Headache comorbidity in refugees and migrants with post-traumatic stress disorder

Abstract

Background

Headache is often comorbid with post-traumatic stress disorder yet overlooked in health assessments of refugees.

Objectives

To describe prevalence of severe headache among refugees with post-traumatic stress disorder and compare severity of post-traumatic stress disorder symptoms and treatment outcomes among those with and without severe headache.

Methods

This follow-up study used data from the Danish Database on Refugees with Trauma. Participants were recruited from 2009 to 2015 at a specialized psychiatric clinic. Prevalence of severe headache was computed by age, sex, and history of head injury or torture. Severe headache was defined as maximum headache scores on the Hopkins Symptom Checklist, Symptom Checklist-90 or the Visual Analogue Scale. Groups with and without severe headache were described according to validated questionnaires before and after 12–18 months of multidisciplinary treatment for post-traumatic stress disorder. Regression analyses were used to analyze associations between headache at start of treatment and symptom burden post-treatment, controlled for pre-treatment scores and possible confounders.

Results

Among 403 female and 489 male participants, nearly all (97.5%) complained of headaches. Severe headache prevalence was 31.4% to 50.0% (depending on which questionnaire was used) and was significantly more common among females and those aged 30–49 years. There was no clear relationship between headache and head injury or torture. Participants with severe headache had heavier symptom burdens compared to those without severe headache. Post-treatment, headache prevalence by age and sex did not change significantly. Those without severe headache showed a tendency toward improvement in outcome measures; this was not seen among those with severe headache. Pre-treatment headache scores were correlated with all outcome measures. Regression analyses controlled for pre-treatment scores of the outcome variables showed associations between pre-treatment headache scores (Hopkins Symptom Checklist or Symptom Checklist-90) and post-treatment scores for intrusion, numbing, hyperarousal, anxiety, disability, and quality of life (all p < 0.02).

Conclusion

Headache is a prevalent comorbid condition among refugees with post-traumatic stress disorder. Measures of pre-treatment headache severity appear to predict post-traumatic stress disorder treatment outcomes. Severe headache adversely affects post-traumatic stress disorder prognosis. Assessment and treatment options should be studied further.

Keywords

PTSD anxiety depression disability quality of life

Introduction

The United Nations Refugee Agency estimates that there are 26.6 million refugees as of mid-2021 (1). The prevalence of post-traumatic stress disorder (PTSD) among refugees is difficult to study under conditions of ongoing conflict, unstable access to basic needs, and lack of health care services. Refugees could be about ten times more likely to have PTSD than matched general populations in western countries (2). Studies among refugees in various settings report a wide range of PTSD prevalence estimates: from 3% to 86% (2 –4).

Headache is a debilitating comorbid condition in many people with PTSD (5 –8). Headache disorders by themselves are associated with impaired quality of life (QoL), disability, substantial lost productivity, and high economic costs (9,10). Headache-related disability is greater in the presence of psychiatric comorbidity (11).

Unfortunately, headache is often overlooked in health assessments of refugees possibly because headache is viewed as a relatively minor, non-life-threatening problem. There are therefore very few prevalence studies on headache in refugee populations. A systematic literature search on comorbid chronic pain and PTSD among refugees yielded only 15 studies between 1996 and 2017. Only four estimated headache prevalence; with reports ranging from 50% to 99% (4).

This follow-up study aims to describe the prevalence of severe headache among refugees with PTSD; and to compare severity of PTSD symptoms and treatment outcomes among those with and without severe headache.

We hypothesized a high prevalence of headache in our clinic-based study population. We further hypothesized that those with severe headache would have poorer treatment outcomes compared to those without severe headache.

Methods

Recruitment

The primary data source was the Danish Database on Refugees with Trauma (DART) compiled through integrated clinical research practice at the Competence Centre for Transcultural Psychiatry (CTP), Mental Health Centre Ballerup, a specialized tertiary care clinic in the Capital Region of Denmark (12). Patients referred to CTP are considered particularly complex and require specialized care.

Participants from four recruitment periods were included (2009 to 2011, 2011 to 2012, 2011 to 2015, and 2013 to 2015) (13 –17). Inclusion criteria were: aged 18 years or older; referred to CTP for treatment of trauma-related symptoms; recognized as a refugee or family reunified with a refugee; had experienced psychological trauma related to being a refugee (e.g., torture, political persecution, imprisonment, experience of war); and diagnosed with PTSD according to ICD-10 research criteria.

Participants were excluded if they had a psychotic disorder (ICD-10 diagnoses F2x and F30.1–F31.9); abused drugs or alcohol (F1x.24–F1x.26); had been admitted to a psychiatric hospital; or had a physical illness or handicap that would not allow for participation (Figure 1).

Data collection

Data on headache were gathered uniformly throughout the recruitment periods. Interdisciplinary treatment was carried out using manuals developed specifically for this patient group. Typical treatment duration was 12 to 18 months (12 –17).

Baseline clinical and demographic data were gathered by medical doctors (including JC and MLW) and social workers. Information was gathered on age, sex, country of origin, civil status, children (under 18 years), education, work status, and nature of trauma.

Participants filled out self-administered questionnaires on their first and last visits. Questionnaires were available in Danish, English, Arabic, Persian, and Russian. Participants had the option of filling in questionnaires themselves or with the help of a translator.

The following data collection instruments were used:

Harvard Trauma Questionnaire (HTQ)

Sixteen questions on symptoms of PTSD and functional distress were used to quantify trauma-related symptoms (18,19). Responses were coded as (1) not at all, (2) a little, (3) quite a bit, and (4) extremely. Items were analyzed according to four symptom domains adapted from previous studies (20,21):

Intrusion (intrusive thoughts, flashbacks, disturbing dreams, physical or psychological reactions to reminders of trauma) items 1, 2, 3, 16.

Avoidance (avoiding activities, thoughts and feelings associated with the traumatic event) items 11 and 15.

Numbing (diminished interest, lack of vision for the future) items 4, 5, 12, 13, 14.

Hyperarousal (anxious arousal – feeling jumpy, feeling on guard) items 6 and 9; and (dysphoric arousal – difficulty concentrating and sleeping, irritability) items 7, 8 and 10.

A mean score (range 1 to 4) was calculated for each domain and for the total score.

Hamilton Depression and Anxiety Questionnaires

Severity of depression was analyzed using the Hamilton Depression scale (HAM-D) with 17 questions rated from 0 to 2 or 0 to 4 (22,23). Summary scores ranged from 0 to 52.

Severity of anxiety was analyzed using the Hamilton Anxiety scale (HAM-A) with 14 questions rated from 0 to 4 (24). Summary scores ranged from 0 to 56.

Sheehan Disability Scale (SDS) and the WHO Well-being Index (WHO-5)

Disability was measured using three items of the SDS indicating the extent to which symptoms disrupted life in three areas: work/school, social life, and family life. Each domain was scored from 0 to 10, with 0 being no disruption, and 10 being extreme disruption (25).

WHO-5 was used as a measure of QoL (26,27). Five questions were scored from 0 to 5 and summed for a total of 0 to 25. A score from 0 to 100 was calculated by multiplying the raw score by four. Higher numbers represented better QoL. A score below 13 was considered poor well-being.

Hopkins Symptom Checklist (HSCL), Symptom Checklist-90 (SCL-90), and Visual Analogue Scale (VAS)

Portions of the HSCL (28) and SCL-90 (29) were used to identify somatic comorbidities with a one-week recall period.

Participants were asked the following questions related to headache:

“How much did headache bother or distress you in the last week, including today?” Responses were coded as (1) not at all, (2) a little, (3) quite a bit, and (4) extremely.

