusted for gender, socioeconomic status, age, body weight, smoking, alcohol, and baseline cholesterol
Cholesterol stress response and cholesterol 3-years later
Steptoe & BrydonHealth Psychol, 2005
Applications of Psychobiological Measures
• Epidemiological studies
• Clinical studies
• Experimental studies
• Naturalistic monitoring studies
Naturalistic / ambulatory monitoring studies
• Dynamic responses in everyday life• Covariation of biology, events and
reactionsAmbulatory blood pressureSalivary cortisol
120
122
124
126
128
130
Morning Mid-day Afternoon Evening
mm
Hg Low job
control
High jobcontrol
Adjusted for gender, age, occupational grade,smoking, bmi, and physical activity
Ambulatory systolic pressure: working day
Steptoe & WillemsenJ Hypertension, 2004
8.00
-8.3
0
10.0
0-10
.30
12.0
0-12
.30
14.0
0-14
.30
16.0
0-16
.30
18.0
0-18
.30
20.0
0-20
.30
22.0
0-22
.30
0
4
8
12
16
20
Cor
tisol
in n
mol
/l Low job control
High job control
Cortisol and job control
Men, age-adjusted Kunz-Ebrecht et alSoc Sci Med, 2004
Naturalistic / ambulatory monitoring studies
• Dynamic responses in everyday life• Covariation of biology, events and
reactions
Limited range of biological measuresControl for confounders
6 year risk of stroke in older men and women
Variable• Positive affect• Negative affect• Age• Education• Smoking• Previous heart attack• Diabetes• Blood pressure
Relative risk / p0.74 (0.62-0.88) 1.01 (0.97-1.05)0.0040.080.0080.0020.0010.001
Ostir et al, Psychosom Med, 2001
10 year mortality in older men and women
Variable• Age ≥ 75• Male• Unmarried• Low education• Cognitive impairment• Functional impairment• Low positive affect
Conditional hazard ratio1.78 (1.61-1.96)2.12 (1.91-2.36)1.30 (1.17-1.50)1.22 (1.10-1.35)1.64 (1.45-1.87)2.07 (1.87-1.28)1.12 (1.05-1.18)
Blazer & Hybels, JAGS, 2004
Psychobiology and Happiness
Aims• To assess associations between
happiness and psychobiological responses relevant to health and disease risk
• To determine whether associations of psychobiological responses and happiness are independent of negative emotional states
Measurement of happiness
• Repeated sampling every 20 minutes from morning (07:30 – 09:30 start) until evening (22:30) on a working day
• Happiness on 5-point scale:1 = very low to 5 = very high
• Division into 1-3 vs 4-5• Proportion of happy ratings (4-5) over
the day
Happiness in men and women
0
20
40
60
80
100
1 Low 2 3 4 5 High
Happiness quintiles
% h
appy
ratin
gs
Men Women
Steptoe et alPNAS, 2005
Salivary cortisol – working day
5
6
7
8
9
10
1 Low 2 3 4 5 High
Happiness quintiles
nmol
/l
8 samples (08:00 – 22:30)Adjusted for gender, age, occupational grade, smoking, bmi, and GHQ
P = .009
Fibrinogen stress responses
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
1 Low 2 3 4 5 High
Happiness quintiles
g / L
Adjusted for gender, age, occupational grade, smoking, bmi, haematocrit, baseline fibrinogen, and GHQ
P = .003
Happiness ratings - 3 years
0
20
40
60
80
100
1 Low 2 3 4 5 High
Happiness quintiles (baseline)
% h
appy
ratin
gs
Men Women
Cortisol and happiness – 3 year
4
5
6
7
8
1 Low 2 3 4 5 High
Happiness quintiles
nmol
/l
N = 144Adjusted for gender, age, occupational grade, work at follow-up, smoking, bmi, GHQ
P = .064
Systolic BP and happiness – 3 year
115
120
125
130
1 Low 2 3 4 5 High
Happiness quintiles
mm
Hg
Adjusted for gender, age, occupational grade, work at follow-up, smoking, bmi, GHQ. N = 160
P = .030
Applications of Psychobiological Measures
• Epidemiological studies
• Clinical studies
• Experimental studies
• Naturalistic monitoring studies
