Depression, anxiety, health-related quality of life and pain in patients with chronic fibromyalgia and neuropathic pain

1. Introduction 

Chronic pain is often associated with anxiety and depression (Bair et al., 2003), resulting in a low health-related quality of life (Otto et al., 2007, Gustorff et al., 2008, Pagano et al., 2004, Dworkin and Caligor, 1988, Lautenbacher and Krieg, 1994). The mechanisms underlying the association between mental symptoms and chronic pain are, however, not clear (Dersh et al., 2002), but abnormalities in pain and mood modulating systems in the brain and spinal cord have been suggested as a common mechanism (Stahl and Briley, 2004, Stahl, 2003). The monoamines serotonin and noradrenaline originate in raphe nuclei of the brainstem and project to the forebrain and the spinal cord and both are known to influence mood and pain processing (Suzuki and Dickenson, 2005, Suzuki et al., 2004, Porreca et al., 2002, Millan, 2002). Importantly, antidepressants such as TCAs and SNRIs used to treat both depression (National Collaborating Centre for Mental Health, 2004) and anxiety disorders (van Balkom et al., 1997, Kapczinski et al., 2003) have a modulating effect on neuropathic pain (Bymaster et al., 2001, Sindrup and Jensen, 1999, Max et al., 1987, Goldstein et al., 2005) and fibromyalgia (Arnold et al., 2004, Arnold et al., 2005, Arnold, 2006).

Previous studies have shown that patients with fibromyalgia (FM) display high levels of both lifetime and current prevalence rates of anxiety and depression compared with controls (Epstein et al., 1999, Ahles et al., 1991, Krag et al., 1994). Also, patients with neuropathic pain (NP) may present with elevated anxiety and depression scores (Nicholson and Verma, 2004, McCarberg, 2003, Haythornthwaite et al., 1991, Haythornthwaite and Benrud-Larson, 2000), but to a lesser extent than described for patients with FM (Gustorff et al., 2008, Mongini et al., 2007, Berger et al., 2004, Rasmussen et al., 2004b).

The aetiology of FM is not known. According to the 1990 criteria of the American College of Rheumatology (Wolfe et al., 1990) FM is defined by (1) widespread pain in combination with (2) tenderness at 11 or more of 18 specific tender point sites. Furthermore, a series of symptoms including fatigue, restless legs, postural problems, dizziness, and bladder problems are also seen (Jones et al., 2008). FM may be part of a group of conditions termed affective spectrum disorders (ASD), which include, e.g., major depressive disorder, attention-deficit disorder, bulimia nervosa, cataplexy, dysthymic disorder, generalized anxiety disorder, irritable bowel syndrome, migraine, obsessive–compulsive disorder, panic disorder, posttraumatic stress disorder, premenstrual dysphoric disorder, and social phobia (Hudson et al., 2003) and share one or more physiological abnormalities (Bradley, 2008, Hudson et al., 2003). It has been suggested that these conditions are characterized by monoamine changes in the brain and the hypothalamic–pituitary–adrenal axis (HPA) leading to the clinical symptoms seen in ASD. As such, ASD including FM may be seen as a reaction to sustained stress with high levels of cortisol production and a certain genetic or psychosocial vulnerability (Bradley, 2008). On this background, a high correlation between mental disorder and fibromyalgia is expected. Other studies have mainly pointed to a disturbance in the pain processing systems as the main pathophysiological factor behind FM (McLean and Clauw, 2005, Ahles et al., 1991).

