Complex pain phenotypes: Suicidal ideation and attempt through latent multimorbidity.

BACKGROUND: Given the relatively high rates of suicidal ideation and attempt among people with chronic pain, there is a need to understand the underlying factors to target suicide prevention efforts. To date, no study has examined the association between pain phenotypes and suicide related behaviors among those with mild traumatic brain injuries.
OBJECTIVE: To determine if pain phenotypes were independently associated with suicidal ideation / attempt or if comorbidities within the pain phenotypes account for the association between pain phenotypes and suicide related behaviors.
METHODS: This is a longitudinal retrospective cohort study of suicide ideation/attempts among pain phenotypes previously derived using general mixture latent variable models of the joint distribution of repeated measures of pain scores and pain medications/treatment. We used national VA inpatient, outpatient, and pharmacy data files for Post-9/11 Veterans with mild traumatic injury who entered VA care between fiscal years (FY) 2007 and 2009. We considered a counterfactual causal modeling framework to assess the extent that the pain phenotypes during years 1-5 of VA care were predictive of suicide ideation/attempt during years 6-8 of VA care conditioned on covariates being balanced between pain phenotypes.
RESULTS: Without adjustment, pain phenotypes were significant predictors of suicide related behaviors. When we used propensity scores to balance the comorbidities present in the pain phenotypes, the pain phenotypes were no longer significantly associated with suicide related behaviors.
CONCLUSION: These findings suggest that suicide ideation/attempt is associated with pain trajectories primarily through latent multimorbidity. Therefore, it is critical to identify and manage comorbidities (e.g., depression, post-traumatic stress disorder) to prevent tragic outcomes associated with suicide related behaviors throughout the course of chronic pain and mild traumatic brain injury management.

Introduction

Mild traumatic injury (mTBI) is a signature injury of war in Iraq and Afghanistan Veterans [1, 2]. Although, the mechanism is unknown, mTBI is associated with up to a two-fold increased risk of suicide [3]. Following mTBI, musculoskeletal and headache pain is commonly reported. Pain is a factor that may interact with mTBI to increase the risk of suicide.

Some individuals with chronic pain consider suicide as an option to eliminate suffering and burden to others when pharmacological, medical, and surgical therapies are exhausted [4-9]. Suicidal ideation, attempts and completion are two to three times more likely in those with chronic pain than in those without pain [10-13]. The intersection between longitudinal patterns of pain, mTBI and suicide presents clear challenges to the patients and providers. However, prior research examining longitudinal patterns of pain trajectories, multifaceted treatment, and subsequent adverse outcomes such as suicidal ideation and attempt (suicide-related behavior; SRB) are limited among patients with mTBI [14].

Our prior work identified pain trajectory phenotypes in Post-9/11 Veterans with mTBI using latent trajectory models. We found that four pain trajectory phenotypes (complex low impact, stable pain; complex low impact, worsening pain; complex moderate impact, worsening pain; complex high impact, stable pain) characterized by different patterns of longitudinal pain trajectories and complex, multidimensional longitudinal treatment regimens including medications with dual indications for use (e.g., antidepressants, anticonvulsants) [15]. Because prior research in Post-9/11 Veterans has also found that chronic pain frequently occurs with mTBI, post-traumatic stress disorder (PTSD), and depression [16, 17], our examination includes not only the pain trajectory phenotypes (hereafter pain phenotypes), but also comorbid conditions that may be reflected in the complex treatment regimens reflected in the pain phenotypes. These comorbid conditions account for the mental pain and suffering that can contribute to suicide risk via defeatist thoughts and emotions that color future expectations and threaten a positive future [18].

Different components of complex pain may have additive effects on suicide ideation and attempt including medications, high pain scores, and accumulated non-cancer pain conditions [5, 19–23]. Complex comorbidity may add to the impact of these components of complex pain on SRB [24-26], or they may account for the relationship between pain phenotypes and SRB among Veterans with mTBI [3, 27–33]. We sought to determine if these pain phenotypes in mTBI were independently associated with SRB, or if the complexity of polymorbidity that may be a latent component of the pain phenotypes accounts for the association between pain phenotypes and SRB.

Materials and methods

Study cohort

The study sample and methods for development of pain phenotypes are described previously [15]. This study’s analytic sample included those who entered VA care October 1, 2007 through September 30, 2009, who were diagnosed with mild TBI (mTBI), and who had at least three years of care during the first 5 years after entering VA care [34]. The University of Texas Health Science Center at San Antonio, the University of Utah, and the Bedford VHA Hospital institutional review boards approved the study, with a waiver of informed consent.