Psychobiological measures:
• Cortisol
• Inflammatory markers: IL-6
• Heart rate variability
Cortisol:
• General marker of activation and associated with health outcomes
• Related to physical health outcomes, mental health (depression) and cognitive function (memory)
• Influenced by current and earlier life psychosocial factors
• Easy to measure in saliva, urine or blood
Cortisol and cognition:
• Wide variation in natural cortisol levels• In aging, some increase, others
decrease cortisol levels• High increases in natural levels with
aging associated with impaired memory, smaller hippocampal volume
Lupien, PNEC, 2005• General cognitive function associated
with higher cortisol in older menMacLullich, PNEC, 2005
7
8
9
10
11
Lower Medium HigherTertiles of Cortisol Reactivity
VPA
-II M
emor
y Sc
ore
Acute cortisol responsivity and memory in old age
Adjusted for age, gender, education, chronic illness,medication, baseline cortisol and time of day
Wright et al, PNEC, 2005
Problems of interpretation ofcortisol data:
• Variation over the day; single readings of limited value
Wak
e
+30
min
8.00
-8.3
0
10.0
0-10
.30
12.0
0-12
.30
14.0
0-14
.30
16.0
0-16
.30
18.0
0-18
.30
20.0
0-20
.30
22.0
0-22
.30
0
5
10
15
20
25
30
35
Cor
tisol
in n
mol
/lCortisol profile over working day
Problems of interpretation ofcortisol data:
• Variation over the day; single readings of limited value
Cortisol awakening response
Levels over the day, slope
Schulz et al, 1998, Stress Med
Cortisol waking response
10
15
20
25
30
35
Wake +30m Wake +30m
nmol
/l
Higher SEPLower SEP
Work day Weekend
Kunz-Ebrecht et alPNEC, 2004
Wake +10 +20 +30 +6010
15
20
25
30
nmol
/l
Lower SEP
Higher SEP
Cortisol waking response
Wright & SteptoePNEC, 2005
Cortisol waking response
Heightened response• Depressive symptoms (Pruessner 2003)• Patients with clinical depression (Bhagwagar
2003)• Chronic work stress (Steptoe 2000)• Working vs weekend days (Schlotz 2004) • Loneliness (Steptoe 2004, PNEC)• Abdominal adiposity (Steptoe, 2004, IJO)
Problems of interpretation ofcortisol data:
• Variation over the day; single readings of limited value
• Which is bad – high or low level?
Buske-Kirschbaum et alPsychosom Med, 2003
Cortisol waking response
Heightened response• Depressive symptoms (Pruessner 2003)• Patients with clinical depression (Bhagwagar 2003)• Chronic work stress (Steptoe 2000)• Working vs weekend days (Schlotz 2004) • Loneliness (Steptoe 2004, PNEC)• Abdominal adiposity (Steptoe, 2004, IJO)Reduced response• Chronic fatigue (Roberts 2004)• Some physical illness groups (Kudielka 2003)
Cortisol and recent trauma:
• 190 low income women from urban Michigan• Samples in clinic, on waking, 30 min after
waking, and before bed• 12 month exposure to trauma (assault, rape,
car crash, sudden death of close friend etc)
Young et al, Biol Psychiatry, 2004
Cortisol and recent trauma:
• 163/190 clinic• 170/190 waking• 173/190 second morning• 166/190 bedtime samples• Second morning sample very late in
more than a third• Pooled data used
Inflammatory markers
• C-reactive protein• Interleukin 6 (IL-6)• Fibrinogen• Tumor necrosis factor alpha
Interleukin 6
An ‘endocrine’ cytokine associated with• Coronary heart disease• Type 2 diabetes, insulin resistance, obesity• Depression • Disability • Cognitive decline
0
10
20
30
40
Low HighQuartiles of IL-6
% h
igh
CES
DIL-6 and depressed mood
3024 men & women aged 70-79Penninx et al, Biol Psychiat
2005
IL-6 and cognitive impairment
4 year follow-up of 2632 men and women aged 70-79 years
Cognitive impairment, adjusting for baseline cognitive score, age, education, race, depression, alcohol, stroke and statins