“To which degree did headache bother you in the past week, including today?” Responses were coded as (0) not at all, (1) a little, (2) quite a bit, (3) a lot, and (4) extremely.

Participants were also asked to indicate degree of headache pain in the last four weeks using a VAS score from 0 (none) to 10 (very severe).

Physical injury

Data on head trauma were systematically gathered only from participants enrolled from 2013 to 2015. Participants were asked if they experienced any head injury related to their trauma (yes/no), whether head injuries resulted in loss of consciousness (yes/no), whether they experienced torture (yes/no), number of torture episodes (once, two to five times, more than five times), and duration of exposure to torture (ranging from less than a month to more than five years).

Definition of terms

A case of headache was defined as a report of headache in at least one of three questionnaires: HSCL (score of 2, 3, or 4), or SCL-90 (score of 1, 2, 3, or 4), or VAS (>=1).

Participants most affected by headache were those who reported a maximum HSCL score of 4, SCL-90 score of 4, or VAS 8 to 10. These participants are referred to in this study as suffering from “severe headache”.

Participants recruited into the study had symptoms of PTSD or were given a diagnosis of PTSD by a physician based on ICD-10 criteria. Criterion E (symptom debut within six months of the stressful event) was disregarded because of the complex and repeated nature of participants’ traumatic experiences.

Statistical analysis

Description of study participants

Demographic data were summarized as counts and percentages for the following variables: sex, age, region of origin, migrant status (refugee, family reunification or other), civil status, number of children, living arrangement, education in home country, current work in host country, income source, torture, and head injury related to a traumatic event. Chi-square tests were used to examine headache prevalence among those who experienced head injury or torture.

Headache prevalence and severity before and after PTSD treatment

Reports of headache in the HSCL and SCL-90 were summarized as counts and percentages. VAS scores were summarized as means and standard deviations. VAS scores were also dichotomized as <8 or 8–10, then counts and percentages were calculated.

Prevalence of severe headache was calculated by age, sex, history of physical trauma to the head, and history of torture. Chi-square test was used to examine distributions by sex. Fisher’s Exact test was used to compare distributions by age group. McNemar’s test was used to examine proportions with severe headache by sex and age group pre- and post-treatment.

Distribution according to headache severity pre- and post-treatment was analyzed using Wilcoxon Signed Ranks test.

PTSD treatment outcomes for participants with and without severe headache

Participants were divided into two groups: with and without severe headache according to their responses to the HSCL, SCL-90 and VAS. Symptom burden, disability, and QoL (as measured in four HTQ domains, HAM-A, HAM-D, SDS and WHO-5) were summarized as mean scores for each headache group pre- and post-treatment.

These scores were then compared using t-tests. P < 0.05 was considered significant at the outset. However, we considered analyses using the three headache measures (HSCL, SCL-90 and VAS) as a family of tests. We applied a more stringent cut-off point to avoid Type 1 error because of multiple comparisons. A p-value of <0.05/3, i.e., p < 0.017 was therefore considered significant.

Cohen’s d was used as a measure of effect size.

A correlation matrix was constructed to describe bivariate relationships between pre-treatment headache scores and outcome measures. Pearson’s correlation coefficients were computed. Significant correlations at the 0.05, 0.01 and 0.001 levels (2-tailed) were marked on the matrix.

Regression analyses were then performed to explore these relationships further.

In Model 1, independent variables were pre-treatment headache score and pre-treatment score of the outcome variable. This analysis was done for each questionnaire (i.e., scores for HSCL, SCL-90 and VAS). Model 2 included variables in Model 1 plus age and sex as possible confounding variables. Standardized beta coefficients and tests of significance for the headache variables were computed (level of significance, p < 0.017).

Dropout analysis

Dropouts were defined as those who did not have scores for all three questionnaires on headache symptoms (HSCL, SCL-90 and VAS) at end of study. Characteristics of dropouts were summarized by age, sex, several demographic variables, headache severity, PTSD severity (four HTQ domains), anxiety, depression, disability and QoL. Dropouts were then compared with study completers according to these characteristics using chi-square tests (p < 0.05).

SPSS Statistics 25 (IBM) was used for all statistical analyses.

Ethical considerations

Study protocols used during the four data collection periods were approved by relevant scientific ethics committees and registered as clinical trials (NCT00917397, NCT01362543, NCT01569685, NCT01955538). Informed consent was secured from all participants. Study aims and methods of data management were approved by the Data Protection Agency in accordance with the Privacy Act.

Data from the Danish Database on Refugees with Trauma (DART) were anonymized prior to analysis. MLW and JC had access to the database. Statistical analyses were done by MLW without access to participants’ Central Person Registry numbers or other personally identifiable information.

Results

Description of study participants

A total of 965 persons were enrolled. Among them, 73 were excluded and 892 were included (403 females, 489 males) (Figure 1). Demographic characteristics are summarized in Table 1.

Table 1.

Demographic data on 892 participants.

Demographic characteristics	All participants at start (n = 892)		All participants at end (n = 730)
Demographic characteristics	n	%	n	%
Sex
Male	489	54.8	386	52.9
Female	403	45.2	344	47.1
Age (years)
18–29	84	9.4	46	6.3
30–39	201	22.5	167	22.9
40–49	351	39.3	292	40.0
50–59	233	26.1	209	28.6
60–69	21	2.4	15	2.1
70–79	2	0.2	1	0.1
Region^a
Middle East	690	77.4	575	78.8
Eastern Europe	160	17.9	129	17.7
Africa	19	2.1	11	1.5
Asia	22	2.5	14	1.9
Missing data	1	0.1	1	0.1
Migrant status
Refugee	643	72.1	523	71.6
Family reunification	241	27.0	199	27.3
Other	2	0.2	2	0.3
Missing data	6	0.7	6	0.8
Civil status
Married or cohabiting	473	53.0	407	55.8
Single	132	14.8	91	12.5
Separated or divorced	228	25.6	199	27.3
Widow/widower	24	2.7	21	2.9
Missing data	35	3.9	12	1.6
Children (under 18 years)
None	301	33.7	237	32.5
1 child	159	17.8	141	19.3
2 children	187	21.0	158	21.6
3 or more children	206	23.1	178	24.4
Missing data	39	4.4	16	2.2
Living arrangement
Alone all the time	221	24.8	174	23.8
Alone part of the time	41	4.6	35	4.8
Living with family (never alone)	592	66.4	508	69.6
Missing data	38	4.3	13	1.8
Education in home country
Higher education	206	23.1	174	23.8
Upper secondary	205	23.0	172	23.6
Primary or lower secondary	322	36.1	270	37.0
Other	12	1.3	9	1.2
No education	98	11.0	73	10.0
Missing data	49	5.5	32	4.4
Current work in host country
None	458	51.3	374	51.2
Less than 10 hours/week	5	0.6	3	0.4
10 to 37 hours/week	18	2.0	14	1.9
Over 37 hours/week (full-time)	28	3.1	25	3.4
Other	244	27.4	207	28.4
Missing data	139	15.6	107	14.7
Income source
Salary	43	4.8	36	4.9
Sick leave benefits	173	19.4	146	20.0
Unemployment benefits	595	66.7	494	67.7
Pension	4	0.4	3	0.4
Education grant	13	1.5	6	0.8
Other	48	5.4	38	5.2
Missing data	16	1.8	7	1.0

^aRegion refers to the following responses to the question on country of origin: Middle East: Afghanistan, Iran, Iraq, Israel, Jordan, Kuwait, Lebanon, Palestine, Syria, Yemen; Eastern Europe: Albania, Armenia, Azerbaijan, Bosnia, Chechnya, Ex-Yugoslavia, Kazakhstan, Poland, Russia, Ukraine; Africa: Algeria, Burundi, Egypt, Ethiopia, Morocco, Rwanda, Somalia, Sudan, Tunisia, Uganda; Asia: Bhutan, Cambodia, China, India, Myanmar, Pakistan, Sri Lanka, Turkey, Vietnam.

