Of the 9384 participants that received the questionnaire, 4855 returned it (51.7%) and 4810 (51.3%) responded to the self-harm questions. The mean age of respondents at the time the questionnaire was completed was 16
years and 8
months (standard deviation (SD) 2.9
months). Those who returned the questionnaire were more likely than non-respondents to be female, have a mother in a non-manual social class, and to have relatively high educational qualifications (see Table for more details).
Comparison of responders and non-responders by key background variables
The total number of respondents who had ever self-harmed was 905 (18.8%, 95% CI 17.7% to 19.9%). Prevalence was higher in females (25.6%) than males (9.1%) and overall prevalence was not substantially different in imputed models taking account of missing data due to questionnaire non-response - the estimated lifetime prevalence for self-harm was 18.4% (95% CIs 17.3%-19.6%) (see Additional file 3
There was strong evidence of an association (p
0.001) between self-harm and both gender and current negative mood (see Table ). Females were at almost 3.5 times more likely to report self-harm than males and individuals with negative mood symptoms were more than 5 times greater risk. The risk of self-harm was also somewhat higher amongst among respondents whose mother scored 13+ on the EPDS five years earlier (OR
1.48 [1.17, 1.86], p
0.001), those whose mothers were of manual social class (OR
1.46 [1.12, 1.90], p
0.005) and those achieving grades A*-C in less than five GCSE/GNVQ exams (1.20 [1.03, 1.34], p
0.025). Finally there was no strong evidence of an association between self-harm and either maternal education or the young person’s ethnicity. Comparison of these results showed good agreement across the two imputation samples considered and we were confident that the estimates were not being overly impacted by noise due to imputation of the outcome (see Additional file 3
Association of ever having self-harmed at age 16 with key demographic variables1,2
Table gives details of frequency, actions and consequences of self-harm behaviour. Just under one third of those who had ever self-harmed, which equated to 5.7% of the total sample, said that they had ever seriously wanted to kill themselves while self-harming, with no gender difference. A majority of those who had self-harmed had done so in the past year, 7.8% in the past week. The single method most often used the last time respondents self-harmed was cutting for both genders (64.1%). This was followed by self-battery (8.3%) - which included actions such as biting, pulling hair and head butting walls - while overdosing (4.1%) and burning oneself (1.9%) were less common methods. Just over 20% engaged in more than one method of self-harm at once. Only 12.4% of respondents had sought medical help following the most recent self-harm act, with no gender difference. More than half of those who had self-harmed had done so more than once in the past year, and 25.3% had done so 6 or more times. Approximately half felt better after self-harming the most recent time, with boys more likely to feel neither better nor worse compared to girls. As the frequency of self-harm increased, so did the likelihood that the person would feel better subsequently (see Figure ). Compared to those who had self-harmed once, those who had self-harmed more than 10 times were three times as likely to feel better rather than worse or the same following self-harm (OR
3.02, 95%CI: 1.93-4.72), (p value for trend across categories
Description of self-harm actions and consequences among the 905 participants who reported self-harm
Association between frequency of self-harm and how respondent felt subsequently.
Figure shows the reasons given for the most recent self-harm episode. The most common reasons were to gain relief from terrible feelings (64.4%) and a desire to punish oneself (41.0%). Of the whole sample, 25.3% gave ‘wanted to die’ as a reason (25% males and 25.4% females). Females were more likely than males to select desire to punish self (43.8% compared to 29.4%, χ2
0.001) and to gain relief from terrible feelings (66.7% compared to 55.9%, χ2
0.003), whereas males were more likely than females to choose ‘superficial reasons’ such as “I was curious” or “because my friend was” (7.8% compared to 2.2%, χ2
0.001). There were no other large gender differences in reasons given.
Reasons given for the most recent time participants self-harmed.
Table compares those who had wanted to die the most recent time they had self-harmed with those who had not given this as a reason. Those who wanted to die were much more likely to have poorer educational qualifications and a greater number of depressive symptoms on the SMFQ at age 16, and slightly more likely to have a mother in a manual social class at the time of their birth. This group were also much more likely to have taken an overdose the most recent time they self-harmed, whereas there was no difference between the groups regarding whether they had cut themselves. Those who had wanted to die were also less likely to feel better after the last episode of self-harm, and much more likely to have sought medical help (30.1% vs. 6.4%).
Comparison of those whose last episode of self-harm was associated with vs. without a desire to die: their characteristics, and the characteristics of the act and its consequences1
Table shows the percentage of respondents who answered yes to the series of questions on suicidal thoughts and plans. Just under one quarter of all respondents had ever felt that life was not worth living, with 15.8% (95% CIs 14.5%-16.6%) having thought of killing themselves, and 4.3% ever having made plans to kill themselves (95% CIs 3.7-4.8). As shown in Figure , females were more than twice as likely as males to have ever self-harmed, felt life was not worth living, wished they were dead and away from it all, had thoughts of killing themselves and made plans to kill themselves (all χ2 test p values <0.001). Large differences in all the questions relating to suicidal thoughts and plans were visible between those who had self-harmed with desire to die the most recent time, those who had self-harmed without desire to die the most recent time and those who have never self-harmed (Table ). Approximately 90% of those who had self-harmed with desire to die and over one third of those who had self-harmed without desire to die had ever thought of killing themselves, compared to 7.7% of the never self-harmed group. And over half of those who had self-harmed with desire to die had ever made plans to kill themselves, compared to 8.7% of those who had self-harmed without desire to die, and just 0.7% in the never self-harmed group.
Prevalence of suicidal thoughts/plans in those whose most recent episode of self-harm was associated with a desire to die, those who self-harmed with no desire to die, and those who have never self-harmed1
Percentage of respondents who have ever self harmed or experienced suicidal thoughts, by gender.
Figure shows the relationship between frequency of self-harm and suicidal thoughts or plans. Compared to those who had self-harmed once, those who had self-harmed 2–5 times were nearly twice as likely to have thought of killing themselves (OR
1.86, 95%CIs: 1.34-2.58), those who had self-harmed 6–10 times were more than three times as likely to have done (OR
3.08, 95%CIs: 1.84, 5.17) and those who had self-harmed more than ten times were nearly five times as likely to have done (OR
4.93, 95%CIs: 3.06, 7.94). Similarly, compared to those who had self-harmed once, those who had self-harmed 2–5 times were more likely to have made plans to kill themselves (OR: 2.59, 95%CIs: 1.61, 4.17), those who had self-harmed 6–10 times were more than three times more likely to have done (OR:3.18, 95%CIs: 1.69, 6.00) and those who had self-harmed more than 10 times were approximately eight times more likely to have done (OR: 8.64, 95%CIs: 5.12, 14.60).
Comparison of self-harm with desire to die and self-harm with no desire to die the most recent time by background variables, and self-harm actions and consequences.