Browse the corpus

Pain after combat injury in male UK military personnel deployed to Afghanistan. BACKGROUND: Chronic pain after injury poses a serious health burden. As a result of advances in medical technology, ever more military personnel survive severe combat injuries, but long-term pain outcomes are unknown. We aimed to assess rates of pain in a representative sample of UK military personnel with and without combat injuries. METHODS: We used data from the ADVANCE cohort study (ISRCTN57285353). Individuals deployed as UK armed forces to Afghanistan were recruited to include those with physical combat injuries, and a frequency-matched uninjured comparison group. Participants completed self-reported questionnaires, including 'overall' pain intensity and self-assessment of post-traumatic stress disorder, anxiety, and depression. RESULTS: A total of 579 participants with combat injury, including 161 with amputations, and 565 uninjured participants were included in the analysis (median 8 yr since injury/deployment). Frequency of moderate or severe pain was 18% (n=202), and was higher in the injured group (n=140, 24%) compared with the uninjured group (n=62, 11%, relative risk: 1.1, 95% confidence interval [CI]: 1.0-1.2, P<0.001), and lower in the amputation injury subgroup (n=31, 19%) compared with the non-amputation injury subgroup (n=109, 26%, relative risk: 0.9, 95% CI: 0.9-1.0, P=0.034). Presence of at least moderate pain was associated with higher rates of post-traumatic stress (RR: 3.7, 95% CI: 2.7-5.0), anxiety (RR: 3.2, 95% CI: 2.4-4.3), and depression (RR: 3.4, 95% CI: 2.7-4.5) after accounting for injury. CONCLUSION: Combat injury, but not amputation, was associated with a higher frequency of moderate to severe pain intensity in this cohort, and pain was associated with adverse mental health outcomes.

This analysis reports on data from the ADVANCE cohort. Full details of the study, including its protocol, are published elsewhere.27 Injured individuals were recruited from a sample of UK armed forces personnel who were deployed to Afghanistan between 2002 and 2014 and had sustained severe physical combat injuries. Eligibility criteria for the injured group were: sustaining a physical combat injury while on deployment to Afghanistan, aeromedical evacuation as a result of the injury resulting in admission to a UK hospital, and no history of cardiovascular, liver, or renal disease before injury. As a comparison group, uninjured personnel were recruited, consisting of individuals who were frequency-matched to the injured cohort on their age, rank, regiment, deployment era, and role on deployment. The ADVANCE study has approval from the Ministry of Defence Research Ethics Committee (protocol number 357/PPE/12). All participants gave written informed consent. Because there were a very small number of female UK military combat casualties in Afghanistan only male UK injured personnel were included. Consenting participants underwent a ‘baseline’ assessment between August 5, 2015 and August 28, 2020. This consisted of a 1-day study visit to the Defence Medical Rehabilitation Centre (DMRC) Headley Court or Stanford Hall, where participants completed self-report questionnaires and underwent clinical examination and targeted cardiovascular, musculoskeletal, and audiological investigation. Participants will be invited to subsequent follow-up visits over 20 yr.27

Defence Medical Rehabilitation Centre (DMRC) Headley Court or Stanford Hall, where participants completed self-report questionnaires and underwent clinical examination and targeted cardiovascular, musculoskeletal, and audiological investigation. Participants will be invited to subsequent follow-up visits over 20 yr.27 Participants were asked to rate their overall pain impression as no pain, mild pain, moderate pain, severe pain, or extreme pain, as part of the European Quality of Life 5-Day scale (EQ5D29). This was used as a measure of ‘overall pain’.

Defence Medical Rehabilitation Centre (DMRC) Headley Court or Stanford Hall, where participants completed self-report questionnaires and underwent clinical examination and targeted cardiovascular, musculoskeletal, and audiological investigation. Participants will be invited to subsequent follow-up visits over 20 yr.27 Participants were asked to rate their overall pain impression as no pain, mild pain, moderate pain, severe pain, or extreme pain, as part of the European Quality of Life 5-Day scale (EQ5D29). This was used as a measure of ‘overall pain’. For site-specific pain, participants were asked to rate their pain intensity during the last week, if any, on a 0–10 numerical rating scale (NRS), 0 indicating no pain at all and 10 indicating the worst pain imaginable. Participants were asked separately to rate (a) pain in the right/left hip or knee (joint pain, if multiple reported, the average was used), and (b) back pain, (c) phantom limb pain (amputees only), and (d) residual limb pain (amputees only). For each of these pain categories, we determined the frequency of participants with (a) no pain (0 on the NRS), (b) mild pain (1–3), (c) moderate pain (4–7), and (d) severe pain (8–10). A score of ≥4 on an 11-point NRS (and at least ‘moderate’ pain), has been demonstrated to distinguish mild from ‘impactful’ or ‘bothersome’ pain when pain intensity measures are compared with patient-reported measures of the impact of pain on physical and psychological function.30,31 Therefore a comparison of frequency of no-to-mild and at least moderate (including extreme) pain was used as the main outcome.

