Introduction
Youth mental ill-health represents a global crisis (McGorry et al., Reference McGorry, Mei, Dalal, Alvarez-Jimenez, Blakemore, Browne, Dooley, Hickie, Jones, McDaid, Mihalopoulos, Wood, Azzouzi, Fazio, Gow, Hanjabam, Hayes, Morris, Pang and Killackey2024). For example, depression and anxiety disorders are the most common mental ill-health disorders (World Health Organization, 2022), with a peak onset occurring during adolescence and young adulthood (Kessler et al., Reference Kessler, Berglund, Demler, Jin, Merikangas and Walters2005). This has a significant impact on young people, including lower social and occupational functioning and reduced quality of life (Bowman, McKinstry, & McGorry, Reference Bowman, McKinstry and McGorry2017; Cotton et al., Reference Cotton, Hamilton, Filia, Menssink, Engel, Mihalopoulos, Rickwood, Hetrick, Parker, Herrman, Telford, Hickie, McGorry and Gao2022; Gibb, Fergusson, & Horwood, Reference Gibb, Fergusson and Horwood2010; Jaycox et al., Reference Jaycox, Stein, Paddock, Miles, Chandra, Meredith, Tanielian, Hickey and Burnam2009). While many young people recover from mental ill-health and healthcare services provide support to them and their families, treatment outcomes have remained modest (Patel et al., Reference Patel, Saxena, Lund, Kohrt, Kieling, Sunkel, Kola, Chang, Charlson, O’Neill and Herrman2023). Further understanding of the etiology of these disorders and barriers and facilitators to recovery is needed to effectively treat them (McGorry et al., Reference McGorry, Mei, Dalal, Alvarez-Jimenez, Blakemore, Browne, Dooley, Hickie, Jones, McDaid, Mihalopoulos, Wood, Azzouzi, Fazio, Gow, Hanjabam, Hayes, Morris, Pang and Killackey2024). The co-occurrence of mental and physical conditions, such as pain (Cotton et al., Reference Cotton, Hamilton, Filia, Menssink, Engel, Mihalopoulos, Rickwood, Hetrick, Parker, Herrman, Telford, Hickie, McGorry and Gao2022; Dudeney et al., Reference Dudeney, Aaron, Hathway, Bhattiprolu, Bisby, McGill, Gandy, Harte and Dear2024; Slater et al., Reference Slater, Waller, Briggs, Lord and Smith2025; Victor et al., Reference Victor, Hu, Campbell, White, Buse and Lipton2010), is often overlooked and may hold the key to improved treatment outcomes (Kroenke et al., Reference Kroenke, Shen, Oxman, Williams and Dietrich2008; Liu et al., Reference Liu, Huang, Bao, Lu, Dong, Wolkowitz, Kelsoe, Shi and Wei2024; Thielke, Fan, Sullivan, & Unützer, Reference Thielke, Fan, Sullivan and Unützer2007).
Pain is defined as an unpleasant sensory and emotional experience associated with, or resembling that associated with, actual or potential tissue damage (Raja et al., Reference Raja, Carr, Cohen, Finnerup, Flor, Gibson, Keefe, Mogil, Ringkamp, Sluka, Song, Stevens, Sullivan, Tutelman, Ushida and Vader2020). In young people, pain might be underdiagnosed or undertreated (Friedrichsdorf et al., Reference Friedrichsdorf, Postier, Eull, Weidner, Foster, Gilbert and Campbell2015; Hassett et al., Reference Hassett, Hilliard, Goesling, Clauw, Harte and Brummett2013), with headache, musculoskeletal, back, abdominal, pelvic, and multisite pain most common (Chambers et al., Reference Chambers, Dol, Tutelman, Langley, Parker, Cormier, Macfarlane, Jones, Chapman, Proudfoot, Grant and Marianayagam2024; Hirsch et al., Reference Hirsch, Dhillon-Smith, Cutner, Yap and Creighton2020). Pain can be classified into acute and chronic pain. Acute pain is considered an appropriate response to tissue trauma or inflammatory-related processes, with many recovering within expected tissue healing times (Cohen, Vase, & Hooten, Reference Cohen, Vase and Hooten2021). However, some individuals with acute pain go on to develop chronic pain, defined as pain lasting or recurring for three or more months (Treede et al., Reference Treede, Rief, Barke, Aziz, Bennett, Benoliel, Cohen, Evers, Finnerup, First, Giamberardino, Kaasa, Korwisi, Kosek, Lavand’homme, Nicholas, Perrot, Scholz, Schug and Wang2019). In the context of acute pain, the presence of mental ill-health predicts the transition to chronic pain in young people (Fisher, Monsell, Clinch, & Eccleston, Reference Fisher, Monsell, Clinch and Eccleston2024; Holley, Wilson, & Palermo, Reference Holley, Wilson and Palermo2017; Rabbitts et al., Reference Rabbitts, Palermo, Zhou, Meyyappan and Chen2020). One-in-five young people are estimated to live with chronic pain at any one time (Chambers et al., Reference Chambers, Dol, Tutelman, Langley, Parker, Cormier, Macfarlane, Jones, Chapman, Proudfoot, Grant and Marianayagam2024), which impairs daily functioning and quality of life (Huguet & Miró, Reference Huguet and Miró2008; Hunfeld et al., Reference Hunfeld, Perquin, Duivenvoorden, Hazebroek-Kampschreur, Passchier, Van Suijlekom-Smit and Van der Wouden2001) and is associated with an increased prevalence of mental ill-health (Dudeney et al., Reference Dudeney, Aaron, Hathway, Bhattiprolu, Bisby, McGill, Gandy, Harte and Dear2024).
