Suicides during the Swedish midsummer holiday: analysis of cause of death data 1980–2018

Figures & data

Figure 1. Paired samples t-tests of deviating mean death frequency on Midsummer’s Eve (ME), Midsummer’s Day (MD) and the mean of both days (MEMD), compared to the mean death frequency on Adjacent Dates (AD). Bars show the mean death frequency, and the 95% confidence intervals represent the annual variability. The dotted horizontal line represents the mean frequency across all days (4.32 deaths per day = 3.31 deaths due to suicide + 1.02 deaths with undetermined intent).

Figure 2. Mean death frequency (diagnosis type = suicide) for each day in the observation period. Midsummer’s Eve (ME) always occurs on a Friday (June 19^th to 25^th) and is followed Midsummer’s Day (MD) on the next day (center of X-axis). The 95% confidence intervals reflect the variance in suicide between the 39 years. Mean daily frequency across all days = 3.31 deaths (dotted horizontal line).

Figure 3. Mean death frequency (diagnosis type = undetermined intent) for each day in the observation period. Midsummer’s Eve (ME) always occurs on a Friday (June 19^th to 25^th) and is followed Midsummer’s Day (MD) on the next day (center of X-axis). The 95% confidence intervals reflect the variance in suicide between the 39 years. Mean daily frequency across all days = 1.02 deaths (dotted horizontal line).

Table 1. Descriptive statistics.

External cause of death	Midsummer’s Eve (ME)		Adjacent days (AD)
	%	% undetermined within category	%	% undetermined within category
Poisoning	36.2	(68.5)	34.8	(40.0)
Hanging/suffocation/strangulation	26.8	(0)	28.4	(0.7)
Drowning/submersion	10.7	(75.0)	9.8	(35.5)
Firearms/explosives	11.4	(0)	10.3	(2.7)
Jumping from height	4.7	(0)	5.8	(20.5)
Vehicle crashes/moving objects	6.0	(77.8)	7.6	(31.5)
Other methods	4.0	(83.3)	3.4	(40.0)
All methods (total)	100	(40.9)	100	(22.8)

Prevalence (%) of various external causes of death observed during Midsummer’s Eve (ME) and Adjacent days (during 1980–2018). The rate (%) of undetermined diagnosis (deaths with undetermined intent) is presented within each category.

The values are rounded to one decimal, approximating 100%. Deaths with undetermined intent are defined by the ICD codes Y10–Y34 (10th revision) or E980–E989 (8th and 9th revision). Suicides (intentional self-harm) are defined by the codes X60–X84 or E950–E959.

Data availability statement

The data used for this study was provided by the Swedish National Board of Health and Welfare, in 2019. For ethical reasons, we do not provide raw data from the Swedish cause of death registry (see ethical statement in section: Methods and materials). Other methodological information to support replication by independent researchers is available from the corresponding author upon request.