Measures and data sources

Primary outcome: Suicide related behavior

We used ICD9-CM codes used in prior studies of SRB to identify suicide ideation (V6284) and attempt (E950, E952, E953, E953, E954, E955, E956, E957, E958, E959) in national VA inpatient and outpatient data in years 6–8 of VA care [35]. Due to small numbers of documented attempts, we examined three SRB categories: suicidal ideation without attempt, suicide attempt regardless of suicide ideation, and no SRB. We also examined the combined outcome of any SRB vs. no SRB.

Primary independent variable: Pain phenotype

As described in previous work by Song, et al. [15], pain phenotypes were derived from general mixture latent variable models based on repeated measures of pain scores and pain medications or other pain treatments (e.g., complex pain clinic, physical therapy) during years 1–5 of VA care. These models identified four pain phenotypes: ‘low impact-stable,’ ‘low impact- worsening,’ ‘moderate impact-worsening,’ and ‘high impact-stable.’

Covariates

We included age, sex, race/ethnicity, education, and military characteristics from the Operation Enduring Freedom/Operation Iraqi Freedom/ Operation New Dawn Roster file through December 2014, and percent service-connected disability from Veteran’s Service Network data. Age was classified as 18–29, 30–39, 40–49 and greater than or equal to 50 years. Sex was classified as male and female based on data reported from the Department of Defense presented at the time of entry to military service. Race/ethnicity was defined as Black, White, Hispanic, or other based on available numbers in each racial/ethnic group. Education was the education level at the time of leaving military service classified as high school or less and some post-high school education and higher. Military characteristics included branch of service (Army, Marine Corps, Navy/Coast Guard, and Air Force), component of service (Active Duty, National Guard/Reserve), military rank (enlisted, officer/warrant officer), and deployment history (single deployment, multiple deployments). Service-connected disability is awarded in 10% increments and was classified as 0–20%, 30–50%, 60–80%, and 90–100%.

We identified diagnosed health conditions hypothesized to be reflected in the pain phenotypes and associated with SRB [5, 10, 24–26, 36]. S1 Table provides ICD9-CM diagnosis codes used; the algorithm required two or more diagnoses at least 7 days apart in outpatient data or at least once in inpatient data during the period of pain trajectory class analysis (years 1–5 of VA care) [37, 38]. Conditions included as covariates were back/neck pain, other musculoskeletal pain, headache, posttraumatic stress disorder (PTSD), depression, anxiety, substance use disorder, insomnia, attention impairment (proxy for potential impulsiveness), and cognitive dysfunction.

Statistical analysis

Descriptive statistics for each pain phenotype were calculated. Chi-square tests were used to compare categorical variables between pain phenotypes.

We considered a counterfactual causal modeling framework [39, 40] to assess the extent that the pain phenotypes during years 1–5 of VA care were predictive of SRB during years 6–8 of VA care should covariates be balanced between pain phenotypes. The propensity scores of pain phenotypes were calculated, and the inverse propensity scores were incorporated as weights in the SRB model to minimize observed confounding for assessing the association of pain phenotype with SRB. To obtain robust estimates of propensity scores associated with pain phenotypes, we used generalized boosted multinomial regression modeling implemented by the twang package in R [41]. To examine the extent to which comorbidities contribute to differential SRB by pain phenotype, we considered two propensity score models with nested sets of predictors. Set 1 (long set) predictors were age, sex, race/ethnicity, education, military rank, multiple deployments, service-connected disability, prior SRB, and diagnosed health conditions that include headache, pain, anxiety, depression, PTSD, insomnia, substance use disorder, attention impairment and cognitive dysfunction. Set 2 (short set) predictors were a subset of Set 1 covariates excluding comorbid health conditions. Propensity score models were evaluated in terms of covariate balance between pain phenotypes using standardized bias—a standardized bias below the threshold 0.25 was deemed as covariate balance [42].

For SRB, we considered models for both a dichotomous and a trichotomous outcome. For any SRB or none, we conducted three separate logistic regression models to estimate the effects of pain phenotypes in SRB conditioned on varying degrees of adjustment of confounding associated with covariates, including (1) all covariates (long set) as predictors and inverse propensity score weights (IPSW) predicted by all covariates; (2) the short list of covariates (that excludes comorbid health conditions) as predictors and IPSW predicted by the short list of covariates; and (3) none. For the trichotomous SRB outcome (any attempt, ideation only, or no SRB), three multinomial logistic regression analyses were conducted with adjustments (1)-(3) as described above.