At the start of treatment, 387 participants (43.9%) scored 4 on HSCL and 280 (31.4%) scored 4 on SCL-90 (Table 2). In addition, 430 (50.0%) indicated that their headache was between VAS 8 to 10, with a mean score of 7.0 (SD 2.9).

Table 2.

Participants who reported severe headache based on responses to three questionnaires, by age and sex, before and after treatment for PTSD.

	Start of treatment for PTSD				End of treatment for PTSD
	n	Frequency	Percent within category	p for category difference	n	Frequency	Percent within category	p for category difference	p for before and after treatment
HSCL = 4
Total	881	387	43.9		728	293	40.2		0.019
Sex
Male	484	167	34.5		385	115	29.9		0.047
Female	397	220	55.4	<0.001*	343	178	51.9	<0.001*	0.223
Age (years)
18–29	84	25	29.8		46	14	30.4		0.227
30–39	199	96	48.2		166	64	38.6		0.112
40–49	343	166	48.4		292	142	48.6		0.743
50–59	232	94	40.5		208	72	34.6		0.272
60–69	21	6	28.6		15	1	6.7		0.125
70–79	2	0	0	0.007*	1	0	0	<0.001*	—
SCL-90 = 4
Total	865	280	31.4		718	238	26.7		1.000
Sex
Male	475	111	23.4		382	96	25.1		0.320
Female	390	169	43.3	<0.001*	336	142	42.3	<0.001*	0.334
Age (years)
18–29	83	19	22.9		46	16	34.8		0.607
30–39	198	69	34.8		164	49	29.9		0.243
40–49	337	126	37.4		289	117	40.5		0.672
50–59	226	63	27.9		204	55	27.0		1.000
60–69	19	3	15.8		14	1	7.1		1.000
70–79	2	0	0	0.021	1	0	0	0.006*	—
VAS 8 to 10
Total	860	430	50.0		710	355	50.0		0.393
Sex
Male	473	191	40.4		377	158	41.9		1.000
Female	387	239	61.8	<0.001*	333	197	59.2	<0.001*	0.208
Age (years)
18–29	82	34	41.5		46	15	32.6		0.041
30–39	191	98	51.3		162	80	49.4		0.770
40–49	340	183	53.8		284	160	56.3		0.906
50–59	225	106	47.1		202	94	46.5		1.000
60–69	20	8	40.0		15	6	40.0		1.000
70–79	2	1	50.0	0.260	1	0	0	0.019	1.000

HSCL, Hopkins Symptom Checklist; PTSD, Post-traumatic stress disorder; SCL-90, Symptom Checklist 90; VAS, Visual Analogue Scale. Severe headache refers to maximum scores in the three questionnaires.

P for sex differences were analysed using chi-square test. P for age differences were analysed using Fisher’s Exact test since several cells had expected counts less than 5. P for before and after treatment was analyzed using Mc-Nemar’s test.

*The level of significance is corrected for multiple comparisons, 0.05 divided by 3 tests. P < 0.017 is considered significant.

Table 2 shows that severe headache is significantly more common among females than males (p < 0.001 for all three questionnaires), and most common among 30–49-year-olds compared to other age groups. Age group differences were not consistently statistically significant across the three questionnaires.

Among all 892 participants, 46.2% reported having been tortured. Among 318 participants who were asked systematically about head injury, 41.8% said they experienced head injury related to a traumatic event.

Table 3 summarizes the proportion of participants who reported severe headache according to history of head injury and torture. Prevalence of severe headache was similar for those with or without a history of head injury: (p > 0.017 for all three questionnaires). Prevalence was generally higher among those who experienced head injury with loss of consciousness compared to those who did not lose consciousness, but the differences did not reach statistical significance. Prevalence of severe headache was similar for those who reported torture and those who did not: (p > 0.017 for all three questionnaires). For those who experienced torture, there was no clear trend in headache complaints according to number or duration of torture episodes.

Table 3.

Proportions of participants with severe headache^a at start of treatment according to history of head injury and torture.

	n	HSCL = 4 (%)	n	SCL-90 = 4 (%)	n	VAS = 8 to 10 (%)
Head injury^b
Yes	133	51 (38.3%)	129	38 (29.5%)	128	67 (52.3%)
No	119	52 (43.7%)	120	36 (30.0%)	120	56 (46.7%)
Do not know	61	26 (42.6%)	57	19 (33.3%)	58	28 (48.3%)
p-value^c		0.669		0.863		0.660
Head injury with loss of consciousness
Yes	97	39 (40.2%)	94	28 (29.8%)	93	51 (54.8%)
No	21	4 (19.0%)	21	4 (19.0%)	21	6 (28.6%)
p-value^c		0.068		0.321		0.030
Experienced torture
Yes	409	180 (44.0%)	402	126 (31.3%)	401	198 (49.4%)
No	469	205 (43.7%)	460	152 (33.0%)	456	230 (50.4%)
p-value^c		0.929		0.594		0.756
Number of torture episodes
1 episode	181	81 (44.8%)	179	56 (31.3%)	177	93 (52.5%)
2 to 5 episodes	137	58 (42.3%)	135	38 (28.1%)	134	59 (44.0%)
>5 episodes	77	40 (51.9%)	75	23 (30.7%)	76	38 (50.0%)
p-value^c		0.390		0.829		0.326
Duration of torture
Less than a month	134	65 (48.5%)	131	40 (30.5%)	132	70 (53.0%)
One month to less than a year	152	68 (44.7%)	150	49 (32.7%)	149	70 (47.0%)
1 to 5 years	83	37 (44.6%)	82	26 (31.7%)	81	40 (49.4%)
More than 5 years	20	8 (40.0%)	20	5 (25.0%)	19	8 (42.1%)
p-value^c		0.853		0.910		0.694

^aSevere headache refers to a score of 4 on the Hopkins Symptom Checklist (HSCL), 4 on the Symptom Checklist-90 (SCL-90), or a score of 8 to 10 on the Visual Analogue Scale (VAS).

^bData on head injury were gathered systematically only during the last recruitment period (n = 318).

^cP values were determined using chi-square test. P < 0.017 is considered significant.

Headache prevalence and severity before and after PTSD treatment

At baseline, headache was reported by 97.5% (870 of 892 participants) in at least one of three symptom questionnaires (HSCL, SCL-90, or VAS). Post-treatment, prevalence was 97.4% (711 of 730 participants).

There was no significant change in prevalence of severe headache after treatment for all age and sex categories (all p > 0.017, Table 2).

Using HSCL, 172 had a better rating post-treatment, 146 had a worse rating, and 402 remained unchanged (p = 0.009). Using SCL-90, 194 had a better rating post-treatment, 188 had a worse rating, and 315 remained unchanged (p = 0.095). Pre-treatment VAS scores for all those with headache (i.e., excluding those who scored 0) were extremely high, with a mean of 7.1 (SD = 2.8). Post-treatment, the mean for all with headache decreased slightly to 6.9 (SD = 2.9); 305 had a better score post-treatment, 339 had a worse score, and 45 remained unchanged (p = 0.084).