In the context of youth mental health services, 45% of young people presenting for mental health treatment experience pain that negatively impacts overall quality of life (Cotton et al., Reference Cotton, Hamilton, Filia, Menssink, Engel, Mihalopoulos, Rickwood, Hetrick, Parker, Herrman, Telford, Hickie, McGorry and Gao2022), which could be any type of pain (e.g. acute versus chronic or any location). This is an important issue considering pain is associated with a reduced response to mental health treatment in adults (Kroenke et al., Reference Kroenke, Shen, Oxman, Williams and Dietrich2008; Liu et al., Reference Liu, Huang, Bao, Lu, Dong, Wolkowitz, Kelsoe, Shi and Wei2024; Thielke et al., Reference Thielke, Fan, Sullivan and Unützer2007). Both mental ill-health and pain have been associated with key clinical outcomes for young people including severity of depression and anxiety (Filia et al., Reference Filia, Rickwood, Menssink, Gao, Hetrick, Parker, Hamilton, Hickie, Herrman, Telford, Sharmin, McGorry and Cotton2021; Slater et al., Reference Slater, Waller, Briggs, Lord and Smith2025), suicidal thoughts and behaviors (Filia et al., Reference Filia, Rickwood, Menssink, Gao, Hetrick, Parker, Hamilton, Hickie, Herrman, Telford, Sharmin, McGorry and Cotton2021; Hinze et al., Reference Hinze, Crane, Ford, Buivydaite, Qiu and Gjelsvik2019, Reference Hinze, Ford, Crane, Haslbeck, Hawton, Gjelsvik, Allwood, Aukland, Casey, De Wilde, Farley, Fletcher, Kappelmann, Kuyken, Laws, Lord, Medlicott, Montero-Marin, Nuthall and Wainman2021; Hinze, Karl, Ford, & Gjelsvik, Reference Hinze, Karl, Ford and Gjelsvik2023; Moller et al., Reference Moller, Davey, Badcock, Wrobel, Cao, Murrihy, Sharmin and Cotton2022), substance (mis)use (Filia et al., Reference Filia, Rickwood, Menssink, Gao, Hetrick, Parker, Hamilton, Hickie, Herrman, Telford, Sharmin, McGorry and Cotton2021; Lambarth et al., Reference Lambarth, Katsoulis, Ju, Warwick, Takhar, Dale, Prieto-Merino, Morris, Sen, Wei and Sofat2023; McLaren et al., Reference McLaren, Kamper, Hodder, Wiggers, Wolfenden, Bowman, Campbell, Dray and Williams2017), and impaired social and occupational functioning (Filia et al., Reference Filia, Rickwood, Menssink, Gao, Hetrick, Parker, Hamilton, Hickie, Herrman, Telford, Sharmin, McGorry and Cotton2021; Iorfino et al., Reference Iorfino, Carpenter, Cross, Crouse, Davenport, Hermens, Yee, Nichles, Zmicerevska, Guastella, Scott and Hickie2022; Murray, Groenewald, de la Vega, & Palermo, Reference Murray, Groenewald, de la Vega and Palermo2020). Yet knowledge on the association between the specific characteristics of pain and how these impact young people in youth mental health settings above mental ill-health alone is limited.
Our first aim was to describe the specific characteristics of pain (frequency, intensity, and limitations) in young people attending five primary care youth mental health (headspace) centers in Australia for their first presentation of mental ill-health. Our second aim was to estimate the associations between pain characteristics and symptoms of depression and anxiety, suicidal ideation, social and occupational functioning, and substance use from baseline to 3-month follow-up.
Methods
Our study was a secondary analysis of a previously reported observational study with a 3-month follow-up (Filia et al., Reference Filia, Rickwood, Menssink, Gao, Hetrick, Parker, Hamilton, Hickie, Herrman, Telford, Sharmin, McGorry and Cotton2021). Ethical approval was granted by the University of Melbourne Human Research Ethics Committee (ID: 1645367.1). Written informed consent was obtained from every participant as well as from their parent/guardian if they were under 18 years old. Our study is reported in line with Strengthening the Reporting of Observational studies in Epidemiology guidelines (Supplementary Table 1) (von Elm et al., Reference von Elm, Altman, Egger, Pocock, Gøtzsche, Vandenbroucke and Initiative2008).
Setting
Participants were recruited at five headspace centers across Australia. headspace is the largest Australian non-profit youth mental health network (Rickwood et al., Reference Rickwood, Paraskakis, Quin, Hobbs, Ryall, Trethowan and McGorry2018). Recruitment occurred between September 2016 and April 2018, in three metropolitan and two regional centers in different states (Australian Capital Territory, Queensland, Victoria, Tasmania) to ensure representativeness.
Participants
All young people aged 12–25 years who attended headspace for a first appointment for mental health or substance use-related problems were eligible to participate.
Data sources and measurement
At baseline, an interview was conducted, and participants completed self-report questionnaires on a tablet device. Diagnoses were based on Diagnostic and Statistical Manual for Mental Health Disorders Fourth Edition (DSM-IV) criteria and were obtained from medical records (American Psychiatric Association, 2013). After 3 months, participants were contacted by the research team to complete the follow-up assessment on-site or through an online link. Pain characteristics and clinical outcomes were assessed at both baseline and 3-month follow-up.
Pain characteristics
Pain characteristics over the prior week were obtained from the Assessment of Quality of Life-Six dimensions (AQoL-6D) questionnaire (Allen et al., Reference Allen, Inder, Lewin, Attia and Kelly2013; Richardson et al., Reference Richardson, Peacock, Hawthorne, Iezzi, Elsworth and Day2012), which includes three questions on pain characteristics:
Serious pain frequency: ‘How often do you experience serious pain?’ Responses included very rarely (1), less than once a week (2), three to four times a week (3), most of the time (4).
Pain intensity: ‘How much pain or discomfort do you experience?’ Responses included none at all (1), I have moderate pain (2), I suffer from severe pain (3), I suffer unbearable pain (4).
Pain limitations: ‘How often does pain interfere with your usual activities?’ Responses included never (1), rarely (2), sometimes (3), often (4), always (5).
