Synergistic effect of chronic pain and non-suicidal self-harm on pain sensitivity

Abstract

Abnormalities in pain processing have been observed in patients with chronic pain conditions and in individuals who engage in self-harm, specifically non-suicidal self-injurious behaviors (NSSI). Both increased and decreased pain sensitivity have been described in chronic pain patients, while decreased pain sensitivity is consistently observed in individuals with NSSI. The objective of the study was to identify the differential effects of chronic pain and NSSI on experimental pain sensitivity, specifically pressure pain threshold, in depressed patients. Moreover, the role that hopelessness may play between depression severity and pain sensitivity was also examined. Depressed patients with and without chronic pain, and with and without lifetime self-harm behaviors were analyzed into four groups. Group 1 (N=42) included depressed patients with both Chronic pain(+) and Self-harm(+), Group 2 (N=53) included depressed patients with Chronic pain(+) but no Self-harm(−), Group 3 (N=64) included depressed patients without Chronic pain(−) but Self-harm(+), and Group 4 (N=81) included depressed patients with neither Chronic pain(−) nor Self-harm(−). Healthy controls (N=45) were also recruited from the community. Depressed patients with both Chronic pain(+) and Self-harm(+) reported higher pressure pain threshold measures as compared to the other groups. Mediation analysis indicated that hopelessness mediates the relationship between depression severity and pressure pain threshold. Our findings suggest that a multiprong approach including adequate mental health services and pain control for depressed patients with comorbid chronic pain and non-suicidal self-harm is needed to yield effective outcomes.

Introduction

Chronic pain, a common health problem, presents a major clinical challenge because of its complex natural history, unclear etiology, and poor response to available therapies. It is defined as pain lasting longer than three or six months that adversely affects an individual’s well-being [58, 80]. Over 100 million people in the United States suffer from chronic pain [34]. Moreover, annual health care costs and lower worker productivity due to pain is estimated at over $600 billion [36].

There is a high comorbidity and strong associations between chronic pain and mental illnesses [15, 82, 84]. This phenomenon is highlighted by the high rates of opioid use and misuse in chronic pain patients [17, 44]. Additionally, the prevalence of one condition co-occurring with the other is higher than the presence of only one condition [10]. Generally, depressed patients report more physical comorbidities, including chronic pain than the general population [5, 10, 84]. On the other hand, mood and anxiety disorders have been shown to be highly prevalent in patients suffering from chronic pain [28]. Comorbid depression and chronic pain leads to greater health care costs and burden including disability and poorer health outcomes [10, 69], and substantial clinical challenges remain for physicians to recognize the high comorbidity and to provide effective treatment [39].

Non-suicidal self-injury (NSSI) is defined as the deliberate, self-inflicted damage to body tissue without suicidal intent [75]. Observed in 1–4% of adults, NSSI is a public health problem with potentially serious medical complications and few effective treatment strategies [19, 51]. Individuals engage in NSSI to relieve overwhelming psychological distress or pain [60, 62], whereby psychological pain is externalized by transforming it into a tangible physical sensation [40]. The exact mechanism through which this transformation of pain is achieved is unclear. However, it is suggested that engaging in self-harm allows the individual to overcome the usual powerful aversion to physical pain [78]. Several studies have observed that individuals engaging in NSSI have lower pain sensitivity [37, 52]. Moreover, NSSI is associated with depression and suicide risk [6, 51]. However, to our knowledge, pain sensitivity in depressed patients with and without NSSI has yet to be examined.

Experimental pain paradigms allow for objective measurements using standardized methodology of administering painful stimuli. Both increases and decreases in pain sensitivity have been reported in individuals suffering from chronic pain [63, 66]. In depressed patients, overall lower pain sensitivity was observed as compared to healthy controls in several sensory modalities [29, 79]. Depressed patients show higher heat pain thresholds as compared to patients with chronic low back pain and healthy controls [71]. However, to our knowledge, there are no data regarding experimental pain sensitivity in patients with comorbid depression and chronic pain. On the other hand, individuals who engage in non-suicidal self-injurious behaviors report consistently higher pain threshold and tolerance and less pain intensity to painful stimulation, indicating less pain sensitivity [52]. Individuals who engage in non-suicidal self-injurious behaviors usually reported moderate to severe symptoms of depression [57].