istinguish mild from ‘impactful’ or ‘bothersome’ pain when pain intensity measures are compared with patient-reported measures of the impact of pain on physical and psychological function.30,31 Therefore a comparison of frequency of no-to-mild and at least moderate (including extreme) pain was used as the main outcome. PTSD was measured using the PTSD Checklist (PCL) questionnaire,32 a 17-item inventory that records symptoms of PTSD over the previous month. The cut-off for probable PTSD was ≥50 (range 17–85). Depression was measured using the Patient Health Questionnaire-9 (PHQ9)33 asking for symptoms from the previous 2 weeks in nine items (cut-off ≥10; range 0–27). The Generalised Anxiety Disorder-7 (GAD7)34 questionnaire was used to assess anxiety in sevens items (cut-off ≥10; range 0–21). Global impression of health was assessed using the EQ5D perceived health scale, ranging from 0 (worst health possible) to 100 (best health possible). Frequencies of mild/moderate/severe (including extreme where relevant) pain are presented as absolute numbers. In addition, weighted percentages accounting for undersampling of less severe injury and response characteristics (age, rank, and branch service at time of sampling) are presented along with unweighted cell counts.

encies of mild/moderate/severe (including extreme where relevant) pain are presented as absolute numbers. In addition, weighted percentages accounting for undersampling of less severe injury and response characteristics (age, rank, and branch service at time of sampling) are presented along with unweighted cell counts. Log-binomial regression models were used to investigate the association between injury status (uninjured vs injured and no amputation vs amputation) and overall pain, joint pain, and back pain. Relative risks (RRs) are reported along with 95% confidence intervals (CIs). In a prespecified subgroup analysis, using the same outcomes, we compared those without amputation injury with those who had an amputation. All models were adjusted by including socioeconomic status (defined as military rank at sampling) and age at assessment. Similarly log-binomial regression models were used to investigate the association between overall pain and mental health outcomes (PTSD, depression, and anxiety). To investigate the association of pain and global impression of health, we used an ordinal regression model to compare global impression of health between participants with and without any moderate to severe pain and with and without injury to derive the odds ratio (OR) and 95% CI. Models were adjusted by including exposure, socioeconomic status (defined as military rank at sampling), and age at assessment.

regression model to compare global impression of health between participants with and without any moderate to severe pain and with and without injury to derive the odds ratio (OR) and 95% CI. Models were adjusted by including exposure, socioeconomic status (defined as military rank at sampling), and age at assessment. The missing value rate was <1% per variable. Missing values were not imputed; however, no reported pain was not considered missing, but was considered ‘no pain’, as pain can only be self-reported.

Participants were asked to rate their overall pain impression as no pain, mild pain, moderate pain, severe pain, or extreme pain, as part of the European Quality of Life 5-Day scale (EQ5D29). This was used as a measure of ‘overall pain’. For site-specific pain, participants were asked to rate their pain intensity during the last week, if any, on a 0–10 numerical rating scale (NRS), 0 indicating no pain at all and 10 indicating the worst pain imaginable. Participants were asked separately to rate (a) pain in the right/left hip or knee (joint pain, if multiple reported, the average was used), and (b) back pain, (c) phantom limb pain (amputees only), and (d) residual limb pain (amputees only). For each of these pain categories, we determined the frequency of participants with (a) no pain (0 on the NRS), (b) mild pain (1–3), (c) moderate pain (4–7), and (d) severe pain (8–10). A score of ≥4 on an 11-point NRS (and at least ‘moderate’ pain), has been demonstrated to distinguish mild from ‘impactful’ or ‘bothersome’ pain when pain intensity measures are compared with patient-reported measures of the impact of pain on physical and psychological function.30,31 Therefore a comparison of frequency of no-to-mild and at least moderate (including extreme) pain was used as the main outcome.

istinguish mild from ‘impactful’ or ‘bothersome’ pain when pain intensity measures are compared with patient-reported measures of the impact of pain on physical and psychological function.30,31 Therefore a comparison of frequency of no-to-mild and at least moderate (including extreme) pain was used as the main outcome. PTSD was measured using the PTSD Checklist (PCL) questionnaire,32 a 17-item inventory that records symptoms of PTSD over the previous month. The cut-off for probable PTSD was ≥50 (range 17–85). Depression was measured using the Patient Health Questionnaire-9 (PHQ9)33 asking for symptoms from the previous 2 weeks in nine items (cut-off ≥10; range 0–27). The Generalised Anxiety Disorder-7 (GAD7)34 questionnaire was used to assess anxiety in sevens items (cut-off ≥10; range 0–21). Global impression of health was assessed using the EQ5D perceived health scale, ranging from 0 (worst health possible) to 100 (best health possible).