Outcomes
Symptoms of depression: Depressive symptoms were measured using the nine-item Patient Health Questionnaire (PHQ-9) (Kroenke, Spitzer, & Williams, Reference Kroenke, Spitzer and Williams2001), with total scores ranging from 0 to 27 and higher scores indicating more severe depressive symptoms.
Symptoms of anxiety: Anxiety symptoms were assessed using the seven-item Generalized Anxiety Disorder (GAD-7) scale (Spitzer, Kroenke, Williams, & Löwe, Reference Spitzer, Kroenke, Williams and Löwe2006), resulting in total scores between 0 and 21, with a higher score indicating greater symptoms of anxiety.
Suicidal ideation: For suicidal ideation, the 15-item Suicidal Ideation Questionnaire-Junior (SIQ-JR) was used (Reynolds, Reference Reynolds1987). Scores ranged between 0 and 90, with higher scores indicting greater suicidal ideation.
Social and occupational functioning: Social and occupational functioning was assessed using the Social and Occupational Functioning Assessment Scale (SOFAS) (Goldman, Skodol, & Lave, Reference Goldman, Skodol and Lave1992). A total score ranges from 0 to 100, with higher scores indicating greater functioning. The SOFAS is regularly used in headspace services to capture social, occupational, and school functioning in 12–25-year olds (Rickwood et al., Reference Rickwood, McEachran, Saw, Telford, Trethowan and McGorry2023). This is different from pain-related activity limitations, as it captures social and occupational broadly and not just specific to pain.
Substance use: Substance use risk scores (tobacco, alcohol, cannabis, cocaine, amphetamine, inhalants, sedatives, hallucinogens, and opioids) were obtained from the World Health Organization Alcohol, Smoking and Substance Involvement Screening Test (WHO-ASSIST v3.0) (Humeniuk et al., Reference Humeniuk, Ali, Babor, Farrell, Formigoni, Jittiwutikarn, De Lacerda, Ling, Marsden, Monteiro, Nhiwatiwa, Pal, Poznyak and Simon2008, Reference Humeniuk, Henry-Edwards, Ali, Poznyak and Monteiro2010). Scores range from 0 to 31 for tobacco and 0 to 39 for all other substances, with higher scores indicating greater substance use risk.
Confounders
The following variables were considered as confounders: age, sex assigned at birth (male, female), diagnosis (depression, anxiety, depression and anxiety, other), timepoint (baseline, follow-up), and the study center.
Bias
Based on previously published studies using this data (Cotton et al., Reference Cotton, Hamilton, Filia, Menssink, Engel, Mihalopoulos, Rickwood, Hetrick, Parker, Herrman, Telford, Hickie, McGorry and Gao2022; Filia et al., Reference Filia, Rickwood, Menssink, Gao, Hetrick, Parker, Hamilton, Hickie, Herrman, Telford, Sharmin, McGorry and Cotton2021), attrition and missing data bias were anticipated. We explored differences in baseline demographics, pain characteristics, and outcomes between participants who completed both baseline and follow-up and those who only completed baseline. In addition, restricted maximum likelihood estimations (Brauer & Curtin, Reference Brauer and Curtin2018) and multiple imputation (Mayer, Reference Mayer2024) were used during analysis (see statistical methods section).
Study size
The sample size of the current analysis was based on the available sample of the original study (n = 1,107 at baseline, n = 665 at follow-up) (Filia et al., Reference Filia, Rickwood, Menssink, Gao, Hetrick, Parker, Hamilton, Hickie, Herrman, Telford, Sharmin, McGorry and Cotton2021).
Quantitative variables
All outcomes and pain characteristics were treated as continuous in analyses. Age was also treated as continuous. Covariates of sex assigned at birth, diagnosis, time, and study center were treated as categorical.
Given there were only two timepoints, we baseline-centered pain characteristics prior to multi-level modelling to evaluate whether (1) the baseline pain score (level 2 exposure; between-participant effect) and/or (2) change from the baseline score (level 1 exposure; within-participant effect) was associated with the outcome over time. Interpretation of the baseline-centered coefficients represent (1) if participants with higher baseline pain values had worse clinical outcomes over time relative to participants with lower baseline pain values (level 2 exposure; between-participant effect) and/or (2) if change from baseline pain score was associated with change in clinical outcome within each participant (level 1 exposure; within-participant effect).
Statistical methods
All statistical analyses were conducted in R (version 4.4.1) (The Comprehensive R Archive Network, 2024). Simple tests explored differences in baseline demographics, pain characteristics, and outcomes between participants who did and did not complete follow-up assessments (Wilcoxon rank sum test; Pearson’s Chi-squared test; Fisher’s exact test). Multicollinearity for pain characteristics was checked with correlation analyses.
Multiple imputation through chained random forests was conducted prior to analysis using ‘missRanger v2.6.0’ (Mayer, Reference Mayer2024). For imputation, we specified a wide data set including demographic variables (listed in Table 1), all items (including total and standardized scores) on the AQoL6D questionnaire (Allen et al., Reference Allen, Inder, Lewin, Attia and Kelly2013; Richardson et al., Reference Richardson, Peacock, Hawthorne, Iezzi, Elsworth and Day2012), and all outcomes relevant to our analyses across both timepoints. We wide-imputed 20 datasets with a predicted mean matching score of three and converted these back to long format before subsequent analyses.
Table 1. Baseline demographic, pain, and outcome characteristics of the sample
Abbreviations: PHQ-9 = Nine item Patient Health Questionnaire, GAD-7 = Seven item Generalized Anxiety Disorder Scale, SIQ-JR = Suicidal Ideation Questionnaire-Junior, SOFAS = Social and Occupational Functioning Assessment Scale, ASSIST = World Health Organization Alcohol, Smoking, and Substance Involvement Screening Test.
a Mean (SD); n (%).
b Wilcoxon rank sum test; Pearson’s Chi-squared test; Fisher’s exact test. Indicating the differences between participants with baseline and follow-up data compared to those with baseline data only.