Depressed patients may report more clinical pain, however, in experimental pain studies, decreased pain sensitivity has been consistently observed [29, 79]. The contrasting findings may be attributed to the type of pain the patients are perceiving. Depressed patients may be more prone to negative perceptions of their chronic pain, which may increase their reports of pain, but experimental studies administer external stimulation of skin (e.g. pressure or thermal stimuli), and the pain may not reach the attentional demand threshold [79]. Cognitive and emotional processing deficits observed in depression may affect pain sensitivity measures. For instance, in patients with acute back pain, hopelessness was found to mediate the relationship between pain intensity reports and depression severity, where pain did not directly affect depression severity, but directly affected hopelessness, which in turn influenced depression [45].

A link between depression and hopelessness has been well established. Hopelessness, described as spontaneous, uncontrollable negative thoughts about the future, is one of the three elements of Beck’s cognitive triad of depression [41]. Moreover, hopelessness has also been found to be inversely related to pain sensitivity [23]. However, it is currently unclear the role that hopelessness may play between the influence of depression severity on pain sensitivity. Understanding the relationship that depression has on pain and the effect of maladaptive cognitive symptoms such as hopelessness may provide further insight for treatment strategies.

We aimed to examine pain sensitivity in depressed patients with chronic pain and NSSI, and suitable controls. We also explored the relationship between depression severity and pain processing and examined a possible mediating role of hopelessness. Our hypotheses were: a) experimental pressure pain threshold will be higher in depressed patients with comorbid chronic pain and NSSI as compared to depressed patients who present only one condition and to healthy controls; b) pressure pain threshold will correlate with depression, psychological pain, hopelessness and anxiety; and c) hopelessness will mediate the relationship between depression and pain sensitivity.

Methods

Participants

We recruited 285 male and female adults from ages 18 to 65 between January 2015 and April 2017 as part of an ongoing study on cognitive and biological measures of self-harm behavior [20]. Subjects were recruited from the psychiatric inpatient units, the psychiatric consultation service and the psychiatric outpatient clinics of the University of Arkansas for Medical Sciences (UAMS) Hospital (groups 1–4). All subjects in groups 1–4 fulfilled DSM (Diagnostic and Statistical Manual of Mental Disorders)-5 [8] criteria for major depressive episode and either major depressive disorder (MDD), bipolar disorder, depression not otherwise specified or adjustment disorder. Exclusion criteria were: a) inability to read, write and speak English; b) inability to provide informed consent; c) history of dementia, neurovascular or neurodegenerative conditions; d) use of opioids within the last month; and e) use of nonsteroidal anti-inflammatory agents within the last week. The UAMS Institutional Review Board approved all procedures. Data is available upon request.

The diagnostic criteria for NSSI disorder was defined by the DSM-5 criteria as self-inflicted acts of injury on more than 5 days in the past year [8]. Subjects with presence of NSSI disorder were categorized in the self-harm behavior groups. Chronic pain was defined as pain lasting for greater than 3 months duration. The four groups included: Group 1 (N=42): depressed patients with both chronic pain(+) and lifetime self-harm behavior(+); Group 2 (N=53): depressed patients with chronic pain(+) but no lifetime self-harm behavior(−); Group 3 (N=64): depressed patients with no chronic pain(−) but with presence of lifetime self-harm behavior(+); Group 4 (N=81): depressed patients with neither chronic pain(−) nor lifetime self-harm behavior(−).

Healthy controls (HC; N=45) were recruited from the Little Rock community. The Structural Clinical Interview for DSM-IV Diagnoses (SCID) [33] was administered to ensure that the HC had no diagnosis of mental illness currently or in the past. In addition to the exclusion criteria for the patients, additional exclusions included current or axis I disorder, lifetime history of suicide attempts, chronic pain lasting more than three months, and a history of NSSI.

Procedure

After written informed consent, participants underwent an evaluation and interview to obtain demographic data, psychiatric and medical history, behavioral ratings, and measurement of pressure pain threshold. Four elements of clinical data were documented: i) psychiatric diagnostic assessments were performed by or directly under the supervision of a board-certified psychiatrist as defined by the DSM-5 criteria [8]; ii) presence of an anxiety disorder including generalized anxiety disorder, post-traumatic stress disorder, social phobia, and panic disorder; iii) presence of an active substance use disorder other than tobacco, defined as either positive urine/serum drug screen or the individual endorsing current drug use (methamphetamines, cannabinoids, phencyclidine, cocaine, alcohol, and non-prescribed opioids or benzodiazepines); and iv) psychotropic medications taken during the past three days.