PTSD treatment outcomes for participants with and without severe headache

Pre- and post-treatment, participants without severe headache had better scores compared to those with severe headache in all assessments of PTSD, anxiety, depression, disability, and QoL. Pre-treatment, differences between the two headache groups were statistically significant except for numbing (in all three questionnaires) and avoidance (HSCL and VAS) (Table 4a). Post-treatment, the gap in scores of the two groups appeared to widen and all comparisons showed statistically significant differences (Table 4b).

Table 4a.

Scores for severity of psychiatric symptoms, disability, and quality of life among participants with and without severe headache at start of treatment.

	Hopkins Symptom Checklist					Symptom Checklist-90					Visual Analogue Scale
		No severe headache	With severe headache				No severe headache	With severe headache				No severe headache	With severe headache
	n	Mean (SD)	Mean (SD)	p^a	Cohen’s d^b	n	Mean (SD)	Mean (SD)	p^a	Cohen’s d^b	n	Mean (SD)	Mean (SD)	p^a	Cohen’s d^b
HTQ	854	3.12 (0.42)	3.31 (0.39)	<0.001*	−0.468	838	3.14 (0.42)	3.34 (0.39)	<0.001*	−0.501	832	3.14 (0.41)	3.28 (0.40)	<0.001*	−0.351
Intrusion	878	3.23 (0.54)	3.51 (0.45)	<0.001*	−0.556	862	3.26 (0.54)	3.55 (0.44)	<0.001*	−0.589	857	3.25 (0.55)	3.47 (0.46)	<0.001*	−0.440
Avoidance	878	3.13 (0.73)	3.21 (0.78)	0.109	−0.110	862	3.12 (0.74)	3.26 (0.75)	0.012*	−0.183	857	3.12 (0.73)	3.23 (0.75)	0.021	−0.158
Numbing	879	2.90 (0.58)	3.00 (0.60)	0.020	−0.159	853	2.92 (0.58)	3.02 (0.61)	0.018*	−0.173	848	2.94 (0.58)	2.96 (0.61)	0.535	−0.043
Hyperarousal	877	3.25 (0.51)	3.50 (0.45)	<0.001*	−0.521	851	3.23 (0.50)	3.53 (0.44)	<0.001*	−0.527	845	3.26 (0.52)	3.47 (0.45)	<0.001*	−0.437
HAM-A	818	25.46 (7.61)	28.19 (7.27)	<0.001*	−0.367	804	25.47 (7.60)	29.11 (7.00)	<0.001*	−0.490	802	25.24 (7.48)	28.04 (7.44)	<0.001*	−0.376
HAM-D	821	22.47 (6.14)	24.69 (5.87)	<0.001*	−0.368	807	22.60 (6.26)	25.20 (5.40)	<0.001*	−0.433	805	22.43 (6.14)	24.51 (5.88)	<0.001*	−0.346
SDS work or school	826	7.94 (2.32)	8.52 (2.05)	<0.001*	−0.265	820	7.92 (2.33)	8.76 (1.88)	<0.001*	−0.382	816	7.80 (2.31)	8.63 (2.00)	<0.001*	−0.382
SDS social life	855	7.83 (2.32)	8.45 (2.09)	<0.001*	−0.279	849	7.84 (2.28)	8.62 (2.07)	<0.001*	−0.350	845	7.77 (2.36)	8.46 (2.04)	<0.001*	−0.310
SDS family or home	854	6.72 (2.81)	7.31 (2.64)	0.002*	−0.214	848	6.73 (2.73)	7.54 (2.69)	<0.001*	−0.299	844	6.63 (2.77)	7.37 (2.67)	<0.001*	−0.274
WHO-5	866	17.13 (15.98)	11.48 (13.60)	<0.001*	0.377	858	16.74 (15.62)	10.19 (13.33)	<0.001*	0.439	854	17.41 (16.39)	11.52 (13.16)	<0.001*	0.396

Table 4b.

Scores for severity of psychiatric symptoms, disability, and quality of life among participants with and without severe headache at end of treatment.

	Hopkins Symptom Checklist					Symptom Checklist-90					Visual Analogue Scale
		No severe headache	With severe headache				No severe headache	With severe headache				No severe headache	With severe headache
	n	Mean (SD)	Mean (SD)	p^a	Cohen’s d^b	n	Mean (SD)	Mean (SD)	p^a	Cohen’s d^b	n	Mean (SD)	Mean (SD)	p^a	Cohen’s d^b
HTQ	681	2.90 (0.62)	3.29 (0.40)	<0.001*	−0.701	674	2.94 (0.60)	3.30 (0.43)	<0.001*	−0.658	666	2.84 (0.62)	3.28 (0.43)	<0.001*	−0.782
Intrusion	722	3.01 (0.72)	3.47 (0.50)	<0.001*	−0.715	713	3.05 (0.70)	3.50 (0.52)	<0.001*	−0.685	706	2.95 (0.72)	3.45 (0.52)	<0.001*	−0.440
Avoidance	717	2.88 (0.82)	3.08 (0.79)	<0.001*	−0.250	708	2.90 (0.79)	3.08 (0.85)	0.006*	−0.220	701	2.85 (0.81)	3.07 (0.81)	<0.001*	−0.265
Numbing	703	2.73 (0.71)	3.06 (0.57)	<0.001*	−0.503	695	2.76 (0.70)	3.09 (0.56)	<0.001*	−0.173	687	2.66 (0.70)	3.07 (0.59)	<0.001*	−0.643
Hyperarousal	704	3.01 (0.71)	3.46 (0.45)	<0.001*	−0.733	695	3.06 (0.68)	3.45 (0.48)	<0.001*	−0.636	688	2.94 (0.70)	3.44 (0.49)	<0.001*	−0.845
HAM-A	513	25.37 (9.51)	28.82 (8.09)	<0.001*	−0.384	506	25.60 (9.48)	28.89 (7.79)	<0.001*	−0.367	503	24.27 (9.60)	29.14 (7.99)	<0.001*	−0.549
HAM-D	451	21.62 (7.57)	24.14 (6.76)	<0.001*	−0.347	446	21.80 (7.51)	24.20 (6.72)	0.001*	−0.331	444	20.70 (7.63)	24.37 (6.63)	<0.001*	−0.512
SDS work or school	687	7.23 (2.89)	8.82 (1.75)	<0.001*	−0.637	682	7.37 (2.79)	8.87 (1.91)	<0.001*	−0.591	678	6.99 (2.91)	8.79 (1.89)	<0.001*	−0.735
SDS social life	702	7.35 (2.85)	8.80 (1.68)	<0.001*	−0.595	696	7.46 (2.76)	8.87 (1.75)	<0.001*	−0.568	694	7.09 (2.94)	8.79 (1.71)	<0.001*	−0.705
SDS family or home	698	6.58 (2.99)	7.77 (2.36)	<0.001*	−0.434	692	6.62 (2.93)	7.89 (2.38)	<0.001*	−0.460	688	6.25 (2.99)	7.86 (2.39)	<0.001*	−0.591
WHO-5	715	25.42 (24.56)	11.24 (13.77)	<0.001*	0.679	710	24.49 (24.07)	10.26 (12.85)	<0.001*	0.676	703	28.15 (24.66)	11.57 (15.14)	<0.001*	0.810

Severe headache refers to maximum scores in the three questionnaires (Hopkins Symptom Checklist = 4, or Symptom Checklist-90 = 4, or Visual Analogue Scale = 8 to 10). For WHO-5, a higher value shows better well-being. For all other variables, a higher value shows greater symptom burden.