Linear mixed effects models with restricted maximum likelihood estimations were run across both unimputed and imputed data sets using the ‘lme4 v1.1–35.5’ package (Bates et al., Reference Bates, Maechler, Bolker, Walker, Singmann, Dai, Scheipl, Grothendieck, Green, Fox, Bauer, Krivitsky, Tanaka and Jagan2024). Two sets of models were used for each outcome (depressive symptoms, anxiety symptoms, suicidal ideation, functioning, and substance use) and data set (unimputed and imputed), including:
Model 1 (unadjusted for confounders): Pain characteristics (frequency, intensity, and limitations) were included as separate characteristics (single pain variable model). All pain characteristics were included together as exposures in a combined model (multi-pain variable model).
Model 2 (adjusted for confounders): Analyses were repeated with fixed confounders of age, sex assigned at birth, primary diagnosis, and timepoint.
All models included a random intercept for participants clustered within the five study centers (three-level model). The alpha level for p-values was set at <0.05, with p-value adjustment for the false discovery rate (FDR) (Benjamini & Hochberg, Reference Benjamini and Hochberg1995), given multiple outcomes and exposures. We reported the estimates of adjusted pain characteristics from imputed data sets that were significant across both unimputed and imputed analyses after FDR adjustment of the p-values as the most robust findings.
Results
Descriptive data
The flow of participant selection has been previously reported (Filia et al., Reference Filia, Rickwood, Menssink, Gao, Hetrick, Parker, Hamilton, Hickie, Herrman, Telford, Sharmin, McGorry and Cotton2021). Baseline demographic data are presented in Table 1. The average follow-up time was 13 weeks (SD = 1.8). Comparisons between participants who completed follow-up assessments and those who did not showed differences only in study center location and education status (Table 1). The mean (standard deviation) age of the sample was 18.1 (3.3), with 500 (45%) participants aged between 12 and 17 years and 607 (55%) aged between 18 and 25 years.
Pain characteristics
Pain characteristics at baseline are reported in Table 1 and follow-up in Supplementary Table 2. At baseline, 177 (16%) participants experienced serious pain more than 3 days, 51% (548) reported at least moderate pain, and 257 (25%) experienced activity limitations due to pain in the last week. Of the 346 participants with baseline and follow-up data experiencing moderate or higher pain intensity at baseline, 231 (70%) reported moderate or higher pain intensity at follow-up.
Outcome data
Outcome data at baseline are reported in Table 1 and follow-up in Supplementary Table 2. For substance use, only tobacco, alcohol, and cannabis risk scores were analyzed, given the low proportions of participants with non-zero risk scores for cocaine, amphetamine, inhalants, sedatives, hallucinogens, and opioids (Supplementary Table 3).
Main results
Results of correlation analyses for pain characteristics are reported in Supplementary Table 4. The correlation (r) was 0.68 between serious pain frequency and pain intensity, 0.73 between pain intensity and pain limitations, and 0.67 between serious pain frequency and pain limitations. Results of single- and multi-pain variable models are in Supplementary Tables 5 and 6, respectively, and in Figure 1 (key estimates) and Supplementary Figures 1–3 (all estimates). Full model outputs are reported in Supplementary Tables 7–34. All estimates are reported as beta coefficients and 95% confidence intervals (β[95%CI]).
Figure 1. Forest plot of beta coefficients and 95% confidence intervals of pain characteristics from adjusted multi-pain variable linear mixed effects models with restricted maximum likelihood estimation. Here, we present the significant results following false discovery rate (FDR) adjustment in both datasets with and without imputation for ease of figure interpretation. Figures containing all estimates are available in Supplementary Figures 1–3. Between-participant estimates are the baseline score, indicating if baseline pain was associated with clinical outcomes across the three-month follow-up (level 2 exposure). Within‑participant estimates are baseline-centered indicating if a change from the baseline pain score was associated with a change in clinical outcome over time (level 1 exposure). Functioning refers to social and occupational functioning.
Depressive symptoms
Between-participant effects: Single pain variable models showed that higher baseline serious pain frequency (β[95%CI]: 2.21 [1.86, 2.57]; FDR-p < 0.001), pain intensity (β[95%CI]: 3.28 [2.79, 3.76]; FDR-p < 0.001), and pain limitations (β[95%CI]: 2.20 [1.88, 2.52]; FDR-p < 0.001) were associated with greater symptoms of depression over time. In multi-pain variable models, only higher pain intensity (β[95%CI]: 1.50 [0.71, 2.28]; FDR-p = 0.001) and pain limitations (β[95%CI]: 1.13 [0.63, 1.63]; FDR-p < 0.001) were associated with greater symptoms of depression.
Within-participant effects: Single pain variable models showed that increases in serious pain frequency (β[95%CI]: 1.30 [0.89, 2.57]; FDR-p < 0.001), pain intensity (β[95%CI]: 1.86 [1.33, 2.40]; FDR-p < 0.001), and pain limitations (β[95%CI]: 1.47 [1.10, 1.83]; FDR-p < 0.001) were associated with increases in symptoms of depression. Only increases in pain limitations (β[95%CI]: 0.99 [0.54, 1.43]; FDR-p < 0.001) were associated with increases in symptoms of depression in multi-pain variable models.
Anxiety symptoms
Between-participant effects: Higher serious pain frequency (β[95%CI]: 1.91 [1.61, 2.21]; FDR-p < 0.001), pain intensity (β[95%CI]: 2.60 [2.19, 3.00]; FDR-p < 0.001), and pain limitations (β[95%CI]: 1.62 [1.35, 1.89]; FDR-p < 0.001) were associated with greater symptoms of anxiety over time in single pain variable models. Higher serious pain frequency (β[95%CI]: 0.90 [0.45, 1.35]; FDR-p = 0.001) and pain intensity (β[95%CI]: 1.22 [0.56, 1.89]; FDR-p = 0.002) were associated with greater anxiety symptoms in both multi-pain variable models.