Frequencies of mild/moderate/severe (including extreme where relevant) pain are presented as absolute numbers. In addition, weighted percentages accounting for undersampling of less severe injury and response characteristics (age, rank, and branch service at time of sampling) are presented along with unweighted cell counts. Log-binomial regression models were used to investigate the association between injury status (uninjured vs injured and no amputation vs amputation) and overall pain, joint pain, and back pain. Relative risks (RRs) are reported along with 95% confidence intervals (CIs). In a prespecified subgroup analysis, using the same outcomes, we compared those without amputation injury with those who had an amputation. All models were adjusted by including socioeconomic status (defined as military rank at sampling) and age at assessment. Similarly log-binomial regression models were used to investigate the association between overall pain and mental health outcomes (PTSD, depression, and anxiety). To investigate the association of pain and global impression of health, we used an ordinal regression model to compare global impression of health between participants with and without any moderate to severe pain and with and without injury to derive the odds ratio (OR) and 95% CI. Models were adjusted by including exposure, socioeconomic status (defined as military rank at sampling), and age at assessment.

In total, 579 injured participants and 565 uninjured participants were included in this analysis. Basic patient characteristics and group characteristics can be found in Table 1. Participants were assessed on median 8 yr after deployment/injury, and were median 33 yr old, and predominantly white (90%). Only male participants were included.Table 1Cohort overview and patient characteristics. All participants are male. Percentages are based on study weights. IQR, inter-quartile range. ∗Multiple mechanisms of injury could be present in a single participant.Table 1Total cohort (n=1144)Uninjured group (n=565)Injured group (n=579)Non-amputation injury group (n=418)Amputation injury group (n=161)Age at assessment (yr), median (IQR)33 (7)34 (7)33 (7)33 (8)32 (6)Years since injury, median (IQR)8 (3)9 (3)7 (3)New injury severity score, median (IQR)13 (25)9 (13)34 (27)Mechanism of injury, n (%)∗Blast434 (74)281 (67)153 (95)Gunshot135 (24)126 (30)9 (5)Other13 (2)10 (2)3 (2)Ethnicity, n (%)White1036 (90)512 (90)524 (90)376 (90)148 (92)All other ethnic groups108 (10)53 (10)55 (10)42 (10)13 (8)Serving status at assessment, n (%)Left service520 (48)99 (18)421 (73)277 (67)144 (90)Still serving624 (52)466 (82)158 (27)141 (33)17 (10)Engagement type at time of sampled deployment, n (%)Regular1089 (95)545 (96)544 (94)387 (93)157 (98)Reserve55 (5)20 (4)35 (6)31 (7)4 (2)Rank at sampled deployment, n (%)Junior (OR1-4)753 (72)339 (67)413 (77)285 (74)128 (84)Senior (OR5-9)253 (21)147 (25)106 (17)86 (19)20 (11)Officer (OF1-9)138 (7)79 (9)60 (6)47 (7)13 (5)

ngagement type at time of sampled deployment, n (%)Regular1089 (95)545 (96)544 (94)387 (93)157 (98)Reserve55 (5)20 (4)35 (6)31 (7)4 (2)Rank at sampled deployment, n (%)Junior (OR1-4)753 (72)339 (67)413 (77)285 (74)128 (84)Senior (OR5-9)253 (21)147 (25)106 (17)86 (19)20 (11)Officer (OF1-9)138 (7)79 (9)60 (6)47 (7)13 (5) Cohort overview and patient characteristics. All participants are male. Percentages are based on study weights. IQR, inter-quartile range. ∗Multiple mechanisms of injury could be present in a single participant.