2. Aim of the study 

In the present study, we combined these two perspectives by comparing FM with NP, which is a pain condition with a clear evidence of abnormal pain processing. Peripheral neuropathic pain is a consequence of a lesion or disease affecting the somatosensory system (Treede et al., 2007) and is characterized by localized hypersensitivity within the territory of the damaged nervous system due to an increased input of noxious information from the periphery (Finnerup et al., 2007, Woolf, 2004, Jensen and Baron, 2003). It has been suggested that this increased input to the CNS from the periphery may alter the normal descending inhibitory control and be one of the mechanisms responsible for the pain (Woolf and Mannion, 1999, Witting et al., 2003). Patients with FM, on the other hand, may have a primary disturbance in the brain leading to a generalized hypersensitivity to pain (McLean and Clauw, 2005, Arnold et al., 2004, Arnold et al., 2005, Usui et al., 2006, Bennett, 2005) and a variety of other symptoms from other organ systems (Jones et al., 2008). We would, therefore, assume there is a correlation between pain and mental symptoms in FM. In NP patients who have a localized disturbance driven from the periphery, we would expect no such association between pain and mental symptoms unless the noxious barrage from the nerve injury site has engaged the higher cerebral structures involved in mood and emotional states. In the present study, we examined the presence of depression, anxiety, and health-related quality of life together with pain in patients with FM and NP compared with healthy controls.

3. Methods 

4. Results 

4.1. Subjects and treatment 

Thirty NP patients, 28 patients with FM, and 26 controls, all Caucasian, were included. Table 1 shows a summary of the clinical characteristics of the patients and the healthy controls. The primary diagnosis in the NP group was nerve injury following operation , further diagnoses were nerve compression , traumatic nerve injury , neuropathy , and neurinoma . In the FM group there was no obvious aetiology for the FM syndrome in any of the patients. No patients received non-permitted treatment.

Table 1. Summary of clinical characteristics.

		Patients with neuropathic pain	Patients with fibromyalgia	Healthy controls
	Age (mean, SD)a	51.8 (13.1)	48.3 (10.2)	45 (13.7)
	Gender	14 F, 16 M	26 F, 2 M	21 F, 5 M
	Married	87%	86%	68%
	Pain duration from diagnosis (mean, SD)	6.2 (6.7)	3 (1.8)	0
	Pain on NRSb (median, percentile)	5 (5–7)	6 (5–5.6)	< 4
	Pinprick-induced hyperalgesia (von Frey 5.88)	n=30	ND	ND
	Brush-evoked allodynia	n=18	ND	ND
	Paracetamol max 4g/day (no.)	n=2	n=1	n=1
	Aspirin 75mg/day (no.)	n=3	n=0	n=0

ND=not done.

a No significant differences between groups. b No significant differences between patients.

4.2. Mental distress 

Chronic pain patients had significantly higher scores on all mental distress measures on SCL-92 than healthy controls (Table S1, see the online version at doi:10.1016/j.ejpain.2009.03.010). Furthermore, patients with FM had significantly more anxiety, depression, somatization, obsession–compulsion, and generally more mental distress (Global Severity Index) than NP patients (Fig. 1). A gender-stratified analysis only showed a significant difference between genders in the subgroup of somatization. Men had lower scores than women.

View Large Image Download to PowerPoint

Fig. 1. Symptom Check List-92 self-rated by patients with fibromyalgia and neuropathic pain compared with pain-free age- and gender-matched healthy controls. *=Significant differences were found between fibromyalgia and neuropathic pain. Significant differences were found between patients (fibromyalgia and neuropathic pain) and controls on all subscale measures.

Table S1. Mental scores and comparison of the three groups.

