Mortality in severe mental illness: then and now Psychobiology - - PowerPoint PPT Presentation

Psychobiology Research Group

Mortality in severe mental illness: then and now Prof Nicol Ferrier Emeritus Professor of Psychiatry Newcastle University

Mortality associated with severe mental illness

People with psychosis die on average 15‐20 years younger than the general population (e.g. McGrath et al, 2008) Schizophrenia

More than 20-fold increased risk for death by suicide. Increase in deaths from infectious diseases and accidents. People with schizophrenia are up to 50% more likely to have type 2 diabetes mellitus and obesity. 70% of patients with schizophrenia, compared with approximately 50% of the general population, die of coronary heart disease (CHD). SMR for cardiovascular disease (CVD) deaths is 2-4 times age matched population.

Bipolar disorder

More than 20-fold increased risk for death by suicide Patients with untreated illness have >4-fold higher SMR CVD is the leading causes of premature mortality in this population but hospital admissions for CVD treatment were only slightly increased in BD compared to the general population

Evidence that the mortality gap is widening across the world. In UK, gap gradually increased from 2005 and rapidly from 2010

(Hayes et al, 2017)

Pre Premat ature Dea Death ‐ Causes

• Lifestyle Issues
• More sedentary
• Less likely to eat fruit and vegetables
• Twice as likely to develop type II diabetes mellitus
• 2-3 times more likely to be obese
• Smoking
• More likely to smoke
• 76% with first episode psychosis (FEP) were regular smokers
• Those with schizophrenia have a 10 fold increased death rate from respiratory disease
• Inverse Care
• Discrimination by health professionals
• Do not receive optimal physical health care
• Lowered reporting of physical symptoms
• Genetic Factors
• 15% of drug-naïve individuals with first-episode psychosis have elevated fasting glucose levels, high levels of

insulin and cortisol, and three times as much intra-abdominal fat as age and Body Mass Index -matched control subjects

Cardiovascular disease and deaths in severe mental illness CHD mortality

18-49 3.22 (1.99-5.21) 50 -75 1.86 (1.63-2.12) >75 1.05 (0.92-1.19)

Stroke deaths

2.53 (0.99-6.47) 1.89 (1.50-2.38) 1.34 (1.17-1.54)

Correll et al, May 2017. Meta-analysis of ~3.2 M SMI patients and ~113.3M controls CVD (1.95), cerebrovascular disease (1.57) , CCF (1.71*) and death due to CVD (2.45) all increased. * Unadjusted and cross-sectional

UK data:- 46,136 people with SMI and 300,426 without SMI (Osborn et al. 2007) Hazard ratios (95% CI) in age groups with SMI vs controls :-

CV risk factors: overview

*Data using Framingham study cohort. Dawber et al, Am J Public Health, 1951; 41: 279 Wilson et al, Circulation, 1998; 97: 1837–47

2 4 6 8 10 12 14

HTN DM Smoking BMI >27 TC >220

Single risk factors Multiple risk factors Odds ratios

Smoking + BMI

2

Smoking + BMI + TC >220

3

Smoking + BMI + TC >220 + DM

4

Smoking + BMI + TC >220 + DM + HTN

5

The Framingham Study Cohort*

BMI: body mass index, DM: diabetes mellitus, HTN: hypertension, TC: total cholesterol.

Antipsychotics and Metabolic Dysfunction

• A wealth of evidence exists showing link between antipsychotic

use and metabolic dysfunction (Starrenburg et al, 2009)

• “second generation” ‘worse’ than “first generation”
• H1, 5HT2-C, α2, β3 receptors (regulation of food intake)
• 5HT1-A, M3, α2 receptors (present in β cell in Pancreas –

regulation of insulin release, and insulin sensitivity)

(Adapted from Leucht et al, 2013)

Cardiovascular disease and deaths in severe mental illness HRs for CHD mortality in 50-75 age group of SMI (95% CI) Not on APs 1.4 (1.1-1.8) On APs 2.0 (1.7-2.3) p=0.009 Low dose 1.5 (1.1-1.8) High dose 2.5 (2.0-3.1) p<0.001

Correll et al, May 2017. CVD incidence increased with higher AP use, higher BMI and baseline CVD. Osborn et al. 2007 UK data:- 46,136 people with SMI and 300,426 without SMI

Mortality associated with schizophrenia Role of antipsychotics

• There is little evidence that long-term exposure to

antipsychotics increases mortality in schizophrenia.

