Physical comorbidity with bipolar disorder: lessons from UK data
Daniel Smith
Symposium 33: ‘Big data’ and bipolar disorder in the UK
“A failure of social policy and health promotion, illness prevention and care provision.”
Life expectancy at birth of people with mental disorders in the period of 2007–09 (N = 31,719).
Chang C-K, Hayes RD, Perera G, Broadbent MTM, et al. (2011) Life Expectancy at Birth for People with Serious Mental Illness and Other Major
Disorders from a Secondary Mental Health Care Case Register in London. PLoS ONE 6(5): e19590. doi:10.1371/journal.pone.0019590
http://www.plosone.org/article/info:doi/10.1371/journal.pone.0019590
UK data:
• Bipolar Disorder Research Network (n=8,000)
• Scottish Primary Care data (n=1.8 million)
• UK Biobank (n=0.5 million)
• Glasgow Psychosis cohort (n=7,500)
Cardiometabolic disease in BDRN cohort:
0
5
10
15
20
25
Diabetes Hypertension High
cholesterol
Coronary heart
disease
Stroke
Bipolar disorder
Controls
%
Forty et al, in submission
The analysis of ‘SPICE data’ was conducted as part of the Living Well
with Multimorbidity Programme (CSO Grant ARPG/07/1) with
Professor SW Mercer (Principal Investigator) and Professor Bruce
Guthrie (epidemiology lead).
Multimorbidity and major mental illness in Scotland:
– Data from 314 general practices in Scotland (1.8 million people)
– Schizophrenia and related psychoses and bipolar disorder identified (n=12,504)
– 32 physical health conditions also identified
– Multimorbidity described by age, gender and socioeconomic deprivation
– Some prescribing information
Physical health comorbidities assessed:
Coronary heart
disease
Parkinson’s disease Peripheral vascular
disease
Viral hepatitis
Chronic kidney
disease
Multiple sclerosis Sinusitis Liver disease
Asthma Stroke Chronic obstructive
pulmonary disease
Psoriasis/eczema
Atrial fibrillation Blindness Bronchiectesis Irritable bowel
syndrome
Epilepsy Glaucoma Crohn’s disease Migraine
Cancer (any) Hearing loss Diverticulitis Dyspepsia
Thyroid disorders Hypertension Rheumatoid
arthritis
Constipation
Diabetes Heart failure Prostate disease Pain disorder
Prevalence and odds ratios for physical health comorbidity (standardised by age and gender)
Variable Bipolar, n (%) Not bipolar, n (%) Odds ratio (95% CI)
No physical condition 929 (36.0) 799,179 (56.2) 0.59 (0.54 to 0.63)
One physical condition 662 (25.6) 292,651 (20.6) 1.27 (1.16 to 1.39)
Two physical
comorbidities 427 (16.5) 149,297 (10.5) 1.45 (1.30 to 1.62)
Three or more physical
comorbidities 564 (21.8) 180,669 (12.7) 1.44 (1.30 to 1.64)
Prevalence and odds ratios for physical health comorbidity (standardised by age and gender)
Variable Bipolar, n (%) Not bipolar, n (%) Odds ratio (95% CI)
No physical condition 929 (36.0) 799,179 (56.2) 0.59 (0.54 to 0.63)
One physical condition 662 (25.6) 292,651 (20.6) 1.27 (1.16 to 1.39)
Two physical
comorbidities 427 (16.5) 149,297 (10.5) 1.45 (1.30 to 1.62)
Three or more physical
comorbidities 564 (21.8) 180,669 (12.7) 1.44 (1.30 to 1.64)
Differences in prescribing between bipolar and controls for coronary heart
disease (CHD) and hypertension patients, standardised by age and gender.
Patients Bipolar Controls Odds ratio (95% CI)
CHD patients: n = 170 n = 80,985
Aspirin or clopidogrel, % 69.3 73.6 0.81 (0.58 to 1.12)
Statin, % 70.0 74.9 0.69 (0.50 to 0.96)
No antihypertensive, % 29.4 15.8 2.08 (1.49 to 2.91)
One antihypertensive, % 37.6 31.3 1.29 (0.94 to 1.76)
Two or more
antihypertensives, % 32.9 52.7 0.46 (0.33 to 0.63)
Hypertension patients: n = 462 n = 232,986
Statin, % 36.7 41.5 0.82 (0.68 to 0.98)
No antihypertensive, % 21.8 13.9 1.70 (1.36 to 2.12)
One antihypertensive, % 39.8 32.3 1.38 (1.15 to 1.67)
Two or more
antihypertensives, % 37.8 53.7 0.53 (0.44 to 0.68)
Differences in prescribing between bipolar and controls for coronary heart
disease (CHD) and hypertension patients, standardised by age and gender.