In each adjusted model of SRB, we included all predictors in the propensity score model (short or long set) as predictors along with the inverse propensity score weights adjustment so that the estimates associated with pain phenotypes have the ‘double-robustness’ property, i.e., the estimator is consistent either the propensity score model or the outcome model is correctly specified [43]. Under the four assumptions of consistency, exchangeability, positivity, and no misspecification of both outcome and propensity score models, the IPSW estimates of ORs associated with pain phenotypes based on adjustment (1) are interpreted as the effects of pain phenotypes conditioned on the counterfactual all covariates being balanced among phenotypes [44]. Similarly, the IPSW adjusted OR estimates based on the adjustment (2) are interpreted as the effects of pain phenotype on SRB should the partial set of covariates be balanced among pain phenotypes. Comparing estimates associated with pain phenotypes between these two IPSW adjustments will allow us to assess the extent to which the comorbid health conditions jointly attributed to the differential SRB risks between pain phenotypes.

Results

All 10,717 Veterans with mTBI had complex pain and multiple pain treatment modalities. Proportions of Veterans by phenotype were: ‘low impact-stable’ 33.8%, ‘low impact- worsening’ 19.1%, ‘moderate impact-worsening’ 33.2%, and ‘high impact-stable’ 18.7%. The ’high impact-stable’ phenotype had the highest prevalence of SRB followed by the ’low impact-worsening’ phenotype. Table 1 shows that pain phenotype was associated with well-known SRB predictors: younger age, race/ethnicity, education, rank, service branch, multiple deployments, service-connected disability, pain and mental health conditions, and prior suicidal ideation and attempts. The propensity score model for pain phenotypes predicted by the long set of covariates achieved balances of all covariates between pain phenotypes (standardized biases associated covariates all fell below the 0.25 threshold), but the propensity score model predicted by the short set of covariates did not necessarily balance the health covariates. See S2 Table for binomial logistic regression results and S3 and S4 Tables for the multinomial logistic regression results.

Table 1

Demographic and clinical characteristics by complex pain phenotype.

	Complex pain phenotype, N = 10717
	Low impact, stable		Low impact, worsening		Moderate impact, worsening		High impact, stable
Characteristics	n = 3454 (33.8%)		n = 1955 (19.1%)		n = 3393 (33.2%)		n = 1915 (18.7%)		P value
Age: ≤29	2479	72%	1289	66%	2136	63%	1088	57%	< .001
30–39	548	16%	429	22%	700	21%	517	27%
40–49	342	10%	190	10%	469	14%	269	14%
50+	85	2%	47	2%	88	3%	41	2%
Sex: Male	3226	93%	1828	94%	3184	94%	1779	93%	.61
Female	228	7%	127	6%	209	6%	136	7%
Race: Black	437	13%	204	10%	530	16%	199	10%	< .001
White	2430	70%	1400	72%	2274	67%	1393	73%
Hispanic	433	13%	263	13%	446	13%	236	12%
Other	154	4%	88	5%	143	4%	87	5%
Education: High school or less	3055	88%	1684	86%	3004	89%	1679	88%	.04
Some college plus	399	12%	271	14%	389	11%	236	12%
Rank: Enlisted	3346	97%	1891	97%	3322	98%	1872	98%	.01
Officer/Warrant	108	3%	64	3%	71	2%	43	2%
Service Branch: Army	2237	65%	1407	72%	2431	72%	1443	75%	< .001
Air Force	130	4%	68	3%	138	4%	86	4%
Navy/Coast Guard	240	7%	141	7%	245	7%	133	7%
Marines	847	25%	339	17%	579	17%	253	13%
Component: Active	2484	72%	1363	70%	2423	71%	1340	70%	.24
Reserve/National Guard	970	28%	592	30%	970	29%	575	30%
Multiple Deployments	1905	55%	976	50%	1627	48%	824	43%	< .001
Service Connected Disability: 0–20%	646	19%	187	10%	380	11%	163	9%	< .001
30–50%	670	19%	192	10%	373	11%	73	4%
60–80%	1400	41%	763	39%	1326	39%	556	29%
90–100%	738	21%	813	42%	1314	39%	1123	59%
Suicidal Ideation/Attempt, Y1-5: none	3262	94%	1679	86%	3108	92%	1573	82%	< .001
Suicide Attempt	28	1%	30	2%	29	1%	38	2%
Suicidal Ideation	145	4%	208	11%	214	6%	229	12%
Suicidal Ideation + Suicide Attempt	19	1%	38	2%	42	1%	75	4%
Comorbidities, Y1-5a
Headache	1327	38%	1071	55%	1968	58%	1337	70%	< .001
Back/neck pain	1700	49%	1298	66%	2776	82%	1746	91%	< .001
Other musculoskeletal pain	1507	44%	1126	58%	2343	69%	1628	85%	< .001
Anxiety	993	29%	882	45%	1107	33%	867	45%	< .001
Depression	1608	47%	1386	71%	2025	60%	1477	77%	< .001
Post-traumatic stress disorder	2534	73%	1820	93%	2846	84%	1821	95%	< .001
Insomnia	645	19%	800	41%	812	24%	784	41%	< .001
Substance use disorder	1071	31%	767	39%	1126	33%	874	46%	< .001
Attention impairment	185	5%	174	9%	225	7%	194	10%	< .001
Cognitive dysfunction	126	4%	110	6%	173	5%	132	7%	< .001