• No increased CVD mortality in placebo controlled trials
• Mortality increased in drug free populations in both

cross-sectional and prospective studies but high quality data lacking.

• Some positive dose-response studies find increased

CVD deaths in patients on higher AP doses but a major confounding factor may be a higher risk factor load for somatic disorders in the most severely mentally ill. More rigorously designed, prospective studies are urgently needed.

Recent developments:‐ First Episode Psychosis (FEP)

• FEP strongly associated with insulin resistance (IR) (OR=5.14) compared to matched controls,

after relevant adjustments. Weaker and less convincing associations with lipids (low HDL and raised triglycerides were weakly associated with FEP). (Perry et al, 2016)

• Poor glycaemic control linked with increased levels of CRP, IL6, TNFα, IL1β (Dandona P et al ,

(2004)

• IR significantly associated with pre-clinical psychotic symptoms (OR=2.32) after adjusting for:

sex, ethnicity, gestational age, birthweight, physical activity, BMI, smoking, antipsychotic

• medication. No association lipids / metabolic syndrome – suggests IR comes first. The

relationship was non-linear, and exponential

• IL-6 was found to act as a ‘moderator’ in the relationship, suggesting that inflammation, IR and

psychotic symptoms are closely linked Even before the onset of psychotic illness, metabolic dysfunction is prominent.

• This effect is driven by dysglycaemia (insulin resistance)
• Raised IL-6 interacts with IR to cause psychotic symptoms

Morbidity and mortality in C19 asylums

• Many early reports that the insane died in excess and prematurely.

Burrows (1828) “insanity tends to the shortening of human life”.

• Farr (1841) started data collection and confirmed the high mortality
• f asylum inmates compared to the general population. Death rates

(deaths/inmate) were 5‐7 times higher that of matched populations and the “mortality gap” ~ 15 years. Some of the excess mortality linked to deaths from conditions like GPI, epilepsy and imbecility but it was also found in mania and melancholia.

• More marked in men, early in admission, in migrants and the
• malnourished. Lower is some asylums e.g. the Retreat in York
• These issues led, in part, to the development of asylums and

subsequently to the 1870 recommendation from the Commissioners in Lunacy that asylums perform PMs on all deaths.

Mortality in Severe Mental Illness in Victorian asylums

Dying to Get Out of the Asylum: Mortality and Madness in Four Mental Hospitals in Victorian Canada, c. 1841–1891 Wright D, Jacklin L, Themeles T (2013) Bulletin of the History of Medicine, 87, 590‐ 621

General Adult Population.

Moulsford/Berkshire Asylum

The Boroughs of Reading and Newbury built Moulsford Asylum by the Thames and the first patients were admitted in 1870. The asylum was designed to accommodate 285 patients but was almost full to capacity within the first year. The asylum was extended in 1880 to accommodate 609 patients and in asylum was expanded to a capacity of 800. In 1897 its name was changed from Moulsford Asylum to the Berkshire Lunatic Asylum. In 1948, when it was incorporated into the National Health Service, the name was changed again to Fair Mile Hospital. In 2003 Fair Mile was closed as it no longer provided appropriate accommodation.

Sunnyside Royal Hospital, Montrose

Originally founded in 1781 by Susan Carnegie as the Montrose Lunatic Asylum, Infirmary & Dispensary. Obtained a Royal Charter in 1810. In 1834, the Governors of the asylum, carrying out the wishes of Mrs Carnegie (who had strongly advocated the appointment of a medical specialist in insanity) appointed the phrenologist William Browne as medical superintendent. In 1858, a new improved asylum was completed but overcrowding was a problem. The asylum's patient numbers had grown to 670 by 1900 leading to further building. The site was officially closed in late 2011 and most patients were sent to a new £20 million build at Stracathro Hospital - the Susan Carnegie Centre. Sunnyside was open for 230 years before its closure and was the oldest psychiatric hospital in Scotland.

Sunnyside Asylum Numbers of inmates and deaths 1892‐1901

9% 8% 11%

Death rate

Berkshire Asylum Numbers of inmates and deaths 1896‐1905

8% 13% 9%

Death rate

Notes on numbers and deaths

• Both asylums had very favourable Commissioner reports (including the

staff and the food) although it was agreed there was overcrowding.