The Columbia-Suicide Severity Rating Scale (C-SSRS) [67] and Beck Depression Inventory (BDI-II) [14] were used to quantify current suicidal ideation and behavior, as well as depression severity. The Cronbach’s α measured for the suicidal ideation severity subscale as 0.76 in the current study. The other subscales of suicidal ideation (i.e., duration, deterrents, reasons for ideation) as well as the suicidal behavior subscales were ordinal scales and unrelated (i.e., number of aborted vs interrupted attempts), therefore, internal consistency was not measured. The Cronbach’s α for the BDI-II was 0.93. Known risk factors associated with suicide were characterized with the Beck Anxiety Inventory (BAI) [13] and Beck Hopelessness Scale (BHS) [12]. The Cronbach’s α for the BAI and BHS was 0.94 and 0.93, respectively. Physical and psychological pain were measured with the Physical and Psychological Pain Scale [64]. For physical and psychological pain, the Cronbach’s α scores were 0.92 and 0.88, respectively. The Self-Harm Inventory, a 22 item self-reported scale, was used to measure self-harm behaviors [72]. The Cronbach’s α for the Self-Harm Inventory was 0.79 in the current study.

Pressure pain threshold was measured using standard procedures [9]. Patients were placed supine, and a point five inches below the patella in the medial border of the tibia was identified in both legs. Pressure was applied using a hand-held gauge with a 1-cm² rubber tip (Wagner Instruments, Greenwich, CT, USA) alternately to the left and the right tibia for a total of three times per side. Participants were instructed to verbally report when their sensations changed from pressure to pain or discomfort. A mean tender point threshold (lb) was calculated from the six measurements [9].

We calculated our sample size from our previous studies, using means and pooled standard deviation from 3 patient groups and a healthy control group [22]. For the current study, our sample size calculation indicate that 200 participants would yield 90.3% power to detect a difference in pressure pain threshold [22]. Also, for current physical and psychological pain measures, 72 participants would provide 90.6% power and 24 participants would provide 94.8% power to detect differences, respectively [22].

Statistical Analyses

Descriptive statistics were calculated. An analysis of variance (ANOVA) was used to compare mean differences in continuous variables between all 5 groups including: Chronic pain(+) Self-harm(+); Chronic pain(+) Self-harm(−); Chronic pain(−) Self-harm(+); Chronic pain(−) Self-harm(−); and HC. Tukey’s HSD post-hoc analysis was used to determine mean differences between 2 groups. Analysis of covariance (ANCOVA) were performed using a univariate linear regression model with pressure pain threshold, current physical pain ratings and current psychological pain ratings as continuous dependent variables, controlling for age, depression severity, gender and education level. Separate ANCOVAs were performed for the four patient groups (excluding HC) with pressure pain threshold, current physical pain ratings, current psychological pain ratings, and self-harm scores as continuous dependent variables, controlling for age, depression severity, gender and education level.

Pearson correlations were performed for correlations between pressure pain threshold, current physical pain ratings, current psychological pain ratings, Self-Harm Inventory, depression severity, anxiety severity, hopelessness scores, and number of suicide attempts in the patient groups. Depression severity (BDI-II) was examined for the last two weeks previous to the study visit, and hopelessness (BHS) was assessed based on the participant’s symptoms from one week prior to the study visit. Pressure pain threshold was measured at the moment of the study visit. A mediation analysis was used to test the direct effect of depression severity on pressure pain threshold, and whether hopelessness mediated this effect. All statistical analyses were conducted using SPSS 27 (SPSS Inc., Chicago, IL, USA) and PROCESS macro [42].

Results

Demographic and clinical characteristics of the study participants are detailed in Table 1. There was a significant difference in age between the groups (F=16.84; p<.001), with Group 2: Chronic pain(+) Self-harm(−) representing the oldest age group. The majority of the patients were receiving antidepressant medications. The four patient groups did not significantly differ in their depression scores (F=0.78, p=0.51), with a mean BDI-II score of 36.5 ± 11.4, indicating severe depression. The mean BAI score was 29.6 ± 13.0, representing severe anxiety. There was a significant difference in anxiety levels between the depressed groups (F=3.19, p=.024). Post-hoc analysis revealed higher anxiety scores in Group 1: Chronic pain(+) Self-harm(+) than in Group 4: Chronic pain(−) Self-harm(−) (p=.013). No significant group differences were observed regarding hopelessness.