ngagement type at time of sampled deployment, n (%)Regular1089 (95)545 (96)544 (94)387 (93)157 (98)Reserve55 (5)20 (4)35 (6)31 (7)4 (2)Rank at sampled deployment, n (%)Junior (OR1-4)753 (72)339 (67)413 (77)285 (74)128 (84)Senior (OR5-9)253 (21)147 (25)106 (17)86 (19)20 (11)Officer (OF1-9)138 (7)79 (9)60 (6)47 (7)13 (5) Cohort overview and patient characteristics. All participants are male. Percentages are based on study weights. IQR, inter-quartile range. ∗Multiple mechanisms of injury could be present in a single participant. Frequency of no, mild, moderate, severe (and extreme) pain can be found in Table 2. Overall pain was reported by 704 (62%) of the cohort. At least moderate pain was reported by 202 participants (18%) and was more prevalent in the injured group (n=140 [24%]) compared with the uninjured group (n=62 [11%], RR: 1.1, 95% CI: 1.0–1.2, P<0.001).Table 2Frequency of no, mild, moderate, and severe pain across the cohort and subgroups. Percentages are based on study weights.Table 2Overall painNo pain, n (%)440 (39)279 (50)161 (29)115 (29)46 (30)Mild pain, n (%)502 (43)224 (39)278 (46)194 (45)84 (50)Moderate pain, n (%)164 (15)57 (10)107 (18)83 (19)24 (15)Severe pain, n (%)31 (3)4 (1)27 (5)21 (5)6 (4)Extreme pain, n (%)7 (1)1 (<1)6 (1)5 (1)1 (1)Joint painNo joint pain, n (%)483 (42)262 (46)221 (38)148 (35)73 (47)Mild joint pain, n (%)464 (40)226 (39)238 (40)177 (42)61 (35)Moderate joint pain, n (%)188 (17)75 (14)113 (20)87 (21)26 (18)Severe joint pain, n (%)9 (1)2 (0)7 (1)6 (2)1 (1)Back painNo back pain, n (%)412 (36)237 (42)175 (30)117 (28)58 (36)Mild back pain, n (%)461 (40)225 (39)236 (40)180 (42)56 (34)Moderate back pain, n (%)237 (21)92 (17)145 (25)104 (25)41 (25)Severe back pain, n (%)34 (4)11 (2)23 (5)17 (5)6 (4)

75 (14)113 (20)87 (21)26 (18)Severe joint pain, n (%)9 (1)2 (0)7 (1)6 (2)1 (1)Back painNo back pain, n (%)412 (36)237 (42)175 (30)117 (28)58 (36)Mild back pain, n (%)461 (40)225 (39)236 (40)180 (42)56 (34)Moderate back pain, n (%)237 (21)92 (17)145 (25)104 (25)41 (25)Severe back pain, n (%)34 (4)11 (2)23 (5)17 (5)6 (4) Frequency of no, mild, moderate, and severe pain across the cohort and subgroups. Percentages are based on study weights. The injured group reported higher rates of at least moderate joint (n=120 [22%] vs 77 [15%], RR: 1.1, 95% CI: 1.0–1.1, P=0.001) and back pain (n=168 [30%] vs n=103 [19%], RR: 1.1, 95% CI: 1.1–1.2, P<0.001) compared with the uninjured group. The amputation cohort included 161 participants (lower limb: n=157, upper limb: n=17, single limb amputation: n=85, multiple limb amputation: n=76). The amputation injury subgroup reported lower rates of at least moderate overall pain (n=31 [19%]) compared with the non-amputation injury subgroup (n=109 [26%], RR: 0.9, 95% CI: 0.9–1.0, P=0.034). Rates of joint pain were higher for the non-amputees compared with amputees, however, after adjusting for confounders there was only weak evidence of an association (n=93 [23%] vs n=27 [18%], RR: 0.9, 95% CI: 0.8–1.0, P=0.080). No difference was found for back pain (n=121 [30%] vs n=47 [29%], RR: 1.0, 95% CI: 0.9–1.1, P=0.810).

joint pain were higher for the non-amputees compared with amputees, however, after adjusting for confounders there was only weak evidence of an association (n=93 [23%] vs n=27 [18%], RR: 0.9, 95% CI: 0.8–1.0, P=0.080). No difference was found for back pain (n=121 [30%] vs n=47 [29%], RR: 1.0, 95% CI: 0.9–1.1, P=0.810). In addition to any joint or back pain, 115 (71%) reported phantom limb pain and 122 (76%) reported residual limb pain. At least moderate phantom limb pain was reported by 52 participants (32%), and 51 (32%) reported at least moderate residual limb pain, with 31 (19%) reporting both at least moderate phantom and residual limb pain.

to any joint or back pain, 115 (71%) reported phantom limb pain and 122 (76%) reported residual limb pain. At least moderate phantom limb pain was reported by 52 participants (32%), and 51 (32%) reported at least moderate residual limb pain, with 31 (19%) reporting both at least moderate phantom and residual limb pain. Mental health outcomes in relation to pain are displayed in Fig. 1, and in further detail in relation to cohort subgroups and pain are presented in Table 3. As reported previously,28 frequency of PTSD, depression, and anxiety increased in the injured compared with the uninjured group. There was an additional association between at least moderate pain and adverse mental health outcomes after taking exposure into account (for PTSD, RR: 3.7, 95% CI: 2.7–5.0, P<0.001; for depression, RR: 3.4, 95% CI: 2.7–4.5, P<0.001; for anxiety, RR: 3.2, 95% CI: 2.4–4.3, P<0.001). In the injured group, rates of PTSD, depression, and anxiety were four times higher comparing those with at least moderate pain to no/mild pain (see Table 3).Fig 1Reported PTSD (post-traumatic stress disorder), depression, and anxiety for all participants, and separately for those with and without at least moderate pain, indicating increased frequency of adverse mental health outcomes in the presence of pain. Percentages are based on study weights.Fig 1Table 3Mental health outcomes in relation to cohort subgroup and pain. Percentages are based on study weights. PTSD, post-traumatic stress disorder. ∗Weighted percentage of cohort or group (e.g. of uninjured or injured, percentage of columns). †Weighted percentage of pain subgroup (e.g. of injured with no or mild pain, percentage of rows).Table 3Total cohort (n=1144)Uninjured group (n=565)Injured group (n=579)PTSD, n (%∗)133 (13)47 (9)86 (16)No or mild pain, n (%†)69 (8)34 (8)35 (9)At least moderate pain, n (%†)64 (35)13 (22)51 (40)Depression, n (%∗)193 (18)77 (15)116 (21)No or mild pain, n (%†)106 (12)54 (12)52 (12)At least moderate pain, n (%†)87 (46)23 (40)64 (49)Anxiety, n (%∗)159 (15)59 (11)100 (19)No or mild pain, n (%†)89 (10)43 (9)46 (11)At least moderate pain, n (%†)70 (38)16 (28)54 (42)