	Groups/factors	1. NP patients		2. Controls		3. FM patients		p-value (1–2, 1-3, 2-3)
		Median	Percentiles	Median	Percentiles	Median	Percentiles
	SCL-92
	Global severity index	0.48	0.2–1.0	0.14	0.1–0.2	1.1	0.5–1.3	(∗∗∗ ∗∗ ∗∗∗)
	Depression	0.62	0.2–0.9	0.15	0.1–0.6	1.0	0.6–1.9	(∗ ∗∗ ∗∗∗)
	Anxiety	0.5	0.2–0.8	0.2	0.1–0.4	0.9	0.5–1.4	(∗∗ ∗∗ ∗∗∗)
	Somatization	0.9	0.4–1.2	0.2	0.1–0.4	2	1.4–2.3	(∗∗∗ ∗∗∗ ∗∗∗)
	Interpersonal sensitivity	0.4	0.2–0.7	0.1	0.1–0.4	0.7	0.4–0.9	(∗∗ NS ∗∗∗)
	Phobic anxiety	0	0.0–0.3	0	0.1–0.1	0.1	0.0–0.6	(∗∗ NS ∗∗)
	Obsession–compulsion	0.6	0.3–1.0	0.3	0.1–0.6	1.6	0.5–2.1	(NS ∗∗ ∗∗∗)
	Hostility	0.3	0.2–0.7	0.2	0–0.2	0.5	0.3–0.7	(∗∗∗ NS ∗∗∗)
	Paranoid ideation	0.2	0.0–0.5	0	0–0.2	0.5	0.2–1.0	(∗∗ NS ∗∗∗)
	Psychoticism	0.1	0.0–0.4	0	0–0.1	0.3	0.1–0.7	(∗∗ NS ∗∗∗)
	MDI	3	3–5	2	1–3	7	4–9	(NS ∗∗ ∗∗∗)
	HDS	7	4–10	2	1–4	10	8–14.5	(∗∗ ∗∗ ∗∗∗)
	GAD–10	7	4–11	2	1–5.5	14	10–18	(∗∗ ∗∗ ∗∗∗)
	HAS	6	2–10	1.5	0–4	13	10–16.5	(∗∗ ∗∗∗ ∗∗∗)
	SF-36
	Physical functioning	65	40–80	100	95–100	55	40–65	(∗∗∗ NS ∗∗∗)
	Role function–physical aspect	50	0–100	100	100–100	0	0–25	(∗∗∗ ∗∗∗ ∗∗∗)
	Bodily pain	41	27–51	92	84–100	31	22–41	(∗∗∗ NS ∗∗∗)
	General health perception	57	50–77	91	82–92	40	25–60	(∗∗∗ ∗∗ ∗∗∗)
	Vitality	60	40–80	75	70–85	30	18–45	(∗∗ ∗∗∗ ∗∗∗)
	Social function	87.5	63–100	100	100–100	56	50–81	(∗∗∗ ∗∗ ∗∗∗)
	Role function–emotional aspect	83.3	67–100	100	67–100	100	33–100
	Mental health	78	64–88	84	76–92	74	56–88	(∗∗∗ ∗∗ ∗∗∗)

,,,NS=not significant, NP patients=Patients with neuropathic pain, FM patietns=Patients with fibromyalgia, SCL- 92=Symptom checklist-92, MDI=Major Depression Inventory, HDS=Hamilton Depression Scale, GAD-10=Anxiety Inventory, HAS=Hamilton Anxiety Scale, SF-36=health-related quality of life.

4.3. Depression 

The severity of depression measured by MDI showed a difference between the groups (Table S1). A rank sum test (median, percentiles) showed that patients with FM (median: 7, 4–9) had significantly higher scores than controls (median 2, 1–3) and NP patients (median 3, 3–5) . However, no difference between controls and NP patients was found . The gender-stratified analysis showed no significant differences.

Very few patients were depressed according to MDI: Only 3.3% of NP patients and 7.1% of the FM patients had the diagnostic criteria for moderate to severe depression according to ICD-10. None of the controls were depressed. If the men were excluded from the data analysis, 3.3% of NP patients and 3.5% of FM patients had moderate or severe depression according to ICD-10.

When the patients were evaluated by a medical doctor with the Hamilton Depression Scale, a significant difference was seen between all three groups both on the Kruskal-Wallis and rank sum tests (Fig. 2). There were no significant gender differences on neither the total nor the subtotal Hamilton depression score.

View Large Image Download to PowerPoint

Fig. 2. Hamilton Depression Scale rated by a doctor in patients with fibromyalgia and neuropathic pain compared with pain-free age- and gender-matched healthy controls. .