a Additional comorbidity variables in the long model

Modeling any SRB as a dichotomous outcome

Overall, 95% veterans had no SRB and 5% veterans had SRB (Table 2). In the unadjusted binomial logistic regression, ‘low impact-worsening’ pain, ‘moderate impact-worsening’ and ‘high impact-stable’ pain phenotypes were all associated with increased SRB compared to the ‘low impact-stable’ pain phenotype (Fig 1). After adjusting for demographics and prior SRB among pain phenotypes (IPSW estimates of the effects of pain phenotypes derived from the propensity scores predicted by the short set of covariates), the ‘low impact-worsening’ and ‘high impact-stable’ pain phenotypes remained significantly associated with increased risk of SRB. The magnitudes of the odds ratios associated with pain phenotypes were attenuated compared to those in the unadjusted model. After balancing diagnosed health conditions between pain phenotypes (IPSW estimates of the effects of pain phenotypes derived from the propensity scores predicted by the long set of covariates), pain phenotypes were no longer associated with SRB risk.

Table 2

Dichotomous and trichotomous analysis suicidal ideation and/or attempt by complex pain phenotype.

	Complex pain phenotype, No. (%)
	Low impact, stable		Low impact, worsening		Moderate impact, worsening		High impact, stable
Outcome by analysis	n = 3454 (33.8%)		n = 1955 (19.1%)		n = 3393 (33.2%)		n = 1915 (18.7%)
No suicidal ideation/attempt, n = 9751(95.3%)	3360	97.3%	1846	94.4%	3272	96.4%	1756	91.7%
Dichotomous analysis
Suicidal ideation/attempt, n = 483 (5%)	94	2.7%	109	5.6%	121	3.6%	159	8.3%
Trichotomous analysis
Suicidal ideation, n = 390 (3.8%)	74	2.1%	92	4.7%	99	2.9%	125	6.5%
Suicidal attempt, n = 90 (0.9%)	20	0.6%	17	0.9%	22	0.7%	34	1.8%

Fig 1

Effects of pain phenotypes on suicidal ideation or attempta.

a Forest plot of odds ratios and 95% confidence intervals associated with pain phenotypes derived from logistic regression analyses: unadjusted, adjusting away confounding associated with short set of covariates using IPSW, and adjusting away confounding associated with long set of covariates using IPSW. b The short set covariates includes sociodemographic characteristics (excludes comorbid health conditions) as predictors. c The long set covariates includes all covariates sociodemographic characteristics and comorbidities.

Modeling no SRB, suicidal ideation, and suicidal attempt as a trichotomous outcome

For the trichotomous outcome of no SRB, suicidal ideation only, and suicide attempt contained 9751, 390, and 93 veterans, respectively (Table 2).