Within-participant effects: Single pain variable models showed that increases in serious pain frequency (β[95%CI]: 1.25 [0.90, 1.61]; FDR-p < 0.001), pain intensity (β[95%CI]: 1.58 [1.12, 2.05]; FDR-p < 0.001), and pain limitations (β[95%CI]: 1.23 [0.91, 1.55]; FDR-p < 0.001) were associated with increases in symptoms of anxiety. No associations were observed in multi-pain variable models.
Suicidal ideation
Between-participant effects: Single pain variable models indicated that greater serious pain frequency (β[95%CI]: 5.58 [4.45, 6.71]; FDR-p < 0.001), pain intensity (β[95%CI]: 8.05 [6.52, 9.58]; FDR-p < 0.001), and pain limitations (β[95%CI]: 5.39 [4.38, 6.39]; FDR-p < 0.001) were associated with suicidal ideation over time. Only higher pain intensity (β[95%CI]: 3.47 [0.98, 5.96]; FDR-p = 0.020) was associated with greater suicidal ideation in multi-pain variable models.
Within-participant effects: Increases in serious pain frequency (β[95%CI]: 2.71 [1.56, 3.85]; FDR-p < 0.001), pain intensity (β[95%CI]: 4.27 [2.76, 5.78]; FDR-p < 0.001), and pain limitations (β[95%CI]: 3.08 [2.04, 4.12]; FDR-p < 0.001) were associated with increases in suicidal ideation in single pain variable models. No within-participant effects were observed in multi-pain variable models.
Social and occupational functioning
Between-participant effects: Single pain variable models showed that higher serious pain frequency (β[95%CI]: ‑1.60 [−2.17, −1.02]; FDR-p < 0.001), pain intensity (β[95%CI]: ‑2.64 [−3.42, −1.87]; FDR-p < 0.001), and pain limitations (β[95%CI]: −1.77 [−2.27, −1.26]; FDR-p < 0.001) were associated with lower social and occupational functioning over time. No associations between pain characteristics and social and occupational functioning were observed in multi‑pain variable models.
Within-participant effects: Results from single pain variable models showed that increases in serious pain frequency (β[95%CI]: ‑1.19 [−1.82, −0.55]; FDR-p < 0.001), pain intensity (β[95%CI]: ‑1.81 [−2.65, −0.97]; FDR-p < 0.001), and pain limitations (β[95%CI]: −1.49 [−2.06, −0.91]; FDR-p < 0.001) were associated with decreases in social and occupational functioning. In multi-pain variable models, only increases in pain limitations (β[95%CI]: −1.08 [−1.78, −0.38]; FDR-p = 0.009) were associated with decreases in social and occupational functioning.
Tobacco use risk scores
Between-participant effects: Results from single pain variable models showed that greater serious pain frequency (β[95%CI]: 1.23 [0.75, 1.71]; FDR-p < 0.001), pain intensity (β[95%CI]: 1.51 [0.85, 2.16]; FDR-p < 0.001), and pain limitations (β[95%CI]: 1.06 [0.63, 1.48]; FDR-p < 0.001) were associated with higher tobacco use risk scores over time. No associations were observed with tobacco use risk scores in multi-pain variable models.
Within-participant effects: Only increases in pain intensity (β[95%CI]: 0.96 [0.45, 1.46]; FDR-p = 0.001) was associated with increases in tobacco use risk scores in single pain variable models. Increases in pain intensity (β[95%CI]: 1.09 [0.48, 1.70]; FDR-p = 0.002) were also associated with increases in tobacco use risk scores in multi-pain variable models.
Alcohol use risk scores
No effects of pain characteristics in both single- and multi-pain variable models were observed for both between- and within-effects for alcohol use risk.
Cannabis use risk scores
Between-participant effects: Only higher serious pain frequency (β[95%CI]: 1.26 [0.77, 1.75]; FDR-p < 0.001) was associated with greater cannabis use risk scores across single pain variable models over time. No effects of pain characteristics were observed on cannabis use risk scores in multi-pain variable models.
Within-participant effects: No effects of pain characteristics in both single- and multi-pain variable models were observed for cannabis use risk scores.
Discussion
Mental ill-health and pain conditions can co-occur in young people (Cotton et al., Reference Cotton, Hamilton, Filia, Menssink, Engel, Mihalopoulos, Rickwood, Hetrick, Parker, Herrman, Telford, Hickie, McGorry and Gao2022), yet the characteristics of this pain and the impact on clinical outcomes in the early phases of treatment have remained unexplored. Our findings showed that one-in-six young people accessing early intervention-focused mental healthcare experienced serious pain more than 3 days, one-in-two reported at least moderate pain, and one-in-four reported limitations to their usual activities due to pain in the last week. Furthermore, 70% of participants who reported baseline pain also reported pain at follow-up and may represent a group experiencing chronic pain (Treede et al., Reference Treede, Rief, Barke, Aziz, Bennett, Benoliel, Cohen, Evers, Finnerup, First, Giamberardino, Kaasa, Korwisi, Kosek, Lavand’homme, Nicholas, Perrot, Scholz, Schug and Wang2019). Findings showed that those with higher baseline pain values had worse clinical outcomes over time compared to those with lower baseline pain values. Specifically, higher serious pain frequency was associated with greater symptoms of anxiety, higher pain intensity was associated with greater symptoms of depression, anxiety, and suicidal ideation, and higher pain limitations were associated with greater depressive symptoms. Furthermore, an increase in pain intensity associated with an increase in tobacco use risk scores, while increases in pain limitations were associated with increases in depressive symptoms and decreases social and occupational functioning. Our findings showed the substantial negative impact that specific pain characteristics have on young people with mental ill-health and indicate that youth mental health services should screen for pain at intake. Improved and timely assessment must be complemented with collaborative models of pain care in mental health settings (Patel et al., Reference Patel, Saxena, Lund, Kohrt, Kieling, Sunkel, Kola, Chang, Charlson, O’Neill and Herrman2023), to improve clinical outcomes for young people with mental ill-health.