^aIndependent samples t-tests were used to compare means (two-sided p values after testing for equality of variances).

^bCohen’s d is used as a measure of effect size. Negative values are retained to show that means for the no severe headache group are less than means for the severe headache group in all measured variables except for WHO-5.

HAM-A, Hamilton Anxiety Assessment Scale; HAM-D, Hamilton Depression Assessment Scale; SDS, Sheehan’s Disability Scale; WHO-5, World Health Organization Well-being Index; SD, standard deviation of the mean.

*P < 0.017 is considered significant.

Intrusion and hyperarousal showed strong effect sizes pre-treatment (Cohen’s d > 0.500, Table 4a). Post-treatment, in addition to intrusion and hyperarousal, QoL and several disability scores also showed high effect sizes (Cohen’s d > 0.500, Table 4b). The largest differences between group means were seen in hyperarousal and WHO-5 scores post-treatment (Cohen’s d > 0.800 for VAS scores, Table 4b).

Bivariate correlations were computed using pre-treatment headache scores and outcome measures (Table 5). The three questionnaires were strongly correlated with each other (correlation coefficients >0.600). All headache scores were significantly correlated with outcomes, except in the case of HSCL and avoidance. The coefficients were positive, i.e., higher pain scores correlated with higher scores on the outcome measures, indicating worse outcomes. The exceptions were correlations with WHO-5, which showed negative coefficients since high pain scores correlated with lower QoL. The strengths of the relationships were low (correlation coefficients <0.300), but the analyses likely had enough statistical power to identify even these weak effects.

Table 5.

Correlation matrix showing bivariate relationships between headache scores at start of treatment and outcome measures at end of treatment.

	HSCL	SCL-90	VAS
HSCL	–
SCL-90	0.694***	–
VAS	0.635***	0.747***	–
HTQ	0.220***	0.237***	0.163***
Intrusion	0.224***	0.249***	0.180***
Avoidance	0.033	0.093*	0.098**
Numbing	0.144***	0.136***	0.084*
Hyperarousal	0.243***	0.231***	0.160***
HAM-A	0.204***	0.185***	0.101*
HAM-D	0.175***	0.175***	0.133**
SDS work or school	0.214***	0.198***	0.128***
SDS social life	0.179***	0.190***	0.123**
SDS family or home	0.161***	0.146***	0.114**
WHO-5	−0.180***	−0.165***	−0.134***

All values are Pearson correlation coefficients. The strongest correlations (>0.500) are highlighted.

HAM-A, Hamilton Anxiety Assessment Scale; HAM-D, Hamilton Depression Assessment Scale; HTQ, Harvard Trauma Questionnaire; HSCL, Hopkins Symptom Checklist; SCL-90, Symptom Checklist-90; SDS, Sheehan’s Disability Scale; VAS, Visual analogue scale; WHO-5, World Health Organization Well-being Index.

*Correlation is significant at the 0.05 level (2-tailed). **Correlation is significant at the 0.01 level (2-tailed). *** Correlation is significant at the 0.001 level (2-tailed).

Regression analyses were performed using pre-treatment headache score as the independent variable and the various outcome measures as dependent variables (Table 6).

Table 6.

Regression analyses showing relationships between headache (at start of treatment) and treatment outcome.

	Headache score
	HSCL		SCL-90		VAS
Treatment outcome	Coeff B	p	Coeff B	p	Coeff B	p
MODEL 1: headache score is the independent variable adjusted for baseline score of the treatment outcome
HTQ	0.120	<0.001*	0.114	0.002*	0.083	0.021
Intrusion	0.124	<0.001*	0.129	<0.001*	0.093	0.008*
Avoidance	0.016	0.645	0.069	0.056	0.073	0.042
Numbing	0.113	0.001*	0.090	0.010*	0.072	0.038
Hyperarousal	0.151	<0.001*	0.122	0.001*	0.075	0.043
HAM-A	0.121	0.003*	0.080	0.056	0.020	0.630
HAM-D	0.085	0.045	0.079	0.068	0.049	0.252
SDS work or school	0.154	<0.001*	0.133	<0.001*	0.065	0.089
SDS social life	0.137	<0.001*	0.143	<0.001*	0.078	0.037
SDS family or home	0.108	0.003*	0.088	0.016*	0.062	0.090
WHO-5 index	−0.117	<0.001*	−0.092	0.010*	−0.067	0.063
MODEL 2: headache score is the independent variable adjusted for baseline score of the treatment outcome, age, and sex
HTQ	0.136	<0.001*	0.138	<0.001*	0.099	0.007*
Intrusion	0.138	<0.001*	0.149	<0.001*	0.106	0.003*
Avoidance	0.024	0.517	0.081	0.029	0.084	0.023
Numbing	0.131	<0.001*	0.116	0.001*	0.091	0.010*
Hyperarousal	0.161	<0.001*	0.137	<0.001*	0.084	0.025
HAM-A	0.119	0.004*	0.076	0.076	0.013	0.749
HAM-D	0.085	0.049	0.077	0.084	0.043	0.321
SDS work or school	0.166	<0.001*	0.146	<0.001*	0.071	0.065
SDS social life	0.144	<0.001*	0.152	<0.001*	0.080	0.036
SDS family or home	0.130	<0.001*	0.112	0.003*	0.078	0.034
WHO-5 index	−0.130	<0.001*	−0.108	0.003*	−0.077	0.037

Coeff B, Standardized beta coefficients for headache as an independent variable; HAM-A, Hamilton Anxiety Assessment Scale; HAM-D, Hamilton Depression Assessment Scale; HSCL, Hopkins Symptom Checklist; HTQ, Harvard Trauma Questionnaire; SCL-90, Symptom Checklist-90; SDS, Sheehan’s Disability Scale; VAS, Visual analogue scale; WHO-5, World Health Organization Well-being Index.

*The level of significance is corrected for multiple comparisons, 0.05 divided by 3 tests. P < 0.017 is considered significant.

Model 1 was adjusted for the baseline score of the outcome variable and showed significant relationships between pre-treatment headache score (HSCL and SCL-90) and post-treatment intrusion, numbing, hyperarousal, anxiety, disability, and QoL. Similar results were seen in Model 2, which was additionally adjusted for age and sex. The relationships between headache scores and HTQ, intrusion and numbing were significant in all three questionnaires after these adjustments.

Results for HSCL and SCL-90 analyses were similar. In analyses using VAS scores, only HTQ, intrusion and numbing, were significant (p < 0.017), although several analyses for disability and QoL had p-values <0.05 (Table 6). Overall, results appeared to indicate that severe headache complaints at start of treatment predicted poor outcomes.

Dropout analysis

There were 162 participants lost to follow up (18.2%) while 730 (344 females and 386 males) completed treatment. Dropouts were disproportionately male (21.1% vs. 14.6% among females, p = 0.013), under 30 years old (45.2% dropout) or over 60 years old (30.4% dropout, p < 0.001). There were also significant differences in schooling and work profiles (p < 0.042); with a higher proportion of dropouts enrolled in education programs (4.6%) compared to completers (0.8%).

The proportion of those with severe headache was similar for dropouts and completers in the three questionnaires (p = 0.056 for HSCL, p = 0.299 for SCL-90, and p = 0.060 for VAS).

Regarding PTSD symptoms, those who dropped out had less severe avoidance scores at baseline (mean 3.04) compared to completers (3.19; p = 0.024). Differences in intrusion scores were marginally significant (p = 0.041). There were no significant differences according to severity of numbing (p = 0.157) and hyperarousal (p = 0.246).