4.3.1. Anxiety 

When anxiety was measured on GAD-10, significant differences were found among all three groups both on the Kruskal–Wallis and rank sum tests (Table S1). Anxiety scores (median, percentiles) of FM patients (median: 14, 10–18) were indicative (severity) of doubtful anxiety (scores 8–14) or mild anxiety disorder (scores 15–19). The scores of NP patients (median 7, 4–11) and controls (median 2, 1–5.5) were indicative of doubtful or no anxiety disorder. This means that FM patients had the highest scores on GAD-10, although they did not have moderate (scores 20–29) or severe (scores 30–50) anxiety disorder scores. None of the conclusions were changed when the gender-stratified analysis was performed.

The medical doctor’s evaluation of the semi-structured interview Hamilton Anxiety Scale showed significant differences between all three groups both on the Kruskal–Wallis and rank sum tests (Fig. 3 and Table S1). Anxiety scores of FM patients (median 13, 10–16.5) were indicative (severity) of doubtful anxiety disorder (scores 8–14) or mild anxiety disorder (scores 15–19) and those of NP patients (median: 6, 2–10) were indicative of no or doubtful anxiety disorders, and those of controls (median: 1.5, 0–4) of no anxiety disorder. These conclusions were not changed by the gender-stratified analysis.

View Large Image Download to PowerPoint

Fig. 3. Hamilton Anxiety Scale rated by a doctor in patients with fibromyalgia and neuropathic pain compared with pain-free age- and gender-matched healthy controls. .

4.4. Health-related quality of life 

Chronic pain patients have significantly lower scores on all health-related quality of life subscales (SF-36) than healthy controls. Furthermore, FM patients as opposed to NP patients had significantly lower scores on role function–physical aspect (RP) , general health perception (GH) , vitality (VT) , social function (SF) , and mental health (MH) (Fig. 4 and Table S1). A gender-stratified analysis on all subgroups did not show any significant difference between the genders.

View Large Image Download to PowerPoint

Fig. 4. Health-related quality of life (SF-36) self-rated by patients with fibro-myalgia patients and neuropathic pain compared with pain-free age- and gender-matched healthy controls. *=Significant differences were found between patients with fibro-myalgia and neuropathic pain . Significant differences were found between patients (fibromyalgia and neuropathic pain) and controls on all subscale measures.

4.6. Pain vs. mental distress and health-related quality of life 

Correlation analyses within the patient groups between the SCL-92 or SF-36 and NRS were done. There was no correlation in the NP group between SCL-92 subscales and pain measures. In the FM group, there were significant and highly positive correlations between several SCL-92 subscales and average pain (Table 2). The correlations were found on somatization (0.8, ), anxiety (0.65, ), phobic anxiety (0.48, ), depression (0.58, ), psychoticism (0.48, ), and global severity index (0.7, ). Correlations were also observed in FM for SF-36 subscales (Table 2). The correlations coefficient and p-values were for function–physical aspect (−0.5, p<0.009), bodily pain (−0.4, p<0.04), general health perception (−0.58, p<0.02), and role function–emotional aspect (−0.42, p<0.04). For NP the only correlation in the SF-36 was seen in the bodily pain subscale. Furthermore, correlations were seen between NRS and HDS (0.52, ), MDI (0.59, ), and GAD (0.54, ) in the FM group, while no correlations were seen in the NP group.