Our results showed that 51% of young people care seeking for mental ill-health report pain. In contrast, one study showed a chronic pain prevalence of ~70% in adolescents with psychiatric disorders (Mangerud, Bjerkeset, Lydersen, & Indredavik, Reference Mangerud, Bjerkeset, Lydersen and Indredavik2013); however, this was in a tertiary care sample. Participants may have had more severe presentations compared to our sample, which was recruited from primary care settings. Another difference is that our study did not examine chronic pain specifically. In adults with mental ill-health, studies have reported a prevalence of pain (combined acute and chronic) ranging from 42% to 65% (Bair, Robinson, Katon, & Kroenke, Reference Bair, Robinson, Katon and Kroenke2003; Kroenke et al., Reference Kroenke, Shen, Oxman, Williams and Dietrich2008). The reasons for this co-occurrence remain poorly understood; however, neurobiological (brain alterations [Bair et al., Reference Bair, Robinson, Katon and Kroenke2003; Hooten, Reference Hooten2016] and neuroimmune [Campos et al., Reference Campos, Antunes, Matsumoto, Pagano and Martinez2020; Walker et al., Reference Walker, Kavelaars, Heijnen and Dantzer2014]), and behavioral, psychological, and environmental (Khan, Michelini, & Battaglia, Reference Khan, Michelini and Battaglia2020) mechanisms have been proposed. For example, pain conditions are associated with areas of brain processing related to emotion (Martucci & Mackey, Reference Martucci and Mackey2018), which overlap with those linked to mental ill-health in young people (Soltani, Kopala-Sibley, & Noel, Reference Soltani, Kopala-Sibley and Noel2019). Furthermore, persistent neuroinflammation has been observed to influence the onset and persistence of both mental ill-health and pain (Campos et al., Reference Campos, Antunes, Matsumoto, Pagano and Martinez2020). Finally, various behavioral (e.g. sleep quality), psychological (e.g. affect, rumination), and environmental factors (e.g. particularly parental and familial factors in young people) are known to influence both mental ill‑health and pain (Soltani et al., Reference Soltani, Kopala-Sibley and Noel2019). Our findings showed that the prevalence of pain in young people with mental ill‑health accentuates the urgent need to understand and treat this co-occurrence as early as possible to prevent it from continuing into adulthood, where there is the potential for continued functional and social limitations (De La Rosa et al., Reference De La Rosa, Brady, Ibrahim, Herder, Wallace, Padilla and Vanderah2024). It emphasizes the need for researchers to better understand shared mechanisms to optimize treatment decisions.
Higher baseline pain intensity and limitations were found to be associated with greater depressive symptoms over time. These findings are similar to those of a population-based study on Australian young people (Kamper et al., Reference Kamper, Michaleff, Campbell, Dunn, Yamato, Hodder, Wiggers and Williams2019), which found that higher pain frequency was associated with poor mental health. Other studies in young people experiencing pain have shown comparable results (Hu et al., Reference Hu, Liu, Wang, Jia and Liu2022; Zvolensky et al., Reference Zvolensky, Kauffman, Shepherd, Viana, Bogiaizian, Rogers, Bakhshaie and Peraza2020). Yet here we have demonstrated the importance of co-occurring pain in young people with mental ill-health. Depressive symptoms could be worsened by pain due to factors such as less sleep, stress, lower self-efficacy, hopelessness, lower functioning, or limited social participation and positive reinforcement; however, this relationship is likely bidirectional (Hazeldine-Baker, Salkovskis, Osborn, & Gauntlett-Gilbert, Reference Hazeldine-Baker, Salkovskis, Osborn and Gauntlett-Gilbert2018; Thompson et al., Reference Thompson, Broadbent, Fuller-Tyszkiewicz, Bertino and Staiger2019; Wise et al., Reference Wise, Meyers, Dessaiah, Mallinckrodt, Robinson and Kajdasz2008). Pain intensity and limitations contribute to these factors, potentially explaining the association between these specific characteristics and depressive symptoms in the current study. A recent systematic review (Liu et al., Reference Liu, Huang, Bao, Lu, Dong, Wolkowitz, Kelsoe, Shi and Wei2024) showed effect sizes for treatments of depressive symptoms in adults with depression are smaller in those experiencing pain compared to those without pain. Therefore, screening for the possible presence of pain and optimizing pain management in patients with depressive symptoms could improve treatment efficacy.
Greater pain intensity and serious pain frequency at baseline were associated with higher symptoms of anxiety. Pain has been linked to symptoms of anxiety in other studies (Blaauw et al., Reference Blaauw, Dyb, Hagen, Holmen, Linde, Wentzel-Larsen and Zwart2014; Hommer, Lateef, He, & Merikangas, Reference Hommer, Lateef, He and Merikangas2022; Kamper et al., Reference Kamper, Michaleff, Campbell, Dunn, Yamato, Hodder, Wiggers and Williams2019), but our study is the first to examine this relationship in young people experiencing mental ill-health. The fear-avoidance model (Lethem, Slade, Troup, & Bentley, Reference Lethem, Slade, Troup and Bentley1983) may be used to explain the relationship between pain and anxiety suggesting that these conditions not only often co-occur but also exacerbate each other (Hooten, Reference Hooten2016; Vlaeyen & Linton, Reference Vlaeyen and Linton2000). When pain is perceived as catastrophic, it can lead to intense fear and avoidance of physical activity, ultimately resulting in more pain (Vlaeyen & Linton, Reference Vlaeyen and Linton2000). Strategies to manage increased pain episodes should be used to support anxiety in young people with mental ill-health.