In unadjusted multinomial logistic regression models, the ‘low impact-stable’ pain phenotype had significantly lower odds of suicidal ideation compared to other pain phenotypes (Fig 2). The ‘low impact-stable’ pain phenotype had significantly lower odds of suicide attempt-only compared to the ‘high impact-stable’ pain. After adjusting for demographic covariates and prior SRB among pain phenotypes (IPSW estimates of the effects of pain phenotypes derived from the propensity scores predicted by the short set of covariates), the ‘high impact-stable’ pain and ‘low impact-worsening’ phenotypes remained associated with increased risk for suicidal ideation but with a lesser magnitude compared to the unadjusted model. The ‘moderate impact-worsening’ pain phenotype was no longer associated with the risk of suicide ideation conditioned on balanced demographic covariates, while the effects of high impact and low impact worsening remained significant but with reduced magnitudes. Pain phenotype was no longer associated with increased risk of suicide attempt in either adjusted models (Fig 3).

Fig 2

Effects of pain phenotypes on suicidal ideation.

a Forest plot with odds ratio and 95% confidence intervals of the pain phenotypes with low impact-stable pain as reference from the multinomial logistic regression models. The figure shows the effect sizes with and without weights and using the long and short set of covariates for suicidal ideation. b The short set covariates includes sociodemographic characteristics (excludes comorbid health conditions) as predictors. c The long set covariates includes all covariates–sociodemographic characteristics and comorbidities.

Fig 3

Effects of pain phenotypes on suicide attempt.

a Forest plot with odds ratio and 95% confidence intervals of the pain phenotypes with low impact-stable pain as reference from the multinomial logistic regression models. The figure shows the effect sizes with and without weights and using the long and short set of covariates for suicide attempt. b The short set covariates includes sociodemographic characteristics (excludes comorbid health conditions) as predictors. c The long set covariates includes all covariates–sociodemographic characteristics and comorbidities.

Discussion

To understand the relatively high-risk of SRB in those with complex chronic pain, we examined the association of pain complexity, comorbidities, and suicide. Without adjustment, our previously identified pain phenotypes were significant predictors of SRB. However, this assessment did not account for the unobserved presence of comorbidities in this population. Therefore, we utilized propensity scores to balance the comorbidities present in the pain phenotypes, pain intensity and treatment trajectories and found they were no longer significantly associated with SRB. This indicates that comorbidities were a significant factor associating chronic pain with SRB. This further indicates that assessing, treating, and managing comorbid conditions in people living with pain is important to address risk for SRB.

Veterans in the high impact-stable phenotype were most likely to report suicidal ideation and/or attempt suicide. Individuals in this phenotype had the highest rates of PTSD, depression, SUD, and past suicide attempts. Veterans in the low impact-worsening phenotype were the next most likely to endorse SRB and had the second highest rates of PTSD, depression, SUD, and past suicide attempts. Our findings align with prior research on suicide risk in a military polytrauma sample which found that depression and PTSD were significantly associated with suicidal ideation and violent impulses in veterans with chronic pain [36]. Management of these comorbidities may have the most impact on suicide mitigation.

Veterans in the ‘high impact-stable’ and ‘low-impact worsening’ phenotypes had the highest rates of previous suicide attempts and SRB after phenotype development. This finding is consistent with the vast majority of studies that find the best predictor of suicide is prior SRB and that those prior SRB are also associated with comorbidities such as TBI, depression, SUD, and PTSD, all of which are common in Post-9/11 Veterans [35]. Similar to other studies [29, 45, 46], these data suggest that while pain is an important risk factor for suicidal ideation and attempts, psychological comorbidities play a larger role in the development and/or maintenance of suicidal ideation and attempts than pain alone [45]. While the management of chronic pain is paramount within a treatment program, this study highlights the need to add focus to potentially modifiable psychological comorbidities that can drive suicidal ideation and attempts.

The length of the suicidal process may vary among those with chronic pain and mTBI. Individuals with chronic pain alone may have a gradual progression towards suicidal behavior. Suicide may be deliberative and carefully planned, developed, ruminated, and then systematically carried out. In contrast, those with mTBI may have sudden suicidal behaviors which appear hastily decided-upon with little or no planning. Increased impulsivity associated with frontal lobe damages in those with mTBI may contribute to the sudden emergence of suicide [47]. Therefore, depending on whether mTBI is present, tailored interventions are warranted. With the presence of chronic suicidality, long term mental health and medical treatment are indicated. Impulsivity that leads to suicidal behaviors may seem more difficult to treat; however, due to increased utilization of healthcare by Veterans with mTBI there are ample opportunities to screen for suicide and intervene. Fortunately, there are numerous clinical interventions shown to significantly decrease suicidal behavior [48].