• High inflammation RR: 1.66 (1.19 – 2.18)• Low inflammation RR: 1.08 (0.89 – 1.30)
Yaffe et alJAMA 2004
Interleukin 6
An ‘endocrine’ cytokine associated with• Coronary heart disease• Type 2 diabetes, insulin resistance• Depression • Disability
• Sensitive to acute and chronic stress
Caregiver stress and plasma interleukin-6
Kiecolt-Glaser et alPNAS, 2003
Base 30 min 75 min 120 min0.50
1.00
1.50
2.00
2.50
3.00
pg/m
l
Lower SES
Higher SES
No stress
Brydon et alBrain, Behav Immun, 2004
Stress and plasma interleukin-6
Heart rate variability
• Beat-to beat variation in heart rate• Interplay between sympathetic and parasympathetic
(vagal) branches of autonomic nervous system
Assessed in the • Time domain (R-R variability)• Frequency domain (spectral analysis)
High frequency (parasympathetic)Low frequency (sympathovagal balance)Low/High (sympathovagal balance)
Low heart rate variability
• Higher risk of death or recurrent events in patients with CHD (Atrami study, 1998)
• Incident CHD in apparently healthy cohorts (Liao, 1997)
• Future hypertension (Schroeder, 2003)
• Post-stroke mortality (Makikillio, 2004)
• Depressive symptoms (Lim, 2005)
Low heart rate variability
• Poorer cognitive executive function (Hansen et al, 2003)
• Less effective impulse control in children (Allen et al, 2000)
• More negative moods in alcohol abuse (Ingjaldsson et al, 2003)
• Reduced sleep efficiency (Hall et al, 2004)
Whitehall autonomic function study
Low heart rate variability related to:
• Low employment grade• Lower job control• Smoking, inactivity, high alcohol intake• Metabolic syndrome
Hemingway et alCirculation, 2005
Low heart rate variability
• Social isolation (Horsten, 1999)
• High effort/reward imbalance (Vrijkotte, 2000)
• Depressive symptoms (Lim, 2005)
• Depression in post-MI patients (Carney, 2001)
Base Tasks Recovery15
20
25
30
35
40
RM
SSD
(mse
c)
Higher
Lower
Heart rate variability by grade
PsychosocialFactors
AdverseWork stress
NeighbourhoodFinancialDomestic
ProtectiveSocial networks
Coping responsesSelf-esteem
Psychobiologicalprocesses
Psychologicalwell-being
Physicaldisease risk
Affectivedisorders
Health-compromisingbehaviours
Social positionOccupationEducation
Income
Conclusions
• Psychobiological processes are plausible mediators of social influences on disease development
• The involvement of psychobiological responses is determined by the interplay between individual differences in responsivityand exposure to positive and negative life experiences
• Some psychobiological responses appear to be relevant across a range of health outcomes
What we need to know
• The precise pathological significance of some psychobiological responses is poorly understood
• The full sequence from social factor to biological response to disease has seldom been documented
• Whether interventions along this sequence will block health effects and demonstrate causality
Acknowledgements
• University College London:Sir Michael Marmot, Professor Jane Wardle, Dr Sabine Kunz-Ebrecht, Dr Pamela J Feldman, Dr Natalie Owen, Dr Lena Brydon, Dr Mark Hamer, BevMurray, Caroline Wright
• University of DresdenProfessor Clemens Kirschbaum
• University of GlasgowProfessor Gordon Lowe, Dr Ann Rumley
Medical Research CouncilBritish Heart Foundation
Early life stress and adult cortisol stress responses
0
100
200
300
400
500
600
nmol
/l
Control ELS / Nodepression
No ELS /depression
ELS /depression
Heim et al, JAMA, 2000
ELS – history of childhood physical or sexual abuseDepression – current major depressive disorder