Patients Bipolar Controls Odds ratio (95% CI)
CHD patients: n = 170 n = 80,985
Aspirin or clopidogrel, % 69.3 73.6 0.81 (0.58 to 1.12)
Statin, % 70.0 74.9 0.69 (0.50 to 0.96)
No antihypertensive, % 29.4 15.8 2.08 (1.49 to 2.91)
One antihypertensive, % 37.6 31.3 1.29 (0.94 to 1.76)
Two or more
antihypertensives, % 32.9 52.7 0.46 (0.33 to 0.63)
Hypertension patients: n = 462 n = 232,986
Statin, % 36.7 41.5 0.82 (0.68 to 0.98)
No antihypertensive, % 21.8 13.9 1.70 (1.36 to 2.12)
One antihypertensive, % 39.8 32.3 1.38 (1.15 to 1.67)
Two or more
antihypertensives, % 37.8 53.7 0.53 (0.44 to 0.68)
Implications:
• Coronary Heart Disease, Heart Failure, Peripheral Vascular
Disease, Stroke and TIA not more commonly recorded in the bipolar group
• Where cardiovascular diseases were recorded for the bipolar group, evidence of less intensive treatment
• Substantial treatment inequalities for bipolar patients with
coronary heart disease and hypertension.
UK data:
• Bipolar Disorder Research Network (n=8,000)
• Scottish Primary Care data (n=1.8 million)
• UK Biobank (n=0.5 million)
• Glasgow Psychosis cohort (n=7,500)
Mood disorder, cardiovascular disease and
the impact of psychotropic medication
(Martin et al, under review)
0
5
10
15
20
25
30
35
40
45
None One Two Three > Four
Bipolar Disorder (n=1,608)
MDD (n=31,756)
Controls (n=116,079)
Number of comorbidities
%
0
5
10
15
20
25
30
35
40
Any cardiovascular disease
Hypertension Diabetes
Bipolar Disorder (n=1557)
MDD (n=30,990)
Controls (n=113,444)
%
Partially adjusted a Fully adjusted b
OR (95% CI) OR (95% CI)
CVD any:
Control 1 1
Depression 1.29 (1.25, 1.33) 1.15 (1.12, 1.19)
Bipolar 1.50 (1.34, 1.68) 1.28 (1.14, 1.43)
Hypertension:
Control 1 1
Depression 1.27 (1.23, 1.31) 1.15 (1.11, 1.18)
Bipolar 1.44 (1.29, 1.61) 1.26 (1.12, 1.42)
Diabetes:
Control 1 1
Depression 1.29 (1.22, 1.37) 1.07 (1.00, 1.13)
Bipolar 1.37 (1.12, 1.67) 1.01 (0.81, 1.24)
A Partially adjusted: age, sex, deprivation and ethnicity B Fully adjusted: age, sex, deprivation, ethnicity, BMI, smoking status, alcohol consumption and current use of
psychotropic medication.
Diabetes
(N=7,825)
OR and 95%CI
MI
(N=3,129)
OR and 95%CI
Angina
(N=4,222)
OR and 95%CI
Hypertension
(N=38,840)
OR and 95%CI
Stroke
(N=2,066)
OR and 95%CI
Controls, no psychotropic
medication
1 (Reference) 1 (Reference) 1 (Reference) 1 (Reference) 1 (Reference)
Controls on psychotropic
medication
1.91
1.70, 2.14
1.85
1.54, 2.21
2.26
1.96, 2.61
1.59
1.48, 1.70
3.28
2.76, 3.90
Cardiometabolic disease, mood disorder and psychotropic medication (adjusted for age, sex, social deprivation, ethnicity)
Diabetes
(N=7,825)
OR and 95%CI
MI
(N=3,129)
OR and 95%CI
Angina
(N=4,222)
OR and 95%CI
Hypertension
(N=38,840)
OR and 95%CI
Stroke
(N=2,066)
OR and 95%CI
Controls, no psychotropic
medication
1 (Reference) 1 (Reference) 1 (Reference) 1 (Reference) 1 (Reference)
Controls on psychotropic
medication
1.91
1.70, 2.14
1.85
1.54, 2.21
2.26
1.96, 2.61
1.59
1.48, 1.70
3.28
2.76, 3.90
MDD, no psychotropic
medication
1.15
1.08, 1.23
1.33
1.21, 1.47
1.33
1.22, 1.45
1.21
1.18, 1.26
1.45
1.29, 1.63
MDD on psychotropic
medication
2.12
1.93, 2.33
1.83
1.55, 2.15
2.57
2.28, 2.91
1.63
1.54, 1.73
2.97
2.54, 3.48
Cardiometabolic disease, mood disorder and psychotropic medication (adjusted for age, sex, social deprivation, ethnicity)
Diabetes
(N=7,825)
OR and 95%CI
MI
(N=3,129)
OR and 95%CI
Angina
(N=4,222)
OR and 95%CI
Hypertension
(N=38,840)
OR and 95%CI
Stroke
(N=2,066)
OR and 95%CI
Controls, no psychotropic
medication
1 (Reference) 1 (Reference) 1 (Reference) 1 (Reference) 1 (Reference)
Controls on psychotropic
medication
1.91
1.70, 2.14
1.85
1.54, 2.21
2.26
1.96, 2.61