• Recurrent mention of and concern about admissions with no hope of

recovery in both asylums “unfavourable admissions have led to the lowest recovery rates in the history of the asylum” and “overcrowding is telling on the health of patients”.

• Frequent reports of typhoid, dysentery and scarlet fever in Berkshire in

patients and staff. Great concern over water supply and sanitation.

• Both asylums had somewhat raised death rates compared to the norm.

Fluctuations in Berkshire partly related to infectious GI diseases and partly due to nature of admissions. It was noted that aged admissions increased the death rate because of the “to be expected course of events”

Sunnyside Asylum, Montrose Deaths and PMs 1892‐1901

Berkshire Asylum Deaths and PMs 1896‐1905

Labels

Mania Melancholia

Common Mania Melancholia Acute mania Acute melancholia Chronic mania Chronic melancholia Infrequent Senile mania Delusional Mania Melancholia with delusions Mania with delusions Melancholia with stupor Delusional insanity Recurrent mania Occasional Folie circulaire Religious mania Acute delirious mania

Age at death of inmates with mania and melancholia who had a PM

Mean ages of death Mania : 55± 9 Melancholia: 55 ±7

Mortality data

Causes of death groupings from post‐mortems with a diagnosis of Mania or Melancholia in Berkshire and Sunnyside Asylums combined Cause of death grouping Alternatives (in order of frequency) TB

Phthisis pneumonale, Acute TB, TB, Acute miliary TB

Pneumonia

Bronchopneumonia, Acute Pneumonia, Empyema, Pleurisy, Bronchitis

Other Infectious diseases

Enteric fever, Dysentery, Meningitis , Septicaemia, Hydatid cysts

“Surgical”/other

Various cancers, Intestinal obstruction, Volvulus, Perforation, Pyelonephritis, Peritonitis, Rupture of bowel, Ulcerative colitis, Choking

Vascular

Heart disease, Heart failure, Morbis Cordis, Fatty heart Valvular disease of heart, Apoplexy, Cerebral haemorrhage , Congestion of lungs

Brain atrophy

Brain atrophy, Cerebral softening, Senility, Brain congestion

Suicide

Strangulation, Laceration of throat

Exhaustion

Exhaustion from mania, Exhaustion from melancholia

Causes of death groupings from post‐mortems in inmates with diagnosis of Mania

• r Melancholia in Berkshire and Sunnyside Asylums combined in under 55 group

and total group.

Cause of death groupings

Number Total group % of deaths Total group Number Under 55 % of deaths Under 55

TB

76 20 65 38

Pneumonia

50 13 26 15

Infectious diseases

14 4 7 4

“Surgical”/other (including Ca)

41 (Ca 19) 11 (Ca 5% of total) 22 13

Vascular

86 23 26 15

Brain atrophy

63 17 9 5

Suicide

2 1 2 1

Exhaustion

39 11 16 9

Total

371 100 173 100

Notes on PMs

• Very close concordance between PM cause of death and that in the register of

deaths in both asylums.

• In vast majority of cases there was close concordance between PM findings and

cause of death. Only 2 problematical cases. One PM only partial.

• The PMs were performed by staff psychiatrists.
• Evidence of use of appropriate pathological and anatomical terms.
• Earlier PM reports contained drawings etc. but none over study period. Microscopy

mentioned but no results perhaps due to “The want of a good microscope has been felt much lately”

• Allusion in general records to carrying out PMs on those who gave permission but

exceptions mentioned eg. “ 47 post‐mortems were carried out and in 5 the sanction

• f the relatives was not obtained” (Sunnyside Minute book,1894). In Berkshire, it

was described as a “duty” to carry out a post‐mortem on all cases and consent may have been tacit unless objections raised.

• Quite frequent mentions of referrals for inquests (e.g. sudden deaths) Some

proceeded to a PM as part of an inquest. Inquests were carried out on the premises (“rather than the local tavern”).