Disability scores were similar for dropouts and completers (p = 0.151 for work/school, p = 0.288 for social life, and p = 0.102 for family life). However, dropouts had significantly better scores on HAM-A (p = 0.010), HAM-D (0.002), and WHO-5 (p = 0.023).

Discussion

Severe headache was highly prevalent in this study population of refugees with PTSD. Severe headache was more common among women and the middle aged. There was no clear relationship with head trauma and torture episodes. The group with severe headache had significantly heavier symptom loads for PTSD, anxiety, and depression; had higher disability scores; and poorer QoL.

Prevalence of severe headache (by age and sex) did not significantly change after treatment for PTSD. Regression analyses using HSCL or SCL-90 (but not VAS) as measures of headache severity showed that pre-treatment headache appears to predict symptoms of intrusion, numbing, hyperarousal, anxiety, disability, and QoL post-treatment.

Severe headache comorbid with PTSD

In the general population, one-year prevalence of any headache is between 24 to 79% (10). The prevalence of tension-type headache (TTH) is 26.1%, and migraine 14.4% (9). Yet in this study population of refugees with PTSD, point prevalence (one- or four-week recall) of any headache was 97.4%, and depending on the questionnaire used, 31.4 to 50.0% reported severe debilitating headaches.

Headache prevalence differs between local and immigrant populations. Onset or worsening of headache often occurs after immigration, which likely reflects the stress of displacement: uncertain access to basic needs, sleep disturbances, comorbid psychiatric illness, lower access to medical care (due to language barriers or socio-cultural factors), and differences in medication use (30). Neglecting headache diagnosis and treatment among migrants and refugees likely contributes to increased burden of headache worldwide.

Estimating prevalence of comorbid PTSD and pain among refugees is difficult because of the fluidity of the population, multiple traumatic experiences, and limited access to regular health care where psychological and physical health can be assessed. A review of 15 studies on PTSD and pain among refugees shows heterogenous research settings, cultural groups, and phases of integration into host countries (4). Published papers are mainly on people with a refugee background resettled in Western countries (31), where 85.7% of those with PTSD report head pain, compared to only 14.3% among those not diagnosed with PTSD.

General health assessments in primary care settings report lower headache prevalence compared to clinic-based studies. A study of newly arrived Syrian refugees in Norway surveyed 827 adults for physical and mental health problems. Headache was the most prevalent illness, reported by 22% (32). Among refugees treated at a psychiatric center in Norway, however, 98% reported chronic pain in at least one location, with the most prevalent location being the head (80%) (31).

Similarly, in Denmark, a study of 160 newly resettled refugees showed that 54% reported headache during their general health assessment (33). Among refugees with PTSD treated at our psychiatric center, prevalence of headache is much higher: 97.4%. Analysis of a subset of participants included in this study (168 women and 150 men), showed that high pain scores along with chronicity of PTSD symptoms predicted poor treatment response (34).

Our findings support research showing associations between chronic pain and PTSD (6,33,35); as well as high prevalence of PTSD among persons with headache in general (5) and migraine in particular (7,8).

There is high comorbidity between traumatic brain injury, headache, and PTSD (36 –38). In our study, severe headache was more common among those who experienced head injury with loss of consciousness. A study of asylum seekers in the US showed that almost half (42.5%) reported a history of head injury, which is similar to our finding of 40.2% for HSCL (39).

Our findings on poor outcome of treatment for those with severe headache, align with observations from a recent study of refugees who experienced torture or trauma (40). Improvement in PTSD symptoms was observed after treatment, but not pain severity or pain interference (i.e., how pain affects daily functioning). As expected, higher levels of pain interference predicted poorer outcomes.

The current study shows that severe headache could be an effect modifier, where the effect of treatment for PTSD on an outcome variable (improvement in PTSD symptoms) differs depending on the presence or absence of a third variable (severe headache or more generally, pain).

Severe headache could be a marker for PTSD that is refractory to usual treatment. Severe headache in a person with PTSD could be associated with higher pain intensity in other parts of the body, pain catastrophizing, depression, anxiety, disability, higher pain medication use, and lower self-efficacy, all these leading to low QoL (35,40,41).

There are several explanatory theories on the link between chronic pain and PTSD (42), all pointing to shared pathophysiological mechanisms of these two conditions, and maladaptive changes or abnormalities in the somatosensory and cerebral systems (36,43,44).

Pain may trigger memories of trauma. Sensory reexperiencing symptoms of PTSD could include pain (36). Stressful life events (e.g., job loss, divorce, major illness) not necessarily related to previous trauma could lead to pain and PTSD-like symptoms in someone with an existing vulnerability (45).

Implications for diagnosis and treatment

Assessment of pain in patients diagnosed with PTSD is important in treatment and prognostication. Pain self-report questionnaires (e.g., Brief Pain Inventory, the Pain Catastrophizing Scale, or the Short Form McGill Pain Questionnaire) could be used to gather more information (35,45). Characterizing headache pain as being caused by tension, migraine, medication overuse, or physical trauma, could be the starting point for education and appropriate treatment.

The group without severe headache – representing the majority of participants who completed treatment – showed improvement. This is a positive finding for this group of refugees with PTSD, whose symptoms required specialized treatment.

Among people with chronic pain, avoidance contributes to maintenance of PTSD; and among those with PTSD, avoidance contributes to maintenance or aggravation of chronic pain (31). Reduction in avoidance behavior after treatment is maybe the first sign of improvement. Available data do not allow for much optimism, however. A 10-year study of 139 tortured refugees resettled in Denmark reported a 47.5% prevalence of head pain at baseline, and 58.3% at follow-up (46), representing a clear challenge to treatment programs.

There are very few studies on effective treatments for comorbid chronic pain and PTSD. Cognitive behavioral therapy (CBT), Acceptance and Commitment Therapy (ACT), Narrative Exposure Therapy (NET) with biofeedback, Cognitive Processing Therapy (CPT), prolonged exposure, manualized trauma psychotherapy, Traditional Chinese Medicine (TCM) and Emotional Freedom Techniques have been evaluated with positive outcomes for both pain severity and PTSD symptoms (4,35), although these methods do not focus specifically on headache pain.

Since research points to a common underlying mechanism that explains comorbid pain and PTSD, an integrated treatment approach for both conditions might be more effective (36,47). Headache treatment will likely support improvement of PTSD symptoms. Non-pharmacological treatment of TTH – such as stress management and relaxation techniques – can also alleviate symptoms of PTSD. Preventive medications for TTH recommended by the Danish Headache Society include amitriptyline, mirtazapine, and venlafaxine (48), which are, perhaps not coincidentally, among the medications used for treatment of PTSD symptoms at our psychiatric center.

Improved treatment of headache comorbidity could improve PTSD outcomes. Treatment of migraine reduces symptoms of depression and anxiety and improves associated sleep problems and fatigue (49). Treatment of debilitating medication overuse headache (MOH) reduces anxiety, depression, and disability (50). Treatment of chronic post-traumatic headache shows some degree of psychological improvement and slightly better QoL compared to a control group, even if there is no significant improvement in headache and pain parameters (51).

Strengths and limitations

This study is the largest so far conducted on headache in a clinic-based population of migrants and refugees diagnosed with PTSD. Data were gathered consistently across several recruitment periods. Study participants were previously examined by health professionals in primary or secondary care before being referred to specialist care at a tertiary center with a working diagnosis of PTSD. Symptoms of PTSD were then assessed uniformly by medical doctors.

The questionnaires on PTSD, anxiety, depression, pain, and QoL are all previously validated and widely used (18,19,22 –29) These have also been validated for refugee populations (52).