Both mTBI and pain are associated with mental health conditions and physical disabilities leading to difficulty in activities of daily living. These side effects of chronic pain and suicide-related behaviors are hidden disabilities among Veterans with mTBI. Given the high rates of suicide in this population, ongoing monitoring for psychological complications and distress are needed to prevent fatal suicide events. Veterans with mTBI and pain have increased healthcare encounters; thus, multiple opportunities are present to engage patients and caregivers in screening and prevention. Increased screening for suicide within this population is necessary and recommended.

Research related to suicide among patients living with a history of mTBI living with pain is limited. There is insufficient evidence to guide suicide management among those with mTBI and complex comorbid conditions. Among patients with mTBI, the traumatic brain injury is a long-term risk factor for suicide. The current VA/DoD clinical practice guidelines recommend a comprehensive treatment plan by addressing all physical conditions and mental health symptoms simultaneously [49].

Limitations

Known SRB events were limited to the ICD-codes documented in VA electronic medical records. These SRB data were limited compared to other SRB databases: the VA Suicide Prevention Applications Network (SPAN) [50] and the Joint Department of Defense (DoD)—Department of Veterans Affairs (VA) Suicide Data Repository (SDR) [51-53]. Thus our findings should be interpreted with cautions, including potential biases due to measurement errors (e.g., under detection of non-fatal and fatal suicide attempts, or timing of suicide events) or unmeasured confounding (e.g., factors that influenced both SRB and SRB predictors). Since we did not have access to the VA SPAN and the DoD/VA SDR databases, nor information regarding factors underlying the SRB reporting attrition, we are unable to assess the magnitudes nor directions of these potential biases. However, these biases could be limited as our results were consistent with known suicide predictors in the current literature. Nevertheless, future studies would benefit from combining data from SPAN, SDR and DoD/VA medical records and examining conditional probabilities of factors associated with the highest mortality.

While derived from a national sample, these findings cannot be extrapolated to all patients and should be used with caution due to the possibility of unmeasured confounding. For the purpose of inferring practical intervention to mitigate the sources underlying differential SRB risk associated with pain phenotypes, we chose clinical covariates in the propensity score model that could be modified during clinical care such as mental health conditions that can be treated.

Another limitation is related to the reporting and documentation of suicide attempts and ideations in the patients’ medical records. The numbers captured in this study may be lower than the actual number of suicide ideations and attempts that occur among Veterans with mTBI and pain due to reporting and documenting errors. Additionally, our numbers may be underrepresented because not all Veterans who experience SRB seek treatment at the VA or seek treatment at all due to stigma [54, 55]. It is possible that the rate of suicidal behavior is actually higher than what we found among individuals with TBI because those not seeking treatment may be at higher suicide risk.

Conclusion

These data expand on the existing knowledge about the impact of pain intensity, suicidal ideation and attempts, and complex patterns of comorbidities. These findings highlight the need for interdisciplinary care as the effect of the phenotypes disappeared when comorbidities were balanced. Assessing and treating both physical and mental health comorbidities are critical to improving outcomes and mitigating risk for suicide-related behavior.

ICD-9-CM diagnostic code definitions among Post-9/11 Veterans.

(DOCX)

Click here for additional data file.

Binomial logistic regression by complex pain phenotype for suicidal ideation or attempt.

(DOCX)

Click here for additional data file.

Multinomial logistic regression by complex pain phenotype for suicidal ideation.

(DOCX)

Click here for additional data file.

Multinomial logistic regression by complex pain phenotype for suicide attempt.

(DOCX)

Click here for additional data file.

Adjusted odds ratios and 95% confidence intervals associated with pain phenotypes derived from logistic regression analyses: unadjusted, adjusting away confounding associated with sociodemographic, military characteristics, mental health (minus prior suicide-related behavior) covariates using IPSW.

(DOCX)

Click here for additional data file.

Multinomial logistic regression by complex pain phenotype for suicidal ideation and attempt.

Adjusted odds ratios and 95% confidence intervals associated with pain phenotypes derived from logistic regression analyses: unadjusted, adjusting away confounding associated with sociodemographic, military characteristics, mental health (minus prior suicide-related behavior) covariates using IPSW.

(DOCX)

Click here for additional data file.