At least moderate pain, n (%†)64 (35)13 (22)51 (40)Depression, n (%∗)193 (18)77 (15)116 (21)No or mild pain, n (%†)106 (12)54 (12)52 (12)At least moderate pain, n (%†)87 (46)23 (40)64 (49)Anxiety, n (%∗)159 (15)59 (11)100 (19)No or mild pain, n (%†)89 (10)43 (9)46 (11)At least moderate pain, n (%†)70 (38)16 (28)54 (42) Reported PTSD (post-traumatic stress disorder), depression, and anxiety for all participants, and separately for those with and without at least moderate pain, indicating increased frequency of adverse mental health outcomes in the presence of pain. Percentages are based on study weights. Mental health outcomes in relation to cohort subgroup and pain. Percentages are based on study weights. PTSD, post-traumatic stress disorder. ∗Weighted percentage of cohort or group (e.g. of uninjured or injured, percentage of columns). †Weighted percentage of pain subgroup (e.g. of injured with no or mild pain, percentage of rows). Those with at least moderate pain had a lower global impression of health than those with no/mild pain (OR: 3.1, 95% CI: 2.5–3.9, P<0.001) whereas injury had no effect on global impression of health (OR: 1.2, 95% CI: 0.9–1.4, P=0.164, see Fig. 2).Fig 2Reported global impression of health (0=worst health possible, 100=best health possible) participants with no or mild pain (orange bars) and at least moderate pain (red bars) for all participants and uninjured and injured groups as boxplots.Fig 2

impression of health (OR: 1.2, 95% CI: 0.9–1.4, P=0.164, see Fig. 2).Fig 2Reported global impression of health (0=worst health possible, 100=best health possible) participants with no or mild pain (orange bars) and at least moderate pain (red bars) for all participants and uninjured and injured groups as boxplots.Fig 2 Reported global impression of health (0=worst health possible, 100=best health possible) participants with no or mild pain (orange bars) and at least moderate pain (red bars) for all participants and uninjured and injured groups as boxplots.

Frequency of no, mild, moderate, severe (and extreme) pain can be found in Table 2. Overall pain was reported by 704 (62%) of the cohort. At least moderate pain was reported by 202 participants (18%) and was more prevalent in the injured group (n=140 [24%]) compared with the uninjured group (n=62 [11%], RR: 1.1, 95% CI: 1.0–1.2, P<0.001).Table 2Frequency of no, mild, moderate, and severe pain across the cohort and subgroups. Percentages are based on study weights.Table 2Overall painNo pain, n (%)440 (39)279 (50)161 (29)115 (29)46 (30)Mild pain, n (%)502 (43)224 (39)278 (46)194 (45)84 (50)Moderate pain, n (%)164 (15)57 (10)107 (18)83 (19)24 (15)Severe pain, n (%)31 (3)4 (1)27 (5)21 (5)6 (4)Extreme pain, n (%)7 (1)1 (<1)6 (1)5 (1)1 (1)Joint painNo joint pain, n (%)483 (42)262 (46)221 (38)148 (35)73 (47)Mild joint pain, n (%)464 (40)226 (39)238 (40)177 (42)61 (35)Moderate joint pain, n (%)188 (17)75 (14)113 (20)87 (21)26 (18)Severe joint pain, n (%)9 (1)2 (0)7 (1)6 (2)1 (1)Back painNo back pain, n (%)412 (36)237 (42)175 (30)117 (28)58 (36)Mild back pain, n (%)461 (40)225 (39)236 (40)180 (42)56 (34)Moderate back pain, n (%)237 (21)92 (17)145 (25)104 (25)41 (25)Severe back pain, n (%)34 (4)11 (2)23 (5)17 (5)6 (4) Frequency of no, mild, moderate, and severe pain across the cohort and subgroups. Percentages are based on study weights. The injured group reported higher rates of at least moderate joint (n=120 [22%] vs 77 [15%], RR: 1.1, 95% CI: 1.0–1.1, P=0.001) and back pain (n=168 [30%] vs n=103 [19%], RR: 1.1, 95% CI: 1.1–1.2, P<0.001) compared with the uninjured group.