Demographics of sub‐groups in deaths by/with exhaustion in mania or melancholia, 1871‐1906

Total group of death by exhaustion from mania or melancholia. Death by exhaustion with post‐mortem report Group after exclusion

• f cases with clear‐cut

physical disease and senile mania cases Group with case notes records examined. Number 155 90 57 34 Mean Age (years) 55 55 51 47 Sex M/F 73/82 41/49 27/30 16/18 Mania/Melancholia 102/53 67/23 38/19 25/9

Clinical features of death by/with exhaustion

Age Mean age of death by/with exhaustion with mania group was 50yrs. Mean age of death by/with exhaustion with melancholia group was 57yrs Episode First “attack” for 80% Onset ‘Sudden onset’ was specifically stated for ~ half the patients. Length The mania cases showed particularly short total illnesses. Duration of illness added to length of admission of less than two months in majority. Seven patients with mania had illnesses totalling less than two weeks. Melancholia average was between 2 and 6 months Weight 50% described as emaciated on admission, 50% well nourished. About 2/3rds of were emaciated or severely emaciated at death but a third remained well nourished. Food Food refusal was specifically mentioned in 2/3rds The majority of these were tube fed.

Other features

• About half the cases were preoccupied with religious
• delusions. All of these cases exhibited food refusal. One man,

a groom, saw visions of God and felt he should have a hole in his hand to let the poison of his sin against God out. He died

• f severe emaciation as did a female patient from Sunnyside

who continually called on God saying she had been dammed.

• Agitation and marked motor restlessness was common but

stuporose states also seen.

Some examples of precipitants of mania associated with death by exhaustion

• RM female aged 55. Single domestic servant, very temperate and hard working. In service to

a Baptist minister whose wife died. From him talking to her and shaking her hand, the patient ‘developed the idea that he was going to propose to her’ and she had an unshakeable belief in his love. His engagement was announced and RM went ‘clean off’. First attack. Ill for 14 days and died of exhaustion 8 days after admission.

• TB, 48 year old single labourer who has undergone ‘privation and had money worry’. Attack

said to be cause by ‘disappointment in love’ ‐ jilted by his first cousin. First attack of sudden

• nset. Ill for 6 days and died 6 days after admission
• 66 year old intelligent, temperate agricultural labourer but no work for 5 weeks. His wife had

to go to London for an operation for an internal complaint since when he was not able to

• cope. Acutely ill for a week and died two weeks after admission. First attack.
• 35 year old married cook. Cause ‘a fright’. Patient had been at Glamis Castle cooking at the

camp there. Returning to the tent was shocked to see a man in there stealing beef. She got such a fright that she was sleepless resulting in mania. Second attack, two weeks duration and died 12 days after admission

Probable catatonia

• A 19‐year‐old single agricultural labourer was admitted in first attack of 4

months duration.

• The cause was said to be an hereditary predisposition as his uncle was

insane.

• He had gradually become unwell and had to give up work a month before
• admission. He was physically well built but his heart sounded feeble and

pulse was weak. He refused food, only eating “by stealth”. Mental state was said to be melancholic stupor. He was oblivious of surroundings, silent and cataleptic and rigid with beads of sweat on his face.

• Given a stimulant mixture and a cathartic enema. “Silent and stubborn,

refusing food with utmost certainty” and “voluntary muscular action cataleptic... remaining in whichever posture he is placed in”. At night he cried out to the Lord to have mercy on him but stopped when attendants approached.

• Rallied briefly and said he would like some beer which he was given but

soon after became unconscious and died. The cause of death was listed as Exhaustion from melancholia and food refusal. PM unremarkable.

“Acute delirious mania”

• An 18‐year‐old single woman was admitted in her first attack of ‘unknown

cause’.

• She had been unwell for 8 days. Her admitting Certificates described a

state of “great excitement, some at her mother” who she had threatened to kill. She “sings, weeps and shouts alternately” and thinks her food is

• poisoned. She was extremely restless and noisy and moves “exhaustively in

an aimless way”.

• She was particularly restless at night and refuses all food and was tube

feed and also given an enema. She was described as having a “low stuttering delirium”. 2 days after admission she “collapsed” and the following day was noted to be “commencing cardiac failure” and her circulation in left leg was almost absent below the knee. Two days later the limb was “as if” dead.

• She was given “a fair quantity of stimulant” but her “cardiac failure

became more obvious”. She sank into a coma and died 9 days after

• admission. Post‐mortem revealed only congestion of organs and nothing

specific.

Probable delirium

• FC was a 43‐year‐old married wife of a whitesmith with 6 children.
• In her admission records the cause of her insanity was stated as rheumatic
• fever. Although there was a lot of evidence to support rheumatic fever as a

cause of her illness and her death, this was not reflected in either the Register or in the post‐mortem report.