Table 1.

Demographic and clinical characteristics of study population

	Group 1 Chronic pain (+) Self-harm(+)	Group 2 Chronic pain(+) Self-harm(−)	Group 3 Chronic pain(−) Self-harm(+)	Group 4 Chronic pain(−) Self-harm(−)	Group 5 Healthy controls
	n=42 Mean ± SD	n=53 Mean ± SD	n=64 Mean ± SD	n=81 Mean ± SD	n=45 Mean ± SD	p-value	95% CI
Age	35.40 ± 11.16b	43.70 ± 10.32a,c,d	29.75 ± 8.32b,d	37.12 ± 12.73b,c	29.53 ± 8.17b,d	p<.001	33.90–36.62
Sex (female %)	27 (64.3%)	32 (60.4%)	45 (71.4%)	50 (61.7%)	26 (57.8%)	p=.403
Education (years)	13.00 ± 2.55	13.23 ± 2.47	13.33 ± 1.92	13.10 ± 2.90	17.47 ± 1.89 a,b,c,d	p<.001	13.49–14.16
Number of suicide attempts	4.95 ± 4.16	4.94 ± 8.78	3.84 ± 4.03	4.57 ± 5.22	N/A	p=.668	3.77–5.24
Ethnicity
White (%)	31 (73.8%)	38 (71.7%)	46 (74.2%)	55 (68.8%)	35 (77.8%)
Black (%)	8 (19.0%)	14 (26.4%)	12 (19.4%)	25 (31.3%)	7 (15.6%)
Other (%)	3 (7.1%)	1 (1.9%)	4 (6.5%)	0 (0%)	3 (6.7%)
Marital Status
Married (%)	15 (35.7%)	9 (17.0%)	12 (19.0%)	10 (12.5%)	22 (51.2%)
Divorced (%)	2 (4.8%)	0 (0%)	0 (0%)	3 (3.8%)	0 (0%)
Widowed %)	10 (23.8%)	22 (41.5%)	11 (17.5%)	19 (23.8%)	4 (9.3%)
Single (%)	2 (4.8%)	6 (11.3%)	6 (9.5%)	8 (10.0%)	0 (0%)
Never married (%)	13 (31.0%)	16 (30.2%)	34 (54.0%)	38 (47.5%)	17 (39.5%)
Antidepressant use (yes %)	28 (68.3%)	40 (75.5%)	26 (44.1%)	35 (45.5%)	N/A	p=.056
Depression severity (BDI)	35.52 ± 10.69	38.11 ± 11.39	37.22 ± 10.11	35.24 ± 12.40	2.76 ± 3.39 a,b,c,d	p<.001	1.75–3.78
Current suicidal ideation severity	1.54 ± 1.48	2.14 ± 1.76	1.54 ± 1.35	0.98 ± 1.47b	N/A	p=.004	1.34–1.80
Anxiety severity (BAI)	33.57 ± 12.40d	28.85 ± 13.26	30.02 ± 13.20	26.11 ± 12.30a	4.38 ± 6.43 a,b,c,d	p<.001	23.38–26.92
Hopelessness (BHS)	9.85 ± 2.53	10.83 ± 2.52	10.79 ± 2.51	10.57 ± 1.95	8.40 ± 0.92 a,b,c,d	p<.001	9.96–10.48
Diagnosis
Major depression (%)	21 (67.7%)	31 (70.5%)	29 (61.7%)	43 (63.2%)	0 (0%)
Bipolar Disorder (%)	10 (32.3%)	11 (25.0%)	14 (29.8%)	15 (22.1%)	0 (0%)
Depression NOS (%)	0 (0%)	1 (2.3%)	4 (8.5%)	7 (10.3%)	0 (0%)
Psychosis (%)	0 (0%)	1 (2.3%)	0 (0%)	1 (3.3%)	0 (0%)
Anxiety disorder (%)	3 (0%)	5 (0%)	3 (0%)	12 (0%)	0 (0%)
Substance use (%)	29 (70.7%)	25 (47.2%)	34 (57.6%)	41 (53.2%)	0 (0%)

^aStatistically significant from Chronic pain(+) Self-harm(+)

^bStatistically significant from Chronic pain(+) Self-harm(−)

^cStatistically significant from Chronic pain(−) Self-harm(+)

^dStatistically significant from Chronic pain(−) Self-harm(−)

Table 2 shows pain measures of the study sample between the 5 groups. Group 1: Chronic pain(+) Self-harm(+) showed the highest pressure pain threshold at 14.6 ± 4.9 lbs (F=6.291; p<.001). Between the 4 patient groups, there were significant differences in pressure pain threshold (F=3.364; p<.019). Groups 2 and 3 were higher than the depressed and healthy controls but the differences were not significant. There were no significant differences between the groups regarding current psychological pain (F=1.284; p=.281). Post hoc analyses confirmed group distribution regarding presence of chronic pain and self-harm.