Frequency of no, mild, moderate, and severe pain across the cohort and subgroups. Percentages are based on study weights. The injured group reported higher rates of at least moderate joint (n=120 [22%] vs 77 [15%], RR: 1.1, 95% CI: 1.0–1.1, P=0.001) and back pain (n=168 [30%] vs n=103 [19%], RR: 1.1, 95% CI: 1.1–1.2, P<0.001) compared with the uninjured group. The amputation cohort included 161 participants (lower limb: n=157, upper limb: n=17, single limb amputation: n=85, multiple limb amputation: n=76). The amputation injury subgroup reported lower rates of at least moderate overall pain (n=31 [19%]) compared with the non-amputation injury subgroup (n=109 [26%], RR: 0.9, 95% CI: 0.9–1.0, P=0.034). Rates of joint pain were higher for the non-amputees compared with amputees, however, after adjusting for confounders there was only weak evidence of an association (n=93 [23%] vs n=27 [18%], RR: 0.9, 95% CI: 0.8–1.0, P=0.080). No difference was found for back pain (n=121 [30%] vs n=47 [29%], RR: 1.0, 95% CI: 0.9–1.1, P=0.810). In addition to any joint or back pain, 115 (71%) reported phantom limb pain and 122 (76%) reported residual limb pain. At least moderate phantom limb pain was reported by 52 participants (32%), and 51 (32%) reported at least moderate residual limb pain, with 31 (19%) reporting both at least moderate phantom and residual limb pain.

Mental health outcomes in relation to pain are displayed in Fig. 1, and in further detail in relation to cohort subgroups and pain are presented in Table 3. As reported previously,28 frequency of PTSD, depression, and anxiety increased in the injured compared with the uninjured group. There was an additional association between at least moderate pain and adverse mental health outcomes after taking exposure into account (for PTSD, RR: 3.7, 95% CI: 2.7–5.0, P<0.001; for depression, RR: 3.4, 95% CI: 2.7–4.5, P<0.001; for anxiety, RR: 3.2, 95% CI: 2.4–4.3, P<0.001). In the injured group, rates of PTSD, depression, and anxiety were four times higher comparing those with at least moderate pain to no/mild pain (see Table 3).Fig 1Reported PTSD (post-traumatic stress disorder), depression, and anxiety for all participants, and separately for those with and without at least moderate pain, indicating increased frequency of adverse mental health outcomes in the presence of pain. Percentages are based on study weights.Fig 1Table 3Mental health outcomes in relation to cohort subgroup and pain. Percentages are based on study weights. PTSD, post-traumatic stress disorder. ∗Weighted percentage of cohort or group (e.g. of uninjured or injured, percentage of columns). †Weighted percentage of pain subgroup (e.g. of injured with no or mild pain, percentage of rows).Table 3Total cohort (n=1144)Uninjured group (n=565)Injured group (n=579)PTSD, n (%∗)133 (13)47 (9)86 (16)No or mild pain, n (%†)69 (8)34 (8)35 (9)At least moderate pain, n (%†)64 (35)13 (22)51 (40)Depression, n (%∗)193 (18)77 (15)116 (21)No or mild pain, n (%†)106 (12)54 (12)52 (12)At least moderate pain, n (%†)87 (46)23 (40)64 (49)Anxiety, n (%∗)159 (15)59 (11)100 (19)No or mild pain, n (%†)89 (10)43 (9)46 (11)At least moderate pain, n (%†)70 (38)16 (28)54 (42)

We report the frequency and intensity of pain associated with combat injury, compared with the non-injured, in a large cohort of military personnel 8 yr post-injury/deployment. Although a report of overall pain was high across this cohort (61%), ‘impactful’ pain is often described as at least moderate pain, affecting a much lower proportion (18%). Of the uninjured group, 50% reported pain in our cohort, with 11% stating at least moderate intensity, which, despite differences in sex and age, is roughly comparable to estimated rates in the UK general population: overall pain prevalence is estimated to be between 35% and 75%35,36 with 10–25% thought to have at least moderately disabling pain. Painful musculoskeletal injury, unrelated to specific combat trauma, is common in military cohorts after deployment, possibly as a consequence of load carrying and a high burden of physical activity.37,38 For example, just under half of a US cohort of military personnel reported a new onset of low back pain after deployment, unrelated to injury.15 Pain after combat injury is well recognised2,39, 40, 41 and represents a major health and societal burden as it has been associated with higher healthcare and opioid use, vocational limitation, and family discordance.39,42

military personnel reported a new onset of low back pain after deployment, unrelated to injury.15 Pain after combat injury is well recognised2,39, 40, 41 and represents a major health and societal burden as it has been associated with higher healthcare and opioid use, vocational limitation, and family discordance.39,42 Although traumatic injury increased the risk of ‘impactful’ pain, it was lower than might have been expected, given the intensity of pain reported in other trauma cohorts, both military and civilian.9,43 A number of factors may contribute to this, including differences in demographic features, injury characteristics, healthcare provision, post-injury rehabilitation, pain management strategies, or differences in methods of reporting and measuring pain.

