The latest statistics compiled by the Office for National Statistics (ONS, 2015) indicate that in the UK in 2013, there were 6,233 deaths recorded as suicides in individuals over the age of 15; 4% more than in 2012.

Increasing our knowledge about the factors that may precipitate and motivate suicide, and increasing our understanding of the periods and times during which individuals may be most at risk of taking their own lives, is paramount to suicide prevention efforts. Days of the week, months of the year, and national and religious holidays could be used as indicators of temporal variation in suicide rates.

Previous findings relating to temporal variation in suicide rates in the general population have been somewhat contradictory (Ajdacic-Gross et al., 2010), and relatively few previous studies have examined temporal variation in suicide rates in clinical populations.

Consequently, new research recently published in the Journal of Affective Disorders, led by Brendan Cavanagh, has aimed to investigate temporal variation in suicide rates in England in both general and clinical population samples.

Methods 

Data were collected as part of the National Confidential Inquiry into Suicide and Homicide by People with Mental Illness (NCISH; Appleby et al., 1997). Deaths in England that received either a ‘suicide’ or ‘open’ verdict at the coroner’s inquest between 1 January 1997 and 31 December 2012 were considered to be suicides in this study, following conventions for conducting suicide research in the UK.

• General population sample: Data on suicide rates among individuals aged 10 years or older in the general population were obtained from the ONS.
• Clinical population sample: Data on suicide rates among individuals aged 10 years or older who had been in contact with mental health services in the 12 months before their suicide were obtained from NHS mental health trusts.

Temporal variation in suicide rates was examined by month of the year (to explore seasonal variation), day of the week, and national and religious holidays, adjusted for individuals’ sex and age. The authors used ‘December’ and ‘Sunday’ as reference groups to compare suicide rates during other months of the year and on other days of the week to. The national and religious holidays selected for this study included the Christmas period (consisting of Christmas Eve, Christmas Day, and Boxing Day), the Easter period (consisting of Good Friday, Easter Saturday, Easter Sunday, and Easter Monday), New Year’s Eve, New Year’s Day, Valentine’s Day, Mother’s Day, and Father’s Day. The authors used suicides on all other days of the year as a comparison point for each national or religious holiday.

Drawing on the previous literature in this area, some additional subgroup analyses were also conducted within the clinical population sample. Specifically, comparisons were made between individuals with a primary diagnosis of depression or bipolar disorder and individuals diagnosed with any other psychiatric condition, in terms of their monthly suicide risk. Monthly variation between methods of suicide was also examined. Finally, the researchers explored the potential role of social isolation and employment status on variation in suicide across days of the week.

Results

Findings relating to the general population sample (N = 73,591):

• 55,358 deaths were those of men (75%), and 18,233 were those of women (25%). The median age at death for men was 43 years (with a range of 10-104 years), and for women was 48 years (with a range of 10-102 years).
• New Year’s Day was the day with the highest suicide rate, with a mean rate of 17.6 suicides per day. Christmas Day was the day with the lowest suicide rate, with a mean rate of 9.2 suicides per day. Women experienced a significantly greater reduction in suicide risk (40%), compared to men (20%), during the Christmas period as a whole.
• Trends pointed towards suicide risk being higher in the first part of the year (January and April were associated with the highest risk of suicide), and lower in the final quarter (December was the month with the lowest risk).
• Suicide risk was significantly higher on a Monday (a 20% increase), compared to a Sunday, and trends pointed towards a decrease in risk of suicide over the course of the week.

Findings relating to the clinical population sample (N = 19,318):

• 12,787 deaths were those of men (66%), and 6,531 were those of women (34%). The median age at death for men was 43 years (with a range of 10-98 years), and for women was 47 years (with a range of 13-96 years).
• 21 May was the day with the highest suicide rate, with a mean rate of 4.9 suicides per day. 27 March, 28 December, and New Year’s Eve had the joint lowest suicide rate, with a mean rate of 2.0 suicides per day.
• Trends pointed towards suicide risk being higher in the first part of the year (May was associated with the highest risk of suicide), and lower in the final quarter (December was the month with the lowest risk).
• Suicide risk was significantly higher on a Monday (a 28% increase), compared to a Sunday, and trends pointed towards a decrease in risk of suicide over the course of the week. In addition, individuals aged 50 years or older had a significantly higher incidence of suicide on Mondays versus Sundays than people aged under 35 years, as did individuals living alone compared to those living with others.