Table 2.

Pain measures in the study population

	Chronic pain(+) Self-harm(+)	Chronic pain(+) Self-harm(−)	Chronic pain(−) Self-harm(+)	Chronic pain(−) Self-harm(−)	Healthy controls
	n=42 Mean ± SD	n=53 Mean ± SD	n=63 Mean ± SD	n=81 Mean ± SD	n=45 Mean ± SD	p-value	95% CI
Pressure pain threshold	14.56 ± 4.86	11.95 ± 4.16	12.31 ± 4.91	11.74 ± 5.35a	9.67 ± 4.17a,b,c,d	p<.001	11.42–12.58
Current physical pain	4.88 ± 2.76	6.13 ± 2.50	1.86 ± 2.48a,b	1.91 ± 2.49a,b	0.53 ± 1.18b	p<.001	2.54–3.26
Current psychological pain	7.05 ± 2.53	7.15 ± 2.91	5.95 ± 3.15	6.13 ± 3.22	0.33 ± 0.87	p=.114	5.08–5.92
Self-Harm Inventory score	10.09 ± 4.31	6.33 ± 3.66a,c	10.11 ± 3.71b	5.66 ± 3.81a,c	N/A	p<.001	7.26–8.65

^aStatistically significant from Chronic pain(+) Self-harm(+)

^bStatistically significant from Chronic pain(+) Self-harm(−)

^cStatistically significant from Chronic pain(−) Self-harm(+)

^dStatistically significant from Chronic pain(−) Self-harm(−)

Table 3 outlines the correlations between the variables of interest. Pain threshold was negatively correlated with depression severity (r=−.135, p=.037), anxiety severity (r=−.154, p=.017), and hopelessness (r=−.225, p<.001), but not physical (r=−.123, p=.079) or psychological pain (r=−.111, p=.083).

Table 3.

Correlations between pain ratings and clinical variables

Variables	Pressure pain threshold	Physical pain	Psychological pain	Self-Harm Inventory score	Depression	Anxiety	Hopelessness	# Suicide attempts
Pressure pain threshold	1.00
Physical pain	−.123	1.00
Psychological pain	−.111	.358**	1.00
Self-Harm Inventory score	−.005	.145	.222**	1.00
Depression	−.135*	.227**	.474**	.438**	1.00
Anxiety	−.154*	.249**	.329**	.293**	.522**	1.00
Hopelessness	−.225**	.072	.128*	.033	.213**	.109	1.00
# Suicide attempts	.012	.167**	.088	.283**	.109	.038	.050	1.00

^*Significant at .05 level (2-tailed);

^**Significant at .01 level (2-tailed); NSSI: non-suicidal self-injurious behaviors

The mediation analysis revealed that the association between depression severity and pressure pain threshold was mediated by hopelessness (Figure 1). The direct effect of depression severity was found to positively predict hopelessness (β=.213, p<.001), and the direct effect of hopelessness negatively predicted pressure pain threshold (β=−0.203, p=.002). The total effect of depression severity was found to predict pressure pain threshold (β=−0.144, p=.027) (Figure 1A). When depression severity and hopelessness were jointly included in the same mediation model to predict pressure pain threshold, the significance of depression severity disappeared (β=−.101, p=.122), indicating that hopelessness had significant mediation effects between depression severity and pressure pain threshold (Figure 1B). The indirect effect (−.0190) was statistically significant at 95% confidence interval (−.0384 to 0.0052).

Figure 1.

Proposed model of hopelessness mediating the effect of depression on pressure pain threshold. A. Depression predicts negatively pressure pain threshold. B. Incorporation of hopelessness in the model shows the mediation between depression and pressure pain threshold.

Discussion

Our main findings were: a) relative to the other three patient groups, depressed patients with both chronic pain and lifetime NSSI showed the highest pressure pain threshold; b) pressure pain threshold was negatively correlated with depression, hopelessness and anxiety; c) higher depression severity predicted lower pain threshold, which was mediated by hopelessness.

