Suicides by drowning in Greece: Findings from a 10-year sample (2006–2015)

Introduction

Comparatively few studies have focused on suicide by drowning even though it may be the reason that some individuals have never been found. In fact, drowning – together with poisoning by solids and liquids – is considered to be the commonest cause of undetermined deaths; the latter appear to reduce recorded suicide rates by 10%. ¹

Every case of drowning has to be meticulously investigated whether it is due to homicide, accident, natural causes or suicide; this often represents a forensic challenge. Research suggests that this is a method preferred by elderly women.^2,3

Suicide by drowning appears more prevalent near coastal areas or large water reserves (inland lakes and rivers). ⁴ Apparently, only one study on suicide drowning has reported data from the coastal area of a Mediterranean country. ⁵

We know of no previous study of suicides by drowning in Greece, which is a Mediterranean country with an extensive shoreline (coastline: 14,676 km). ⁶ We aimed to fill this gap by studying the characteristics of a sample from a large coastal area of Greece, including several islands in the Aegean Sea.

Materials and methods

We collected data from the Piraeus Department of Forensic Medicine which includes the Piraeus Greater Area (Piraeus is the port of Athens and the largest port in Greece) and several Aegean islands (Aegina, Hydra, Poros, Spetses, Kythira, often Cyclades, Samos and Ikaria). The population of these areas totals some 700,000 individuals (Hellenic Statistical Authority). ⁷

We studied a period of 10 years (2006–2015). Data were retrieved from the victims’ forensic records, including their toxicological analyses and police notes. We documented the use of antidepressants, antipsychotics, anticonvulsants, benzodiazepines, alcohol and illicit drugs: cocaine, opioids, cannabis and amphetamines (the cut-off point for alcohol use was designated at ≥ 0.10 g/100 ml).

We also considered whether there was a variation in the frequency of drowning or any seasonal effect “before” (2006–2010) and “during” (2011–2015) to take account of the severe economic crisis that has been afflicting Greece since 2010.

Cases were categorized as suicides after complete forensic autopsies and police investigations were performed.

Permission for our study was granted by the head of the Piraeus Department of Forensic Medicine, who provided unconditional access to the victims’ forensic records.

For the statistical analysis, we used SPSS 20.0 for Windows and the χ² Fisher exact test for comparison of percentages. The level of p < 0.05 was chosen to indicate statistical significance. We estimated the presence of seasonality for drowning with a Poisson regression variant of the circular normal distribution as modified by Frangakis and Varadhan. ⁸ The relevant statistical analysis was carried out using R stats (R Foundation for Statistical Computing, Vienna, Austria).

Results

During the 10-year period of our study 435 individuals died by suicide: 341 men (78.4%) and 94 women (21.6%). Drowning was chosen by 69 individuals (15.9% of the sample). Mean age: males 52.2 years (SD: 20.3, range: 27–89) and females 58.8 (SD: 15.2, range: 32–84; p = 0.213).

The characteristics of the sample are shown in Table 1; suicide methods are indicated in Table 2, suicide methods “before” and “during” the crisis in Table 3 and seasonality of suicides by drowning in Table 4.

OPEN IN VIEWER

Table 1.

Suicides by drowning (n = 69).

	%
Gender
Males	73.9
Females	26.1
Suicide in sea	92.8
Wells	7.2
Medications a
Antidepressants	8.9
Benzodiazepines	8.9
Antipsychotics	1.5
Anticonvulsants	0
Illicit drugs a
Cocaine	1.5
Opioids	1.5
Cannabis	0
Amphetamines	0
Alcohol a	20.9

OPEN IN VIEWER

Table 2.

Suicide methods 2006–2015 (n = 435).

	%
Hanging	37.2
Jumping from a height	25.3
Drowning	15.9
Self-shooting	12.4
Medication overdose	3.2
Self-immolation	2.3
Run over by train	1.1
Cutting	1.1
Toxic substances	0.7
Other	0.7

OPEN IN VIEWER

Table 3.

Suicide methods before and during crisis.

	2006–2010	2011–2015	p
	%	%
Hanging	40.1	34.1	0.220
Jumping from a height	24.2	26.4
Drowning	18.1	13.5
Self-shooting	11	13.9
Medication overdose	3.1	3.4
Self-immolation	0.4	4.3
Run over by train	0.9	1.4
Toxic substances	0.4	1
Cutting	1.3	1
Other	0.4	1

 

p< 0.05.

OPEN IN VIEWER

Table 4.

Seasonality of suicides by drowning “before” and “during” crisis.

Seasonality	Suicides (n)	RR	95% CIs	p	Peak month
From 2006 to 2010
Drowning	41	1.33	1–3.18	0.81	October
From 2011 to 2015
Drowning	28	1.36	1–4.52	0.85	April (early)

CIs: confidence intervals; RR: relative risk.

Discussion

Rates of suicidal drownings worldwide vary between 2.86% and 10% of suicidal deaths.^2,4,5,9–11 The frequency of the method appears clearly influenced by the proximity and abundance of watery areas. ¹²

According to the literature, suicide by drowning seems to be the preferred means for elderly women ( > 65 years of age).^3,12,13 In our sample, men chose this method three times more often than women. Nevertheless, the females’ mean age in comparison to the males’ was higher but not significantly so.

The sea was, by far, the commonest choice. This was to be expected due to the large coastal area that our study covered. Not one took his/her own life in a swimming pool, river or a lake; however, there are no large rivers or lakes in the area.

Fewer than one-fifth of the cases were receiving psychiatric treatment, mainly with benzodiazepines or antidepressants. Even in the study by Ahlm et al., ¹³ benzodiazepines were the most commonly found (21%). This could suggest that most victims were undiagnosed and/or undertreated and/or noncompliant with pharmacotherapy. However, given that the victims’ psychiatric records (if any) were rarely available, our study cannot provide definitive answers.

Very few individuals had used illicit drugs prior to their death by comparison with those referred by similar studies. Indeed, Gorniak et al. ⁹ reported that 12% of suicide drownings had consumed illicit drugs, a higher number than the 10% referred by Ahlm et al. ¹³

Alcohol seems to be present more often in deaths by drowning – with frequencies reaching 50% – than in suicides by drowning.^9,10,13 Approximately one-fifth of the victims had consumed alcohol prior to their death: this accords with findings of similar studies, particularly the one by Byard et al., ⁴ who recorded a nearly identical number (20.3%). Others report rates of 12% and 27%.^9,10 It has been suggested that when alcohol is present in a drowning death, there is less likelihood that it was a suicide. ⁴

There was a nonsignificant variation in the frequency of drowning from the third place “before” crisis to the fourth place “during” it (third overall). It is not clear why that happened. Perhaps the relative rise of the more violent methods (jumping from a height, self-shooting and self-immolation) at the expense of the less violent ones (hanging and drowning) within the crisis years could imply more severe psychic pain and/or rage.

Prevention of suicide by drowning is challenging given that access to water areas cannot be prohibited. Early detection and appropriate treatment of psychiatric disorders in patients who live near large fresh water or coastal areas appears the only option.

Drowning is an easily accessible suicide method but, when a body is not found, the pain may be unbearable for the relatives, therefore, its study with a view to prevention is important. Greece, with its large coastline, offers a useful area for further research.

Limitations

The study was retrospective, based exclusively on archival material. The sample was not large and drawn from an area with particular characteristics and should not be considered representative of the entire country. Finally, some suicide deaths could have, despite our efforts, been disguised as accidents.