Conclusions

The strengths of this study lie in the authors’ use of a large national dataset and their exploration of temporal variation in suicide rates in both general population and clinical population samples. In both samples, suicide risk was higher during the first part of the year and lower in the final quarter, with December being the month with the lowest risk. Suicide risk was also higher on a Monday, compared to a Sunday, in both samples. On the other hand, while New Year’s Day was the day with the highest suicide rate in the general population sample, it was 21 May in the clinical population sample.

Having provided evidence for temporal variation in suicide rates, the mechanisms behind this are also important to consider. For instance, the authors suggest that their finding that suicide rates increased on New Year’s Day in the general population sample, and on Mondays in both samples, could support psychological explanations of a ‘broken promises effect’, which occurs when individuals’ hopes of a new beginning at the start of new temporal periods are not met (e.g. Gabennesch, 1988).

Limitations 

As the authors acknowledge, the findings of this study should be considered in the context of the following limitations:

• The cultural context of the findings of this study is important to consider, as the findings may not be generalisable to other countries with, for instance, different national and religious holidays.
• The authors were unable to distinguish the clinical population from within the general population in their analyses of the general population sample.
• The clinical population sample may represent an underestimation of individuals who had been in contact with mental health services before their death, as those who were in contact with clinical services over 12 months prior to their death may not have been included.
• While differences in suicide risk were apparent when the authors compared particular days of the week or months of the year with their respective reference groups (Sundays and December), no differences in risk were found when different months of the year or weekdays were compared with one another – the 95% confidence intervals for the point estimates of risk overlapped.
• The large number of comparisons that the authors conducted in their analyses may have increased the probability of spurious significant findings, which may have occurred by chance. This is known as a Type 1 error.

Implications

Overall, while causality cannot be inferred from the authors’ findings due to the observational nature of their study, their findings may nonetheless have important implications for clinical services, in terms of offering targeted support to individuals at particular times of risk:

Clinical services might consider offering increased monitoring and support for vulnerable individuals at particular times of transition (for example, the beginning of the year or the beginning of the week). (p. 180)

In relation to this, it is important to also highlight the authors’ finding that the majority of the deaths in both samples in this study were those of men. The findings may also have implications for preventive interventions and suicide awareness campaigns. For example, campaigns could be launched to:

raise population suicide awareness at times of increased risk, for example through public health messages prior to New Year’s Day. (p. 180)

Future research in this area could perhaps attempt to explore further who may be most vulnerable at particular times of the year. For instance, the authors found that suicide risk on Mondays was particularly high in the clinical population sample for individuals aged over 50 or who lived alone, but were unable to investigate this in the general population sample.

If you need help

If you need help and support now and you live in the UK or the Republic of Ireland, please call the Samaritans on 116 123.

If you live elsewhere, we recommend finding a local Crisis Centre on the IASP website.

We also highly recommend that you visit the Connecting with People: Staying Safe resource.

Links

Primary paper

Cavanagh B, Ibrahim S, Roscoe A, Bickley H, While D, Windfuhr K, Appleby L. Kapur N. (2016) The timing of general population and patient suicide in England, 1997-2012. Journal of Affective Disorders, 197, 175-181. [Abstract]

Other references

Office for National Statistics (2015). Suicides in the United Kingdom, 2013 Registrations (PDF).

Ajdacic-Gross V, Bopp M, Ring M, Gutzwiller F, Rossler W. (2010) Seasonality in suicide – A review and search of new concepts for explaining the heterogeneous phenomena. Social Science & Medicine, 71, 657-666. [PubMed abstract]

Appleby L, Shaw J, Amos T. (1997) National Confidential Inquiry into Suicide and Homicide by People with Mental Illness. British Journal of Psychiatry, 170, 101-102.

Gabennesch H. (1988) When promises fail: A theory of temporal fluctuations in suicide. Social Forces, 67, 129–145.

Photo credits

• Steven Lilley CC BY 2.0
• Joel Kramer CC BY 2.0
• Paul Wilkinson CC BY 2.0
• CJ Sorg CC BY 2.0
• Tony Alter CC BY 2.0
• nimble photography CC BY 2.0