Experimental studies have shown both increased and decreased sensitivity to pain in individuals suffering from chronic pain. Patients with chronic low back pain (CLBP) exhibit lower pressure pain detection threshold in areas anatomically related and unrelated to the lumbar spine [32, 47]. Central sensitization is believed to play a predominant role [55], with hyperresponsive central nociceptors amplifying action potential output. In effect, the malleable nature of the receptive fields of somatosensory neurons is distorted, resulting in the sensation of pain in response to previously innocuous stimuli. Thus, the experience of pain begins to occur independently of the presence, strength, and duration of noxious peripheral stimuli, resulting in diffuse, globalized hypersensitivity to pain that extends beyond the localized area of damaged tissue [87].

On the other hand, a number of studies have found the opposite to be true. For instance, CLBP patients also show higher pain pressure threshold for both pressure and electrical pain stimuli, as well as higher pain tolerance for pressure stimuli [66], and higher pain thresholds for both radiant heat and uncomfortably loud noises [26, 71]. These findings have been explained with the adaptation level theory, which postulates that chronic pain patients gauge experimental pain in the context of their prior experiences with pain. Given that experimental pain stimuli are rarely as noxious as the intense, daily discomfort and unpleasant medical procedures commonly experienced by chronic pain patients, these individuals often demonstrate higher pain thresholds in experimental settings.

In contrast to the mixed findings in chronic pain patients, individuals with NSSI are consistently shown to have lower pain sensitivity in both experimental and observational studies. Up to a half of adult self-injurers and 80% of adolescent self-injurers report feeling minimal to no pain when engaging in self-injurious behavior [38]. Individuals with a history of NSSI exhibit greater pressure pain thresholds and endure pain for longer than healthy controls [43]. Moreover, within NSSI patients, experimental pain threshold correlated positively with number of years of engaging in self-injury. Similar findings of greater pain threshold and endurance were found in both adolescents and adults with a history of NSSI [35, 38]. In sum, our findings are consistent with previous studies showing decreased pain sensitivity in patients with NSSI. Our findings also show that individuals with both NSSI and chronic pain have a higher pain threshold than those with either condition alone.

Several theories have been proposed to explain altered pain perception in patients with a history of NSSI. Repeated exposure to painful stimuli may result in diminished sensitivity to pain over time via the process of habituation. This theory is supported by studies showing a positive correlation between pain threshold and duration/frequency of NSSI [43]. On the other hand, physical pain may be inflicted to provide relief from distressing dissociative states. However, it is also possible that the reverse is true, such that dissociative symptoms obscure the feeling of pain and reduce the saliency of noxious physical stimuli [38]. Finally, severe self-criticism in patients with NSSI may manifest as an increased willingness to undergo pain as a form of deserved self-punishment [11, 35].

A growing body of evidence reveals a considerable overlap between the subjective response to physical and emotional or psychological pain and the shared neural representation in the “pain network” including the dorsal anterior cingulate cortex and anterior insula in healthy individuals [30, 54] and depressed patients [21]. However, the increased activation following painful stimuli in these brain regions seen in healthy individuals is decreased or absent in individuals with NSSI [16, 73]. This suggests that self-injurers may have altered affective processing of painful experiences. This decrease in affective processing of painful stimuli among NSSI individuals may be linked to the deficits in emotional self-regulation frequently observed in this group [16].

History of NSSI is linked to increased suicide risk. It has been theorized that engaging in self-injury leads to an acquired capability for suicide by habituating individuals to the fear and pain involved in executing an act of suicide [86]. History of prior suicide attempts in university students is linked to increased NSSI frequency, greater number of methods used, and decreased subjective pain during self-injury [3]. Moreover, a stronger association between NSSI frequency and history of suicide attempts was observed in subjects with reduced pain sensitivity. These findings were echoed in a meta-analysis that found that the two strongest predictors of suicide attempt in these individuals, after suicidal ideation, were frequency of NSSI and number of methods used [85]. This is consistent with the notion of acquired capability for suicide in that greater NSSI frequency and diversity of methods lead, to a more powerful erosion of the fear and pain that normally serves as a barrier to suicide [50, 61, 83]. Additionally, chronic pain provides the motivation for suicide – the desire to escape from an intolerable situation. When these two factors are combined, the synergistic effect may be a significantly higher cumulative risk of suicide than that seen in either condition alone. Not surprisingly, there is a considerable increased risk for suicide in individuals suffering from chronic pain, independently of depression or opioid use [18, 24, 46, 77]. Parallels can be drawn to NSSI individuals, who demonstrate increased suicide risk with reduced pain sensitivity, which in turn correlates to greater frequency, intensity, and duration of NSSI. As such, the high risk of co-occurrence of chronic pain and intentional self-harm has been emphasized [25]. The compounded increase in pain threshold in patients with chronic pain and NSSI underscores this concern. Our results indicate a strong association between NSSI and number of suicide attempts. It is possible that a similar mechanism is at play in chronic pain patients, such that enduring high intensity pain for a prolonged period of time results in habituation to painful stimuli and reduced hesitancy to carry out suicide.