Table 2. Associations between mental distress (SCL-92) or health-related quality of life (SF-36) and pain (NRS)

		NP patients	FM patients
	SCL-92
	Somatization	NS	0.8∗∗∗∗
	Anxiety	NS	0.65 ∗∗
	Interpersonal sensitivity	NS	NS
	Phobic anxiety	NS	0.48 ∗
	Obsession–compulsion	NS	NS
	Depression	NS	0.58 ∗∗
	Hostility	NS	NS
	Paranoid ideation	NS	NS
	Psychoticism	NS	0.48 ∗
	Global Severity Index	NS	0.70 ∗∗∗
	SF-36
	Physical functioning	NS	NS
	Role function–physical aspect	NS	−0.5 ∗∗∗
	Bodily pain	−0.55 ∗∗∗∗	−0.4 ∗
	General health perception	NS	−0.59 ∗∗
	Vitality	NS	−0.64 ∗∗∗∗
	Social function	NS	NS
	Role function – emotional aspect	NS	−0.42 ∗
	Mental health	NS	NS

NS=Not Significant. NP patients=Patients with neuropathic pain. FM patients=Patients with fibromyalgia.

∗ = . ∗∗ = . ∗∗∗ = . ∗∗∗∗ = .

5. Discussion 

To our knowledge, this is the first study to examine a series of mental symptoms in FM (Wolfe et al., 1990) and NP (Treede et al., 2007) compared with a healthy control group. The main finding is that both chronic pain groups had higher scores on mental distress and symptoms of depression and anxiety as well as a lower quality of life than an age- and sex-matched control group despite similar pain intensity in the patient groups. The prevalence of a current diagnosis of depression or anxiety disorder is, however, close to that seen in the normal Danish population suggesting that the mental symptoms relate to pain rather than to a mental disorder. Importantly, only in the FM group a significant correlation was seen between pain intensity and mental symptoms. This suggests that there are differences in the underlying pathophysiology of pain and mental symptoms in FM and NP, and how these phenomena are processed in the two conditions. The advantage of the present study is the multitude of patient- and doctor-rated symptoms and diagnostic scales administered in two different groups of patients with chronic pain and the close correspondence obtained between self-rated and doctor-rated scales.

Patients with FM had significantly higher scores than patients with NP on the SCL-92 subscales of depression, anxiety, somatization, and obsession–compulsion. Similarly, FM patients have higher scores on scales of depression and anxiety according to HDS, HAS, MDI, and GAD, but the prevalence of major depression was low. Previous studies have also found significant elevations in depression, anxiety, neuroticism, and hypochondriasis (self-rating) (Epstein et al., 1999) and in interviews measuring depression and anxiety for FM patients compared with controls (Krag et al., 1994, Epstein et al., 1999, Ahles et al., 1991). The current prevalence of depression in patients with FM is reported to be 22–26% and the lifetime prevalence of depression arguments to 70% according to DSM-III-R (Hudson et al., 1985, Epstein et al., 1999). A lower incidence was seen in a Danish study from 1989 (Loldrup et al., 1989), where 6% of chronic non-malignant pain patients were depressed when an observer scale was used. In contrast, the prevalence of depression was 30% if a checklist of depressive symptoms was used. Our study found a relatively small number of patients fulfilling the diagnostic criteria of depression according to ICD-10 (FM 7.1%, NP 3.3%) as measured on the MDI. These prevalence rates are close to the incidence of depression in a normal Danish population of 4.1% (females 4.9%, males 3.1%) (Olsen et al., 2004a). The large variation in the prevalence rates of depression pinpoints the necessity of distinguishing between: (1) severity scores scales, e.g. Hamilton depression and anxiety scales, and diagnosis of depression, e.g. DSM-III-R or ICD-10 (MDI) interviews; (2) observer, e.g. Hamilton depression scales, and self-rating scales, e.g. MDI and (3) ICD-10 and DSM criteria for depression. As shown the concepts are often confused in the pain literature, and this way give rise to artificially high depression scores. Furthermore, pain diagnoses and definitions are also important to distinguish between. The most accurate prevalence of depression is found when the diagnostic criteria of depression are evaluated in a stringently defined group of chronic pain patients as in our study.