1.59
1.48, 1.70
3.28
2.76, 3.90
MDD, no psychotropic
medication
1.15
1.08, 1.23
1.33
1.21, 1.47
1.33
1.22, 1.45
1.21
1.18, 1.26
1.45
1.29, 1.63
MDD on psychotropic
medication
2.12
1.93, 2.33
1.83
1.55, 2.15
2.57
2.28, 2.91
1.63
1.54, 1.73
2.97
2.54, 3.48
Bipolar disorder, no
psychotropic medication
1.43
1.17, 1.75
2.03
1.53, 2.68
1.82
1.40, 2.36
1.48
1.32, 1.66
1.98
1.40, 2.81
Bipolar disorder on
psychotropic medication
2.21
1.62, 3.00
3.10
2.04, 4.71
2.22
1.46, 3.39
1.65
1.36, 2.00
2.95
1.78, 4.90
Cardiometabolic disease, mood disorder and psychotropic medication (adjusted for age, sex, social deprivation, ethnicity)
Chronic multisite pain in major depression
and bipolar disorder: cross-sectional study of
149,612 participants in UK Biobank.
(Nicholl et al, under review)
• Definition of multisite pain:
“In the last month have you experienced any of the following
that interfered with your usual activities?”
– Headache pain
– Facial pain
– Neck or shoulder pain
– Back pain
– Stomach or abdominal pain
– Hip pain
– Knee pain
– “Pain all over the body” (widespread)
RRR (95% CI)*
Model Depression Bipolar disorder
1. Unadjusted (n=149,612)
No chronic pain 1 1
1 site 1.35 (1.31, 1.39) 1.41 (1.24, 1.60)
2-3 sites 1.88 (1.81, 1.94) 2.35 (2.07, 2.38)
4-7 sites 3.12 (2.91, 3.34) 4.43 (3.55, 5.52)
Widespread pain 2.80 (2.53, 3.11) 5.38 (4.04, 7.16)
1. Adjusted for sex, age, ethnicity,
deprivation, employment
status BMI, smoking status
frequency of alcohol
consumption, comorbidity
count (n=145,518)
No chronic pain 1 1
1 site 1.27 (1.23, 1.31) 1.27 (1.12, 1.45)
2-3 sites 1.59 (1.54, 1.65) 1.84 (1.61, 2.11)
4-7 sites 2.13 (1.98, 2.30) 2.39 (1.88, 3.03)
Widespread pain 1.86 (1.66, 2.08) 2.37 (1.73, 3.23)
RRR (95% CI)*
Model Depression Bipolar disorder
1. Unadjusted (n=149,612)
No chronic pain 1 1
1 site 1.35 (1.31, 1.39) 1.41 (1.24, 1.60)
2-3 sites 1.88 (1.81, 1.94) 2.35 (2.07, 2.38)
4-7 sites 3.12 (2.91, 3.34) 4.43 (3.55, 5.52)
Widespread pain 2.80 (2.53, 3.11) 5.38 (4.04, 7.16)
1. Adjusted for sex, age, ethnicity,
deprivation, employment
status BMI, smoking status
frequency of alcohol
consumption, comorbidity
count (n=145,518)
No chronic pain 1 1
1 site 1.27 (1.23, 1.31) 1.27 (1.12, 1.45)
2-3 sites 1.59 (1.54, 1.65) 1.84 (1.61, 2.11)
4-7 sites 2.13 (1.98, 2.30) 2.39 (1.88, 3.03)
Widespread pain 1.86 (1.66, 2.08) 2.37 (1.73, 3.23)
RRR (95% CI)*
Model Depression Bipolar disorder
1. Unadjusted (n=149,612)
No chronic pain 1 1
1 site 1.35 (1.31, 1.39) 1.41 (1.24, 1.60)
2-3 sites 1.88 (1.81, 1.94) 2.35 (2.07, 2.38)
4-7 sites 3.12 (2.91, 3.34) 4.43 (3.55, 5.52)
Widespread pain 2.80 (2.53, 3.11) 5.38 (4.04, 7.16)
1. Adjusted for sex, age, ethnicity,
deprivation, employment
status BMI, smoking status
frequency of alcohol
consumption, comorbidity
count (n=145,518)
No chronic pain 1 1
1 site 1.27 (1.23, 1.31) 1.27 (1.12, 1.45)
2-3 sites 1.59 (1.54, 1.65) 1.84 (1.61, 2.11)
4-7 sites 2.13 (1.98, 2.30) 2.39 (1.88, 3.03)
Widespread pain 1.86 (1.66, 2.08) 2.37 (1.73, 3.23)
7,250 patients with psychotic disorder registered (2013):
– schizophrenia (n=4,787)
– bipolar disorder (n=1,784)
– organic psychosis (n=67)
– psychotic depression (n=452)
– substance-induced psychosis (n=160)
Baseline and annual follow-up information:
ICD-10 diagnosis, Community Health Index (CHI) number, ethnicity, marital status, accommodation status and postcode, employment status, educational attainment, family history of psychosis, psychiatric admissions data, current illness severity (including CGI and HoNOS scores), use of the mental health act, current and previous medications, adverse drug effects, psychosocial interventions received and psychiatric comorbidities.