Hopelessness, wish to escape from pain, catastrophic thinking, and poor problem-solving abilities contribute to increased suicide risk among chronic pain sufferers [68]. The results of our mediation analysis indicate that depression severity positively predicts hopelessness, which in turn, negatively predicts pain threshold. These findings are echoed in the literature, which has consistently shown a strong positive correlation (ranging from r = 0.52 to 0.68) between depression severity and hopelessness scores [1, 2, 76, 81]. Depression severity has been shown to predict hopelessness [59, 70], although the opposite direction has also been reported [31, 45]. Orbach and colleagues reported that hopelessness correlates positively with thermal pain tolerance and threshold among suicidal and nonsuicidal depressed adolescents [65]. The discrepancy with our findings may be explained because all our depressed groups suffered from severe depression. Nonetheless, depression and hopelessness are mutually reinforcing phenomena. Among depressed adolescents, hopelessness is positively correlated with tolerance for thermal pain [65].

Of note, the temporal relationship between the variables in our model substantiate the validity of the mediation analysis results. BDI-II collected information regarding depression symptoms during the two weeks prior to the study visit, BHS examined hopelessness severity during the week prior to the study visit, while measurement of pain threshold was performed during the study visit.

Severe anxiety as measured by the Beck Anxiety Inventory was observed in all four patient study groups. The high comorbidity between anxiety disorders and major depression is well established [4, 48, 49], corroborated in our results. Moreover, patients who had lower anxiety severity reported higher threshold for pressure pain, suggesting that anxiety increased pain sensitivity. The literature shows that patients with depression report lower pain sensitivity [79], patients suffering from anxiety report increased pain sensitivity [53, 56, 74]. In particular, patients with chronic pain and symptoms of depression and anxiety, report more severe pain [56]. The presence of pain may lead to cognitive and behavioral changes including increased attention and pessimism towards the pain [7, 27]. This is in line with our findings that hopelessness mediated the association between severity of depression and pain threshold. That is, higher levels of depression severity predicted lower pain threshold, and this association was affected by increased hopelessness, suggestive of negative thinking about the future.

The main clinical implication of our findings lies in highlighting the role of pain in depressed patients with comorbid chronic pain and NSSI. Use of adequate strategies for pain control and anxiety management, has the potential to decrease suffering and improve function in these patients, who share several commonalities with patients with fibromyalgia. Additionally, this patient population may be particularly vulnerable to suicide, which further emphasizes the need for early identification and a multidisciplinary approach.

One strength of this study is that patients were free of opioids and nonsteroidal anti-inflammatory agents, which allowed for a cleaner interpretation of results. Also, the use of a experimental pain paradigm allowed for an objective and standardized measure of the individual’s pain sensitivity.

A limitation of this study is the absence of biological markers of pain regulation. We used the Self-Harm Inventory, which measures lifetime NSSI, as a trait, our results may have been more precise with a more discrete and recent examination interval. Additionally, the majority of patients were receiving antidepressant medications, which may influence pain sensitivity. However, given the elevated depression severity in all four patient groups, withholding medication would have been unethical. Finally, our study only examined the influence of depression and hopelessness on pain sensitivity. Given the complexity of pain modulation, addition of other variables would add a more complete picture.

In conclusion, our results indicate that in depressed patients, concurrent chronic pain and non-suicidal self-harm has a compounding effect on reducing pain sensitivity as compared to the presence of either alone. Both, depression and hopelessness seem to have a role modulating pain sensitivity. Our findings indicate that a multiprong approach for depressed patients that addresses mental health, including NSSI behaviors, as well as, comorbid chronic pain is needed.