Higher baseline pain intensity and limitations were associated with higher levels of suicidal ideation in our study. These associations could be explained by disability, hopelessness, or a desire to escape from pain (Fishbain et al., Reference Fishbain, Bruns, Meyer, Lewis, Gao and Disorbio2012; Tang & Crane, Reference Tang and Crane2006). This is in line with prior research in both mental ill-health and chronic pain populations (Hinze et al., Reference Hinze, Crane, Ford, Buivydaite, Qiu and Gjelsvik2019, Reference Hinze, Karl, Ford and Gjelsvik2023; Kowal, Wilson, Henderson, & McWilliams, Reference Kowal, Wilson, Henderson and McWilliams2014; Wildeboer, Chambers, Soltani, & Noel, Reference Wildeboer, Chambers, Soltani and Noel2023). For co-occurring mental ill-health and pain, one prior study showed that higher levels of depressive symptoms were associated with suicidality onset in adolescents with chronic pain (Wildeboer et al., Reference Wildeboer, Chambers, Soltani and Noel2023). A systematic review also showed that depression moderated suicidal ideation in young people experiencing pain; however, it did not fully explain the pain-suicidality relationship (Hinze et al., Reference Hinze, Crane, Ford, Buivydaite, Qiu and Gjelsvik2019). These together support the findings of our study and may indicate that, regardless of service (e.g. mental health or pain), clinicians should consider the impact of co-occurring mental ill-health and pain symptoms on suicidal ideation.
Increases in the experience of pain-related limitations were associated with decreases in social and occupational functioning. This could indicate that young people are able to develop coping mechanisms that mitigate the impact of their pain in social and occupational settings. However, if the young person deviates from their usual level of pain limitations, this has the potential to negatively influence their social and occupational functioning. Previous studies have shown that children and adolescents with pain experience more restrictions in daily functioning, such as in school, hobbies, and social activities. (Cohen, Vowles, & Eccleston, Reference Cohen, Vowles and Eccleston2010; Kaczynski, Claar, & LeBel, Reference Kaczynski, Claar and LeBel2013; Logan, Simons, Stein, & Chastain, Reference Logan, Simons, Stein and Chastain2008; Roth-Isigkeit et al., Reference Roth-Isigkeit, Thyen, Stöven, Schwarzenberger and Schmucker2005). This is further supported by other studies in young people with chronic pain (Bateman et al., Reference Bateman, Caes, Eccleston, Noel and Jordan2023; Serbic, Friedrich, & Murray, Reference Serbic, Friedrich and Murray2023; van Alboom et al., Reference van Alboom, Elmer, Boersma, Forgeron, Baert, Bracke and Goubert2022) and mental ill-health (Filia et al., Reference Filia, Rickwood, Menssink, Gao, Hetrick, Parker, Hamilton, Hickie, Herrman, Telford, Sharmin, McGorry and Cotton2021; Iorfino et al., Reference Iorfino, Carpenter, Cross, Crouse, Davenport, Hermens, Yee, Nichles, Zmicerevska, Guastella, Scott and Hickie2022). Our results show the distinct impact of pain on social and occupational functioning, over and above the burden of mental ill-health alone. These results suggest that clinicians should monitor increases in activity limitations due to pain, as this may negatively affect social and occupational functioning in young people with mental health conditions.
Our study showed that increases in pain intensity were associated with increases in tobacco use risk. This indicates that an increase in pain intensity may result in an increased risk of problematic tobacco use. Studies have found pain intensity to be associated with smoking in both adolescents (Kamper et al., Reference Kamper, Michaleff, Campbell, Dunn, Yamato, Hodder, Wiggers and Williams2019) and adults (Barry, Pilver, Hoff, & Potenza, Reference Barry, Pilver, Hoff and Potenza2013; Ditre, Brandon, Zale, & Meagher, Reference Ditre, Brandon, Zale and Meagher2011). This association could be explained by the analgesic effect of nicotine (Kishioka, Kiguchi, Kobayashi, & Saika, Reference Kishioka, Kiguchi, Kobayashi and Saika2014). However, over the longer term, smoking and pain can exacerbate each other through a positive feedback loop (Ditre & Brandon, Reference Ditre and Brandon2008). No significant associations were found for alcohol and cannabis use. This emphasizes the importance of screening for fluctuations in pain intensity and providing support to quit smoking to prevent longer-term negative changes, given the association between pain intensity and tobacco use.
The current study has multiple strengths. A large longitudinal dataset was used; participants represented a broad age range and were from diverse locations across Australia, including measures which are commonly used and well-validated (Allen et al., Reference Allen, Inder, Lewin, Attia and Kelly2013; Goldman et al., Reference Goldman, Skodol and Lave1992; Humeniuk et al., Reference Humeniuk, Ali, Babor, Farrell, Formigoni, Jittiwutikarn, De Lacerda, Ling, Marsden, Monteiro, Nhiwatiwa, Pal, Poznyak and Simon2008, Reference Humeniuk, Henry-Edwards, Ali, Poznyak and Monteiro2010; Kroenke et al., Reference Kroenke, Spitzer and Williams2001; Reynolds, Reference Reynolds1987; Richardson et al., Reference Richardson, Peacock, Hawthorne, Iezzi, Elsworth and Day2012; Spitzer et al., Reference Spitzer, Kroenke, Williams and Löwe2006). Participants had a first presentation of mental ill-health, reflecting the prevalence and influence of pain characteristics on clinical outcomes in youth with mental ill-health during the early treatment stages (first 3 months). We explored both between- and within-participant effects in multi-level mixed models and the potential impact of attrition on effect estimates through comparison of demographic, pain characteristics, and outcome data, restricted maximum likelihood estimations, and multiple imputation. Furthermore, we adjusted p-values for the FDR to account for the testing of multiple exposures and outcomes to reduce the potential of false positive results (Benjamini & Hochberg, Reference Benjamini and Hochberg1995).