in reported in other trauma cohorts, both military and civilian.9,43 A number of factors may contribute to this, including differences in demographic features, injury characteristics, healthcare provision, post-injury rehabilitation, pain management strategies, or differences in methods of reporting and measuring pain. The association between combat injury involving joints and post-traumatic osteoarthritis44 could be an explanation for the prevalence of joint pain in our cohort. Intra-articular fracture, meniscal and ligamentous injury are common in modern combat trauma, particularly in the lower extremities.12,45 The injury-related long-term aberrant loading to cartilage, subchondral bone, and synovium can cause osteoarthritis46 which commonly presents with pain, reduced physical ability, and the potential for removal from active duty.44 Chronic low back pain has also been associated with blast injury and other musculoskeletal injuries amongst military populations.47 However, the majority of studies aimed at identifying risk for these types of chronic pain conditions in military populations have been cross-sectional in nature. It is anticipated that future longitudinal assessment of the ADVANCE cohort, and the analysis of specific joint and back pain-related patient-reported outcome measures will provide more robust information about the trajectory of these conditions after traumatic injury.

y populations have been cross-sectional in nature. It is anticipated that future longitudinal assessment of the ADVANCE cohort, and the analysis of specific joint and back pain-related patient-reported outcome measures will provide more robust information about the trajectory of these conditions after traumatic injury. Although combat-related amputation has a significant impact on quality of life, it is relatively rare compared with other types of injury and to other aetiologies for chronic pain in deployed personnel. Our data suggest that amputation does not increase the frequency of joint or back pain compared with those with non-amputation injury. This is important as studies involving amputees rarely include a comparable group that have undergone injury without limb loss. This could mean that major traumatic injury, not limb loss per se, is a risk factor for chronic pain at distant sites after injury. Amputation, although associated with phantom limb and residual limb pain, may not necessarily lead to poorer pain-related outcomes in severe injury. In specific populations, no difference between chronic pain frequency has been identified between limb salvage and amputation groups in combat injury settings.15,16 This may account for the limited difference noted in overall pain prevalence between the amputation and non-amputation injury subgroups in our dataset.

ry. In specific populations, no difference between chronic pain frequency has been identified between limb salvage and amputation groups in combat injury settings.15,16 This may account for the limited difference noted in overall pain prevalence between the amputation and non-amputation injury subgroups in our dataset. Frequency of at least moderate phantom limb pain in our amputee cohort (32%) may be lower than other similar cohorts where rates of 56–58% have been reported,4,47,48 although this difference may be because of differing methods of pain measurement and cut-offs. Interestingly, the frequency of at least moderate ‘overall pain’ was lower than site-specific pain, perhaps indicating different perception of pain when focussed on a specific body part or different sensitivities of the questionnaires used.

ifference may be because of differing methods of pain measurement and cut-offs. Interestingly, the frequency of at least moderate ‘overall pain’ was lower than site-specific pain, perhaps indicating different perception of pain when focussed on a specific body part or different sensitivities of the questionnaires used. The frequency of PTSD was significant in our cohort, present in 14%, and slightly lower than the frequency described in other studies of UK military personnel in combat roles.48 In our dataset, PTSD was commonly comorbid with pain, which supports findings from other military5, 6, 7, 8,18 and non-military settings.49 In other studies, comorbid pain and PTSD have a greater impact on function than either on its own, and a detrimental influence on the ability to engage with therapy. Therefore, understanding the interaction may elucidate the mechanisms involved in the overlapping conditions, and provide a potential method for exploiting each as a therapeutic intervention for the other condition. For example, it has also been suggested that mitigation of severe pain at point of injury may prevent the development of PTSD.50, 51, 52, 53 Conversely, Li and colleagues54 demonstrated that emotional numbness and avoidance were responsible for the largest variance in pain-related physical function—by targeting these psychological domains, improvement in pain-related function may be achieved.54