A limitation is that headache was not characterized according to location, duration, precipitating/relieving factors, and associated symptoms. Specific headache diagnoses could not be made, although these were most likely TTH, migraine, post-traumatic headache, or MOH (53). The case definition for severe headache does not conform to current diagnostic criteria (53).

Bivariate analyses show that HSCL, SCL-90 and VAS are strongly correlated. However, during preliminary analyses, we found differences in the numbers of participants who reported high scores for headache in these three questionnaires. This made it difficult to choose just one instrument to describe the group most affected by headache, assuming the three instruments are equally valid. Ideally, a validated headache diagnostic questionnaire should be used, and the participants asked to fill in a headache diary.

A sum score for all three questionnaires was considered, but without validation, cut-off scores for what constituted severe headache would have been arbitrary. Analyzing all three questionnaires increased the number of statistical analyses performed, which increased the risk of Type 1 error. We attempted to decrease the error risk by using a lower p-value to indicate statistical significance.

Regression analyses using VAS scores showed fewer significant associations, possibly because of a ceiling effect, where a high percentage of participants scored near the maximum and therefore linear models are not appropriate. Severity of headache was based on one-week recall for HSCL and SCL-90; and four weeks for VAS. This might also account for the differences in the results.

Information on prescription pain medications was gathered systematically only for the last 318 participants, and only at baseline. The most common pain medications prescribed were paracetamol (111 participants), tramadol (39 participants) and triptans (20 participants). They were not systematically asked about use of over-the-counter pain relievers nor misuse of opiates, the latter being more prevalent among patients with PTSD (35,54). In a recent Norwegian study of newly arrived refugees, 21% of those who reported headaches used painkillers every day (32). It was not possible to diagnose MOH which was likely prevalent in this study population.

Participants represent the most severe cases of PTSD requiring referral to a highly specialized center. Their responses to the questionnaires were mostly in the severe end of the scales. Our findings are not generalizable to all groups of migrants and refugees.

Data were anonymized after inclusion. It was not possible to rule out repeated inclusion of the same patient during different recruitment periods. If this occurred, this would have accounted for a very small number of participants. Demand for services is high, and those who had already been through a full treatment program were not prioritized for repeat inclusion.

Finally, a sensitivity analysis of the effect of dropout was not done. We suspect that those who dropped out chose to prioritize work or school and were probably more able to function in these settings because of lower avoidance (and possibly intrusion) scores. This group had a lighter symptom burden for anxiety and depression. They reported better well-being even though their disability scores were similar to those who completed treatment.

The proportion of those with severe headache was similar for dropouts and completers in the three questionnaires. It is therefore less likely that results were biased by missing headache data. However, it is possible that we underestimated the degree of improvement for those without severe headache.

Several questionnaires were not completely filled in by participants, and some did not attend a separate session for HAM-A and HAM-D interviews conducted by blinded interviewers. This aspect of data collection could be improved. We chose to present the results as recorded, rather than impute missing data.

Conclusion

Headache is a highly prevalent comorbid condition in this group of migrants and refugees referred to a specialized center for treatment of symptoms of PTSD. Those with severe headache presented with higher symptom burden, more disability, and lower QoL; and showed little improvement post-treatment likely because of mutual maintenance of PTSD and headache disorders. Measures of pre-treatment headache severity appear to predict PTSD treatment outcomes. Assessment and treatment options for these debilitating comorbid conditions should be studied further.

Clinical implications

This study on refugees with PTSD shows that assessment of headache is important in treatment and prognostication.

Severe headache adversely affects PTSD prognosis. Improved treatment of headache comorbidity might improve PTSD outcomes.

Since research points to a common underlying mechanism that explains comorbid pain and PTSD, an integrated treatment approach for both conditions might be more effective.

Footnotes

Acknowledgements

The authors thank the clinical and research staff and secretaries at the Competence Centre for Transcultural Psychiatry, Erik Lykke Mortensen for invaluable advice on statistical analysis, and most of all, our patients through the years who contributed their time and engagement, knowing that their efforts would help improve the treatment of psychiatric illness among migrants and refugees.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Data collection for the Danish Database on Refugees with Trauma (DART) has been ongoing since 2009. Funds were received from TrygFonden under grant numbers 7-10-1002 and 102265 URL: . The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

ORCID iDs

Maria Lurenda Westergaard

Jessica Carlsson

References

United Nations High Commissioner for Refugees (UNHCR). Mid-year trends 2021. https://www.unhcr.org/statistics/unhcrstats/618ae4694/mid-year-trends-2021.html UNHCR. (Accessed 11 January 2022).

Fazel

Wheeler

Danesh

Prevalence of serious mental disorder in 7000 refugees resettled in western countries: a systematic review. Lancet 2005; 365: 1309–1314.

Slewa-Younan

Uribe Guajardo

Heriseanu

, et al. A systematic review of post-traumatic stress disorder and depression amongst Iraqi refugees located in western countries. J Immigr Minor Health 2015; 17: 1231–1239.

Rometsch-Ogioun El Sount

Windthorst

Denkinger

, et al. Chronic pain in refugees with posttraumatic stress disorder (PTSD): A systematic review on patients’ characteristics and specific interventions. J Psychosom Res 2019; 118: 83–97.

Siqveland

Hussain

Lindstrøm

, et al. Prevalence of posttraumatic stress disorder in persons with chronic pain: A meta-analysis. Front Psychiatry 2017; 8: 164.

Fishbain

Pulikal

Lewis

, et al. Chronic pain types differ in their reported prevalence of post-traumatic stress disorder (PTSD) and there is consistent evidence that chronic pain is associated with PTSD: An evidence-based structured systematic review. Pain Med 2016; 18: 711–735.

Peterlin

Tietjen

Brandes

, et al. Posttraumatic stress disorder in migraine. Headache 2009; 49: 541–551.

Peterlin

Nijjar

Tietjen

GE.

Post-traumatic stress disorder and migraine: epidemiology, sex differences, and potential mechanisms. Headache 2011; 51(6): 860–868.

GBD 2016 Headache Collaborators. Global, regional, and national burden of migraine and tension-type headache, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol 2018; 17: 954–976.

10.

Saylor

Steiner

TJ.

The Global Burden of Headache. Semin Neurol 2018; 38: 182–190.

11.

Lipton

Seng

Chu

, et al. The effect of psychiatric comorbidities on headache-related disability in migraine: Results from the Chronic Migraine Epidemiology and Outcomes (CaMEO) Study. Headache 2020; 60: 1683–1696.

12.

Carlsson

Sonne

Silove

From pioneers to scientists: challenges in establishing evidence-gathering models in torture and trauma mental health services for refugees. J Nerv Ment Dis 2014; 202: 630–637.

13.

Buhmann

Nordentoft

Ekstroem

, et al. The effect of flexible cognitive behavioural therapy and medical treatment, including antidepressants on post-traumatic stress disorder and depression in randomized refugees: pragmatic randomized controlled clinical trial. Br J Psychiatry 2016; 208: 252–259.

14.

Sonne

Carlsson

Bech

, et al. Treatment of trauma-affected refugees with venlafaxine versus sertraline combined with psychotherapy – a randomized study. BMC Psychiatry 2016; 16: 383.

15.

Sonne

Carlsson

Bech

, et al. Psychosocial predictors of treatment outcome for trauma-affected refugees. Eur J Psychotraumatol 2016; 7: 30907.

16.

Carlsson

Sonne

Vindbjerg

, et al. Stress management versus cognitive restructuring in trauma-affected refugees – A pragmatic randomized study. Psychiatry Res 2018; 266: 116–123.

