Number of homicides related to mental disorder has fallen since 1970s in England and Wales, study finds

BMJ 2008; 337 doi: (Published 01 August 2008) Cite this as: BMJ 2008;337:a1113
  1. Susan Mayor
  1. 1London

    The number of homicides attributed to mental disorders has fallen in England and Wales over the past 30 years, while that of other homicides has risen, a new analysis of official statistics shows.

    The study looked at four sets of homicide statistics from 1946 to 2004. Researchers compared the numbers of homicides that were due to mental disorder—categorised by legal outcome (not guilty by reason of insanity or diminished responsibility, defendant unfit to plead, infanticides, and cases where the offender committed suicide at the same time as or after the homicide)—with homicides that were not considered to be due to mental disorder (British Journal of Psychiatry 2008;193:130-3).

    The results showed that the annual numbers of all homicides and of those that were due to mental disorder rose steadily until the mid-1970s. The number of homicides associated with mental disorder rose from fewer than 50 a year in 1957 to well over 100 by the 1970s. The highest annual rate was 0.24 per 100 000 population in 1973, and the absolute number peaked in 1979. From 1957 to 1980 the figures show a strong correlation between homicides due to mental disorder and total homicides (P<0.0001).

    However, after 1981 numbers of homicides associated with mental disorder fell, to levels below those seen in the early 1950s. The rate has remained at 0.07 homicides per 100 000 population or lower since 2000. At the same time, other homicides have continued to rise, to above 1.5 per 100 000.

    The authors, led by Matthew Large, a psychiatrist in private practice in Sydney, Australia, said that the reasons for the rise and fall in homicides attributed to mental disorder were not clear. They suggested, “The decline in homicides may be due to improvements in psychiatric treatment and service organisation.”

    Dr Large said, “The main explanation for the decline is a reduction in homicides by people with schizophrenia.” He added, “During the 1970s the health service started to organise community mental health services, and a lot more patients were treated.” In comparison with many other countries, he noted, England has a shorter duration of untreated psychosis.

    Another possibility may be an informal change to the legal tests for the finding of homicide resulting from mental disorder, although the actual criteria have not changed since 1957.

    Dr Large initiated the study, which received no outside funding, after observing that around 60% of people in New South Wales who committed homicide and who were considered to be mentally ill were not treated. He wanted to look at similar rates elsewhere, and England and Wales are two of only three countries in the world (the other is Finland) that have routinely collected data on homicide and mental disorder over the long term.

    It was “surprisingly difficult” to get the study published, Dr Large said, and some reviewers were sceptical about the findings. He said that his study’s findings contrasted with those in other studies—particularly in other countries—which found no consistent downward trend in numbers of homicides related to mental disorder and a direct correlation with the total homicide rates.

    “There seems to be resistance to the idea that homicide rates may be modifiable,” he concluded.


    Cite this as: BMJ 2008;337:a1113