In terms of limitations, only 62% of recruited participants completed both baseline and follow-up assessments. There was no available information on the location or type of pain, underlying medical conditions that could indicate secondary pain conditions (rather than primary), duration of pain (acute or chronic), or if pain was assessed in standard intake procedures. This could have been insightful, since multisite and/or chronic pain is likely to have a larger influence on clinical outcomes (Mangerud et al., Reference Mangerud, Bjerkeset, Lydersen and Indredavik2013) and should be considered in future research. Given that our study only had two timepoints, future research could validate these findings in more intensive longitudinal designs to explore how pain and mental ill-health co-fluctuate and to further understand their individual and combined effects on treatment engagement and outcomes.
Our study emphasizes the need for clinicians and researchers to consider the co-occurrence between mental ill-health and pain. It would be prudent for clinicians in youth mental health settings to screen for pain at intake, work with the young person to determine real-world impacts of co-occurring pain on their mental health and functioning, and provide advice or referrals to integrated treatment as appropriate. For researchers, there is a knowledge gap regarding youth experiences of co-occurring mental ill-health and pain (Soltani et al., Reference Soltani, Kopala-Sibley and Noel2019); there have been no trials that have recruited young people experiencing both mental ill-health and pain and provided targeted, integrated, and accessible treatments to this population (Ma et al., Reference Ma, Romano, Ashworth, Smith, Vancampfort, Scott, Gaughran, Stewart and Stubbs2024). The impact of co-occurring mental ill-health and pain can continue into adulthood in a reinforcing negative cycle that can then lead to occupational and relationship problems (De La Rosa et al., Reference De La Rosa, Brady, Ibrahim, Herder, Wallace, Padilla and Vanderah2024). Similar to mental health conditions (Solmi et al., Reference Solmi, Radua, Olivola, Croce, Soardo, de Pablo, Il Shin, Kirkbride, Jones, Kim, Kim, Carvalho, Seeman, Correll and Fusar-Poli2022), pain develops in early adolescence (Chambers et al., Reference Chambers, Dol, Tutelman, Langley, Parker, Cormier, Macfarlane, Jones, Chapman, Proudfoot, Grant and Marianayagam2024) and can lead to mental ill-health (Bondesson, Bolmsjö, Pardo, & Jöud, Reference Bondesson, Bolmsjö, Pardo and Jöud2024). There is an urgent need to recognize the emergence of pain in adolescence and promote efforts for early intervention and prevention to reduce the burden of pain and its mental health impacts, as well as evaluate this in longer-term follow-up studies. Overall, these results indicate there is an urgent need for researchers to answer key gaps in our understanding and treatments for young people with co-occurring mental ill-health and pain.
Conclusion
We explored the prevalence and impact of pain in young people with mental ill-health accessing early intervention primary mental healthcare services. Results showed that one-in-two young people with mental ill-health also report pain at intake. Serious pain frequency, intensity, and limitations were found to have a negative impact on symptoms of depression, anxiety, suicidal ideation, social and occupational functioning, and tobacco use for young people in the early treatment stages of mental ill-health. These results highlight the need for early pain recognition in mental health settings to support the one-in-two young people with mental ill-health who report pain. Developing more integrated and collaborative care strategies is paramount to reduce the major burden of this co-occurrence among young people and mitigate potential negative longer-term outcomes into adulthood.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/S0033291725101104.
Data availability statement
Due to ethical considerations for this study, the code and dataset used in the current study are available from the corresponding author on reasonable request.
Acknowledgments
We thank the project staff for coordinating the project and team of research assistants involved in collecting the data. We thank the headspace center staff and young people whose participation made this study possible.
Author contribution
Conceptualization: JM, LS, SMC, SDT. Data curation: VAO, CG, SDT. Formal Analysis: VAO, CG, SDT. Funding acquisition: PM, SC, DR, SH, AP, IH, HH. Investigation: KF, JM, AW. Methodology: All. Project administration: KF, JM, AW. Resources: NA. Software: SDT. Supervision: JN, LS, SDT. Validation: JN. Visualization: VAO, CG, SDT. Writing – original draft: VAO, SDT. Writing – review & editing: All. Approved final manuscript: All.
Funding statement
The study was supported by a National Health and Medical Research Council (NHMRC) Partnership Grant (APP1076940). This was a joint project between Orygen, The University of Melbourne and headspace National Youth Mental Health Foundation. PDM is supported by a National Health and Medical Research Council (NHMRC) Senior Principal Research Fellowship (1155508). IBH is supported by a NHMRC Research Fellowship (1136259), an Investigator Grant Leadership 3 (2016346), and a Centre for Research Excellence grant (1171910 and 1061043). LS is supported by an NHMRC Investigator Grant (2017962), University of Melbourne Dame Kate Campbell fellowship and NIH RO1 MH129742, RO1 MH129832 and RO1 MH117601 grants. SDT was supported by RO1 MH129832 and RO1 MH117601 grants and is currently supported by a University of Melbourne Sir Randal Heymanson Fellowship. The funding sources had no role in the writing or publication of the manuscript.
Competing interests
IBH is the Co-Director, Health and Policy at the Brain and Mind Centre (BMC) University of Sydney, Australia. The BMC operates early-intervention youth services at Camperdown under contract to headspace. IBH has previously led community-based and pharmaceutical industry‑supported (Wyeth, Eli Lily, Servier, Pfizer, AstraZeneca, Janssen Cilag) projects focused on the identification and better management of anxiety and depression. IBH is the Chief Scientific Advisor to, and a 3.2% equity shareholder in, InnoWell Pty Ltd which aims to transform mental health services using innovative technologies. PDM is a founding director, patron, and former founding board member of headspace. PDM is the executive director of Orygen, Australia’s National Centre of Excellence in Youth Mental Health and lead agency for five headspace centers across northwest Melbourne. PDM is a past President of the International Association for Youth Mental Health, and a past President of the IEPA; Early Intervention in Mental Health and of the Schizophrenia International Research Society. All other authors declare no conflicts of interest.
Open access