Health-related quality of life was lower in chronic pain patients (NP and FM) than in healthy controls in accordance with previous studies (Otto et al., 2007, Gustorff et al., 2008, Pagano et al., 2004, Bergman et al., 2004, Haythornthwaite and Benrud-Larson, 2000). In addition, one study suggested that pain was the leading cause of low health-related quality of life (Svendsen et al., 2005). For example, multiple sclerosis (MS) patients with pain had lower scores than MS patients without pain. In our study, FM patients had significantly lower scores on a variety of parameters (MH, RP, VT, GH, SF) than NP patients despite similar pain scores. This may indicate that pain patients, especially FM patients may be difficult to treat because health-related quality of life is suggested to predict treatment success (Otto et al., 2007).

An important finding in this study is the correlation between several mental symptoms and pain intensity in the FM but not in the NP group. The mechanism responsible for this differential effect is not clear. One explanation could be different pain intensities in the two groups, but this was not the case in this study. Another possibility could be the moderate pain intensity (5–6 on NRS) in both groups. If the pain intensity had been higher, the results might have been different. Another possible explanation for the correlation between mental systems and pain intensity in the FM group is that FM patients have a central “top-down” disorder (McLean and Clauw, 2005, Schweinhardt et al., 2008), whereas NP have a “bottom-up” disturbance with neuronal abnormality driven from the periphery (Finnerup et al., 2007, Woolf, 2004, Jensen and Baron, 2003). In FM patients, a generalized hyperexcitability is present, possibly due to a disturbance in serotonin and noradrenaline transmission causing mental symptoms. These systems may also have an effect on the descending pain control from the brain stem to the spinal cord, which explains the association between mental symptoms and pain in FM. In contrast, NP patients are thought to have a localized pathology driven from the periphery, leading to central sensitization in the pain-modulating pathways in the spinal cord. This central sensitization in the spinal cord may in some patients subsequently lead to changes in the brain stem with a possible influence on brain structures involved in anxiety and mood. The failure to see an association between mental symptoms and pain in the NP group may then reflect upon a lack of recruitment of brain systems controlling mood phenomena.

Our study argues for the hypothesis that FM and NP share certain mechanisms leading to a central sensitization. While NP has a localized disorder driven from the periphery with few mental and somatic symptoms, FM has a generalized central driven sensitivity that may be responsible for the variety of somatic and mental symptoms seen in the study. Chronic stress due to environmental stress or genetic vulnerability may be an important factor in the development and maintenance of FM.

There are potential sources of error in this study. The distribution of gender and age was a result of the patient population that was willing to participate in the study. In this sample, NP patients are represented by a majority of middle-aged men, while fibromyalgia patients are mostly represented by females. The gender issue is an important one and was dealt with using a gender-stratified version of the rank sum test, which was done on all results. The analysis showed that only somatization was influenced by gender with women having more somatization than men. FM patients were recruited by advertising in newspapers, which may represent a potential bias. However, it is difficult to say whether it is the patients with the mildest or worst symptomatology who will respond to an advertisement. Previous studies (Smith et al., 2004a, Smith et al., 2004b, Edwards et al., 2006) have shown that suicidal ideation is an important parameter in chronic pain patients. Patients with suicidal ideation have more mental distress, depression, and anxiety than patients without this feature. This also applies to patients with alcohol and drug dependence. We cannot exclude that the present frequencies of depression and other mental symptoms are underestimated, but consider this possibility less likely since none of the patients in this study were excluded for these reasons. We believe that the present sample represents a chronic pain population as they are often seen in a neuropathic pain clinic. Another problem is related to the multiple testing carried out in the study. By correcting our results in a conservative manner using the Bonferroni test, the majority of p-values are still significant. Some of the patients received acetylsalicylic acid 75mg per day or paracetamol with a maximum of 4 g orally per day. Even though the drugs are known to influence pain, no differences in pain levels were seen at inclusion. It is therefore unlikely that this medication will interact with the results.

6. Conclusion 

In conclusion, we have shown that patients with FM have more mental symptoms such as depression and anxiety than patients with NP. Associations between pain intensity and mood phenomena were only found in FM patients. This finding points to different mechanisms responsible for the development of mood disorders in the two patient groups.