nt of injury may prevent the development of PTSD.50, 51, 52, 53 Conversely, Li and colleagues54 demonstrated that emotional numbness and avoidance were responsible for the largest variance in pain-related physical function—by targeting these psychological domains, improvement in pain-related function may be achieved.54 We found that rates of PTSD, anxiety, and depression were elevated when accompanied by pain, while the global impression of perceived health decreased. Adverse mental health outcomes and chronic pain have been long shown to affect each other, and it should be noted that we cannot with certainty assert the direction of the effect in this work. There are limited longitudinal studies to provide evidence to more fully understand the proposed models of how pain and PTSD are interconnected.25 Some evidence suggests that baseline PTSD predicts subsequent physical symptoms, and that baseline pain predicts PTSD,6,55 and a recent longitudinal study indicated that although an initial bidirectional relationship was observed between physical symptoms and PTSD, subsequently a unidirectional interaction was determined, with an increase in physical symptoms apparently driving the increase in PTSD.56 Certainly, both pain and PTSD appear to share certain vulnerabilities, including catastrophising, attentional bias, hypervigilance,26 and even genetic risk.57 The longitudinal nature of the ADVANCE study will generate long-term insights into this question and provide an opportunity to test these hypotheses.

in PTSD.56 Certainly, both pain and PTSD appear to share certain vulnerabilities, including catastrophising, attentional bias, hypervigilance,26 and even genetic risk.57 The longitudinal nature of the ADVANCE study will generate long-term insights into this question and provide an opportunity to test these hypotheses. These data allow neither mechanistic classification nor an assessment of the impact of pain specifically on physical and psychological domains. Measurement of pain using a unidimensional intensity scale cannot provide a complete assessment of the biopsychosocial nature of chronic pain. The current cross-sectional analysis only includes a single snapshot of pain and mental health outcomes, therefore the direction of relationships between these measures cannot be determined. It is anticipated that future data collection time points will provide further insight into these complex relationships. This analysis did not follow a prespecified protocol, and while we have aimed to avoid inflation of hypothesis-testing statistics and reducing the reliance on P-values, subsequent longitudinal analyses will have to be prespecified and registered in a transparent process.

urther insight into these complex relationships. This analysis did not follow a prespecified protocol, and while we have aimed to avoid inflation of hypothesis-testing statistics and reducing the reliance on P-values, subsequent longitudinal analyses will have to be prespecified and registered in a transparent process. Just under one fifth of uninjured and injured UK military personnel report moderate-to-severe pain, several years after deployment to Afghanistan. Pain was more frequently reported in those who sustained a severe combat injury than in an uninjured matched control cohort. Comorbidity of pain and mental health outcomes was described, and participants living with at least moderate pain reported a lower median global impression of health. This demonstrates a potential area for further exploration to understand pain-generating mechanisms in traumatic injury and may represent a potential therapeutic target for the management of chronic pain.

These data allow neither mechanistic classification nor an assessment of the impact of pain specifically on physical and psychological domains. Measurement of pain using a unidimensional intensity scale cannot provide a complete assessment of the biopsychosocial nature of chronic pain. The current cross-sectional analysis only includes a single snapshot of pain and mental health outcomes, therefore the direction of relationships between these measures cannot be determined. It is anticipated that future data collection time points will provide further insight into these complex relationships. This analysis did not follow a prespecified protocol, and while we have aimed to avoid inflation of hypothesis-testing statistics and reducing the reliance on P-values, subsequent longitudinal analyses will have to be prespecified and registered in a transparent process.

Just under one fifth of uninjured and injured UK military personnel report moderate-to-severe pain, several years after deployment to Afghanistan. Pain was more frequently reported in those who sustained a severe combat injury than in an uninjured matched control cohort. Comorbidity of pain and mental health outcomes was described, and participants living with at least moderate pain reported a lower median global impression of health. This demonstrates a potential area for further exploration to understand pain-generating mechanisms in traumatic injury and may represent a potential therapeutic target for the management of chronic pain.

Study design and management: PC, NCT, AB, AMJB, CB Experiments: EMPZ, WM, CW, MK, DCR, FF, TM, KS, LS, DF, PLH, EK Data acquisition, handling, and analysis: JV, AK, ECC, DD, ZG, EFM, SS, SS Drafting the manuscript: JV, HIK Revising the manuscript: all authors

The ADVANCE study is funded through the ADVANCE Charity. Key contributors to this charity are the Headley Court Charity (principal funder); HM Treasury (LIBOR grant); Help for Heroes; Nuffield Trust for the Forces of the Crown; Forces in Mind Trust; National Lottery Community Fund; Blesma, The Limbless Veterans; and the UK Ministry of Defence. We wish to thank all of the research staff at both Headley Court and Stanford Hall who helped with the ADVANCE study, including Maria-Benedicta Edwards, Helen Blackman, Melanie Chesnokov, Emma Coady, Sarah Evans, Guy Fraser, Meliha Kaya-Barge, Maija Maskuniitty, David Pernet, Helen Prentice, Urszula Pucilowska, Lalji Varsani, Anna Verey, Molly Waldron, Danny Weston, Tass White, Seamus Wilson, and Louise Young.