7 Mar 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Marco Innamorati

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf  and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

3. Please upload a copy of Figures 2 and 3, to which you refer in your text on page 9. If the figure is no longer to be included as part of the submission please remove all reference to it within the text.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This is an interesting and rare paper focusing on the association of pain and suicdal behaviour and specifically if pain phenotypes are independently associated with suicidal idation or attempts, or via comoribidites, in a longitudinal retrospective cohort study in 9/11 veterans. The objectives are both sicentifically and clinically relevant, methods well chosen and described, results are well presented, thoroughly discussed, limitations are exhaustively addressed and the conclusions are fully supported by the data.

As this is a very well designed study and the results are reported in a simialry excellently written paper I only have a few renarks:

1. Please make the title more espressive of the actual major findings of the study.

2. In the abstract methods section please describe pain phenotypes and also provide some iformation on the statistical analyses performed.

3. Given the reluctance of psychiatric patients to seek help in part due to stigma but also as a symptom of their psychiatric morbidity, and knowing that the risk of suicidal behaiovur is higher among untreated psychiatric patients (including PTSD patients), this should be mentioned as a limitation.

4. What was the reason for not including completed suicides as an outcome?

5. Why were alcohol or substance use disorders not included as covariates?

Reviewer #2: The authors reported original research results derived from their investigation on suicidal ideation and attempt among Post-9/11 Veterans with mild traumatic brain injury and complex pain. The article is overall well-written and of interest to the journal. However, it is unclear how authors assessed suicidal ideation and suicide attempts. Especially for suicidal ideation, the assessment might be more critical. Furthermore, I suggest providing some understanding of the mental pain occurring in the suicidal mind, discussing papers such as Critical appraisal of major depression with suicidal ideation. Ann Gen Psychiatry. 2019

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

17 Mar 2022

Reviewer #1: This is an interesting and rare paper focusing on the association of pain and suicidal behavior and specifically if pain phenotypes are independently associated with suicidal ideation or attempts, or via comorbidities, in a longitudinal retrospective cohort study in 9/11 veterans. The objectives are both scientifically and clinically relevant, methods well-chosen and described, results are well presented, thoroughly discussed, limitations are exhaustively addressed and the conclusions are fully supported by the data. As this is a very well designed study and the results are reported in a similarly excellently written paper I only have a few remarks:

1. Please make the title more expressive of the actual major findings of the study. We edited the title per reviewer #1’s recommendations.

2. In the abstract methods section please describe pain phenotypes and also provide some information on the statistical analyses performed. We updated the abstract per reviewer #1’s recommendations.

3. Given the reluctance of psychiatric patients to seek help in part due to stigma but also as a symptom of their psychiatric morbidity, and knowing that the risk of suicidal behavior is higher among untreated psychiatric patients (including PTSD patients), this should be mentioned as a limitation. We updated the limitations and added two references (#53 & 54) to address the reviewer’s comments, see the limitations section, paragraph three, lines 24-28.

4. What was the reason for not including completed suicides as an outcome? At the time the study was completed, we did not have access to data on completed suicides. Thus, our aim was to identify suicidal ideation and suicide attempts that were documented in the medical records.

5. Why were alcohol or substance use disorders not included as covariates? Substance use disorders were included as covariates, see table 1.

Reviewer #2: The authors reported original research results derived from their investigation on suicidal ideation and attempt among Post-9/11 Veterans with mild traumatic brain injury and complex pain. The article is overall well-written and of interest to the journal.

1. However, it is unclear how authors assessed suicidal ideation and suicide attempts. Especially for suicidal ideation, the assessment might be more critical. We used ICD9-CM codes used in prior studies of SRB to identify suicide ideation (V6284) and attempt (E950, E952, E953, E953, E954, E955, E956, E957, E958, E959) in national VA inpatient and outpatient data in years 6-8 of VA care.

2. Furthermore, I suggest providing some understanding of the mental pain occurring in the suicidal mind, discussing papers such as Critical appraisal of major depression with suicidal ideation. Ann Gen Psychiatry. 2019. We added a sentence to the introduction, paragraph three, lines 22-24 along with the suggested citation (#18).

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

18 Apr 2022

Complex pain phenotypes: suicidal ideation and attempt through latent multimorbidity

PONE-D-21-38828R1

Dear Dr. Song,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Marco Innamorati

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors have addressed all comments and recommendations. The paper can be accepted in its present version.

Reviewer #2: The authors addressed my comments and the article appears suitable for possible publication in the journa.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

20 Apr 2022

PONE-D-21-38828R1

Complex pain phenotypes: suicidal ideation and attempt through latent multimorbidity

Dear Dr. Song:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Marco Innamorati

Academic Editor

PLOS ONE