• Her insanity had all the hallmarks of a delirium with disorientation in

‘place and time’ and she was described as being ‘delirious’. She had an intermittent fever and when the temperature was high she developed transient delusions that the nurses had murdered her children. She had physical signs of rheumatic heart disease.

• At post mortem the heart showed ‘vegetations’ on her valves, the

pathognomic sign of infective endocarditis.

• This condition had been recognised in 1851 and had been extensively

reviewed by Osler in 1885 in his three Gulstonian Lectures and was known to cause the sufferer to be exhausted. Here, all the signs and symptoms of the condition had been recognised and reported and it is not clear why it was not recorded as a cause of death.

Non‐Accidental injury?

• JH was a 63‐year‐old pedlar, married with two children. This was his fourth

admission having been in a number of asylums in the South of England.

• He developed melancholia; he was actively suicidal and he expressed the view that

he had offended against ‘moral law’. He heard the voice of the Lord whispering that he ‘was going to Hell’. From being depressed and miserable, he suddenly became excitable, restless and then violent. He struck his head against a wall and tried ‘to tear out his testicles’ which he later said was ‘at the command of God’. He attacked an attendant and a violent scuffle ensued. He was placed in a padded cell with gloves but sank rapidly over the next few days and died.

• Post‐mortem revealed widespread disease of the vessels to the heart and brain

and a cerebral haemorrhage ‘of two to three days duration’.

• Thus, he died not of exhaustion (though this was a feature of his last days) but of a

stroke or a bleed in the brain possibly caused by or exacerbated by a head injury, possibly (but possibly not) self‐induced.

• It is not clear why the cerebral haemorrhage and head injury are not reflected in

his records nor why there was no inquest. This would have cleared the attendants if that was the way the evidence pointed. This must count as a suspicious death.

Potential cause?

• 40 year old single grocer. The cause of his severe melancholia was emphatically stated as
• masturbation. Around this time Bullen, a Victorian alienist with robust views, had attributed

cases of exhaustion to ‘marked degenerative tendencies’ and outlined ‘prolonged physical drain, mental exhaustion, or moral depravation of masturbatic (sic) habits’ as the immediate causes.

• The patient was described as being a religious enthusiast who confessed to having practiced

the habit for many years. He had marked religious delusions and believed he had ‘sinned away the day of grace’ and that despite preaching to others it is he who ‘falls short’. While masturbation seems an unlikely cause of his death from exhaustion with melancholia, it was a marked problem as it was noted that the ‘seminal emissions occur several times a day’

Cardiovascular findings in deaths by exhaustion

• Pulse: ‐ Definite ≥ 100/min and/or irregular; Probable: ‐ rate over 90/min, weak, poor tone etc..
• Heart signs: ‐ Definite Hypertrophy, murmur; Probable: ‐ heart sounds abnormal or weak, muffled etc

Findings of the pulse Heart signs on examination

Atheroma of basal arteries and/or aorta and/or valvular vessels

Major Heart abnormalities at PM Number of

• bservations

34 34 57 57

Definitely abnormal

7 12 29 38 7 with dilatation of

• ventricle. 23 with

fatty heart or equivalent.

Probably abnormal

15 6

Normal

4 5 28 19

Uncertain‐ no mention

8 11

Death from exhaustion in mania and melancholia: Discussion

The current series have the classical clinical picture characterized by severe agitation, restlessness, food refusal with the majority having a rapid onset and short duration The majority had abnormal cardiac findings on admission, evidence of atheroma at post mortem and evidence of circulatory collapse. The finding of degenerated heart muscle probably represent ischaemia. These findings raise the notion that both the hectic mental state and the cardiac dysfunction both relate to severe and uncontrollable stress before and after asylum admission in those with a particular predisposition.

Conclusions

• Infectious diseases were the predominant cause of death in

younger patients with SMI s accounting for 57% of the deaths. In developing countries excess deaths in SMI are driven by deaths from infectious diseases. Increased susceptibility to infection and/or death from infection?

• Vascular, particularly cardiac, pathology was particularly

common in such a young group. The role of nutrition, infection and alcohol to be analysed. These results suggest the role of drugs and obesity in premature death in SMI may have been

• veremphasised and genetic and other factors e.g. early life

experiences, stress, inflammation may be more relevant.

• Suicide, restraint, accidents and NAI are uncommon in this

population

• Antipsychotics have had a bad press…