17.

Nordbrandt

Sonne

Mortensen

, et al. Trauma-affected refugees treated with basic body awareness therapy or mixed physical activity as augmentation to treatment as usual – A pragmatic randomized controlled trial. PloS One 2020; 15: e0230300.

18.

Berthold

Mollica

Silove

, et al. The HTQ-5: revision of the Harvard Trauma Questionnaire for measuring torture, trauma and DSM-5 PTSD symptoms in refugee populations. Eur J Public Health 2019; 29: 468–474.

19.

Mollica

Caspi-Yavin

Bollini

, et al. The Harvard Trauma Questionnaire. Validating a cross-cultural instrument for measuring torture, trauma, and posttraumatic stress disorder in Indochinese refugees. J Nerv Ment Dis 1992; 180: 111–6.

20.

Tay

Mohsin

Rees

, et al. The factor structures and correlates of PTSD in post-conflict Timor-Leste: an analysis of the Harvard Trauma Questionnaire. BMC Psychiatry 2017; 17: 191.

21.

Rasmussen

Smith

Keller

AS.

Factor structure of PTSD symptoms among West and Central African refugees. J Trauma Stress 2007; 20: 271–80.

22.

Hamilton

A rating scale for depression. J Neurol Neurosurg Psychiat 1960; 23: 56–62.

23.

Bech

Allerup

Gram

, et al. The Hamilton Depression Scale: Evaluation of objectivity using logistic models. Acta Psych Scand 1981; 63: 290–99.

24.

Hamilton

The assessment of anxiety states by rating. Br J Med Psychol 1959; 32: 50–55.

25.

Sheehan

DV.

The Anxiety Disease. New York: Charles Scripner and Sons, 1983.

26.

World Health Organisation. Wellbeing Measures in Primary Health Care/The Depcare Project. Copenhagen: WHO Regional Office for Europe, 1988.

27.

Topp

Østergaard

Søndergaard

, et al. The WHO-5 well-being index: A systematic review of literature. Psychother Psychosomat 2015; 84: 167–176.

28.

Parloff

Kelman

Frank

JD.

Comfort, effectiveness, and self-awareness as criteria for improvement in psychotherapy. Am J Psych 1954; 3: 343–351.

29.

Derogatis

Lipman

Covi

SCL-90: An outpatient psychiatric rating scale-preliminary report. Psychopharmacol Bull 1973; 9: 13–28.

30.

Kruja

Headache in migrants and refugees. In: El Alaoui-Faris

Federico

Grisold

(eds) Neurology in Migrants and Refugees. Sustainable Development Goals Series. Switzerland: Springer Nature, 2022, pp.229–238.

31.

Teodorescu

Heir

Siqveland

, et al. Chronic pain in multi-traumatized outpatients with a refugee background resettled in Norway: a cross-sectional study. BMC Psychol 2015; 3: 7.

32.

Strømme

Haj-Younes

Hasha

, et al. Health status and use of medication and their association with migration related exposures among Syrian refugees in Lebanon and Norway: a cross-sectional study. BMC Public Health 2020; 20: 341.

33.

Andersen

Kruse

Frederiksen

, et al. Health status of refugees newly resettled in Denmark. Dan Med J 2020; 67: A08200567.

34.

Nordbrandt

Vindbjerg

Mortensen

, et al. Chronicity of posttraumatic stress disorder and comorbid pain as predictors of treatment response for trauma-affected refugees. J Trauma Stress 2022; 1–12.

35.

Kind

Otis

JD.

The interaction between chronic pain and PTSD. Curr Pain Headache Rep 2019; 23: 91.

36.

Scioli-Salter

Forman

Otis

, et al. The shared neuroanatomy and neurobiology of comorbid chronic pain and PTSD: therapeutic implications. Clin J Pain 2015; 31: 363–374.

37.

Iljazi

Ashina

Al-Khazali

, et al. Post-traumatic stress disorder after traumatic brain injury –A systematic review and meta-analysis. Neurol Sci 2020; 41: 2737–2746.

38.

Lambru

Benemei

Andreou

, et al. Position paper on post-traumatic headache: The relationship between head trauma, stress disorder, and migraine. Pain Ther 2021; 10: 1–13.

39.

McMurry

Tsang

Lin

, et al. Head injury and neuropsychiatric sequelae in asylum seekers. Neurology 2020; 95: e2605–e2609.

40.

Nordin

Perrin

Pre-treatment pain predicts outcomes in multimodal treatment for tortured and traumatized refugees: a pilot investigation. Eur J Psychotraumatol 2019; 10: 1686807.

41.

Benedict

Keenan

Nitz

, et al. Post-traumatic stress disorder symptoms contribute to worse pain and health outcomes in veterans with PTSD compared to those without: A systematic review with meta-analysis. Mil Med 2020; 185: e1481–e1491.

42.

Brennstuhl

Tarquinio

Montel

Chronic pain and PTSD: Evolving views on their comorbidity. Perspect Psychiatr Care 2015; 51: 295–304.

43.

Moeller-Bertram

Keltner

and Strigo

. Pain and post-traumatic stress disorder – review of clinical and experimental evidence. Neuropharmacology 2012; 62: 586–97.

44.

McKernan

Johnson

Crofford

, et al. Posttraumatic stress symptoms mediate the effects of trauma exposure on clinical indicators of central sensitization in patients with chronic pain. Clin J Pain 2019; 35: 385–393.

45.

Gasperi

Afari

Goldberg

, et al. Pain and trauma: The role of Criterion A trauma and stressful life events in the pain and PTSD Relationship. J Pain 2021; 22: 1506–1517.

46.

Olsen

Montgomery

Bøjholm

, et al. Prevalence of pain in the head, back and feet in refugees previously exposed to torture: a ten-year follow-up study. Disabil Rehabil 2007; 29: 163–171.

47.

Otis

Keane

Kerns

, et al. The development of an integrated treatment for veterans with comorbid chronic pain and posttraumatic stress disorder. Pain Med 2009; 10: 1300–1311.

48.

Dansk Hovedpine

Referenceprogram: Diagnostik og Behandling Af Hovedpinesygdomme Og Ansigtssmerter, 3. Udgave. København: DHOS, 2020.

49.

Blumenfeld

Tepper

Robbins

, et al. Effects of onabotulinumtoxinA treatment for chronic migraine on common comorbidities including depression and anxiety. J Neurol Neurosurg Psych 2019; 90: 353–360.

50.

Bendtsen

Munksgaard

Tassorelli

, et al. Disability, anxiety and depression associated with medication-overuse headache can be considerably reduced by detoxification and prophylactic treatment. Results from a multicentre, multinational study (COMOESTAS project). Cephalalgia 2014; 34: 426–433.

51.

Kjeldgaard

Forchhammer

Teasdale

, et al. Cognitive behavioural treatment for the chronic post-traumatic headache patient: a randomized controlled trial. J Headache Pain 2014; 15: 81.

52.

Wind

van der Aa

Knipscheer

, et al. The assessment of psychopathology among traumatized refugees: measurement invariance of the Harvard Trauma Questionnaire and the Hopkins Symptom Checklist-25 across five linguistic groups. Eur J Psychotraumatol 2017; 8: 1321357.

53.

Headache Classification Committee of the International Headache Society. The International Classification of Headache Disorders, 3rd edition. Cephalalgia 2018; 38: 1–211.

54.

Schwartz

Bradley

Penza

, et al. Pain medication use among patients with posttraumatic stress disorder. Psychosomatics 2006; 47: 136–142.