0
100
200
300
400
500
600
700
800
900
Least deprived 2 3 4 Most deprived
De
ath
s p
er
10
,00
0 p
er
5 y
ear
s
Death rates in Major Mental Illness (MMI) by social deprivation: Glasgow and Scotland
MMI
Glasgow
Scotland
Langan Martin et al, BMC Psychiatry, in press.
0
100
200
300
400
500
600
700
800
900
Least deprived 2 3 4 Most deprived
De
ath
s p
er
10
,00
0 p
er
5 y
ear
s
Death rates in Major Mental Illness (MMI) by social deprivation: Glasgow and Scotland
MMI
Glasgow
MMI Excluding Suicide
Scotland
Langan Martin et al, BMC Psychiatry, in press.
NHS Greater Glasgow and Clyde:
SafeHaven
Dumb terminals
DWEducation
DW1
Anonymised servers
NHS Safe Haven
RCB
LPAC
decisions
Application
server
All clinical
datasets
Safe
Have
n G
overn
ance
Data
Requests
Clinical Trials
Research Data
Safe Haven – IT Infrastructure
Db3
Db2
SQL cluster
Clinical Non
health
Db1 Dataset
Dataset CHI
Seeded
Non health
data
Non
health
Clinical
Data is now in a
data warehouse
structure using only
surrogate keys to
link
DWSocialWork
DWEducation
DWSocialWork
DW1
Current Datasets •Datasets in Safe Haven
– SMR00 - Outpatient Attendance
– SMR01 – Acute inpatient & Day Care
– SMR02 – Maternity
– SMR04 – Mental Health
- CHI – GG&C patient population (1.3 million)
- GRO – Births and deaths date for GG&C
- ePrescribing – encashed prescriptions for Glasgow
- GP (LES and Keep Well) 250 practices
- Heart Failure – locally held national Heart Failure database
- Rheumatology – local clinical database
- SCI DC – GGC population of national Diabetic database
- SCI Store results for GGC
- Parkinson – local clinical database for Movement disorders
- Weight Management
- PsyCIS – schizophrenia database
- Clozapine database
- EDIS (A&E now replaced by Trak care)
• REC approval is to submit an amendment every time 6 new databases are added
• In discussion to extend health data to other Boards in NRS West – Lanarkshire, A&A, D&G and Golden Jubilee
P O P U L A T I O N
S P I N E
H E A L T H
D A T A
Clinical Trial Support e Feasibility, e Recruitment, e Data capture & f/up
Real World Clinical Studies Virtual case/control cohorts, epidemiology, pharmaco-
epidemiology
Actionable Data Analytical tools, visual analytics
Health Economic analyses
Increased efficiency &
effectiveness of NHS services
Enrichment with other data sets Education
Social work Housing
Transport Police
CHI linkage CHI seeding and linkage
Virtual population-wide cohorts
e.g. birth, geriatric Followed longitudinally
Understanding, improving and integrating services
Centre for Data-Driven Research & Innovation (name to be decided)
UK data:
• Bipolar Disorder Research Network (n=8,000)
• Scottish Primary Care data (n=1.8 million)
• UK Biobank (n=0.5 million)
• Glasgow Psychosis cohort (n=7,500)
Bipolar Disorder Research Network:
Nick Craddock, Ian Jones, Liz Forty, Lisa Jones
Scottish Primary Care data:
Stewart Mercer, Bruce Guthrie, Gary McLean, Julie Langan Martin
UK Biobank:
Jill Pell, Daniel Martin, Barbara Nicholl, Daniel Mackay
Glasgow Psychosis cohort:
Moira Connolly, John Park, Julie Langan Martin
Thanks
Physical comorbidity with bipolar disorder: lessons from UK data
Daniel Smith
Symposium 33: ‘Big data’ and bipolar disorder in the UK