Hostname: page-component-5f7774ffb-pcg8z Total loading time: 0 Render date: 2026-02-23T12:25:39.039Z Has data issue: false hasContentIssue false
  • English
  • Français

Evaluating context effects on PHQ-8 somatic item scores among people with a chronic medical condition: a scleroderma patient-centred intervention network randomised experiment

Published online by Cambridge University Press:  23 February 2026

Sophie Hu ,
Marie-Eve Carrier ,
Meira Golberg ,
Marie -Claude Geoffroy ,
Linda Kwakkenbos ,
Susan J. Bartlett ,
Catherine Fortuné ,
Amy Gietzen ,
Karen Gottesman  and
Amanda Lawrie-Jones
...Show all authors
Sophie Hu
Affiliation:
Lady Davis Institute for Medical Research, Jewish General Hospital, Montreal, QC, Canada Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, QC, Canada
Marie-Eve Carrier
Affiliation:
Lady Davis Institute for Medical Research, Jewish General Hospital, Montreal, QC, Canada
Meira Golberg
Affiliation:
Lady Davis Institute for Medical Research, Jewish General Hospital, Montreal, QC, Canada
Marie -Claude Geoffroy
Affiliation:
Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, QC, Canada Douglas Mental Health University Institute, Montreal, QC, Canada Department of Psychiatry, McGill University, Montreal, QC, Canada
Linda Kwakkenbos
Affiliation:
Department of Clinical Psychology, Behavioural Science Institute, Radboud University, Nijmegen, the Netherlands Radboudumc Center for Mindfulness, Department of Psychiatry, Radboud University Medical Center, Nijmegen, the Netherlands
Susan J. Bartlett
Affiliation:
Department of Medicine, McGill University, Montreal, QC, Canada Research Institute of the McGill University Health Centre, Montreal, QC, Canada
Catherine Fortuné
Affiliation:
Ottawa Scleroderma Support Group, Ottawa, ON, Canada
Amy Gietzen
Affiliation:
Steffens Scleroderma Foundation, Albany, New York, NY, USA
Karen Gottesman
Affiliation:
National Scleroderma Foundation, Los Angeles, CA, USA
Amanda Lawrie-Jones
Affiliation:
Scleroderma Australia, Melbourne, VIC, Australia Scleroderma Victoria, Melbourne, VIC, Australia
Vanessa L. Malcarne
Affiliation:
Department of Psychology, San Diego State University, San Diego, CA, USA San Diego Joint Doctoral Program in Clinical Psychology, San Diego State University/University of California, San Diego, CA, USA
Michelle Richard
Affiliation:
Scleroderma Atlantic, Halifax, NS, Canada
Maureen Sauvé
Affiliation:
Scleroderma Society of Ontario, Hamilton, ON, Canada Scleroderma Canada, Hamilton, ON, Canada
Luc Mouthon
Affiliation:
Service de Médecine Interne, Centre de Référence Maladies Autoimmunes et Autoinflammatoires Systémiques Rares d’Ile de France, de l’Est et de l’Ouest, Hôpital Cochin, Paris, France Assistance Publique Hôpitaux de Paris-Centre, Hôpital Cochin, Université Paris Cité, Paris, France
Andrea Benedetti
Affiliation:
Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, QC, Canada Research Institute of the McGill University Health Centre, Montreal, QC, Canada Respiratory Epidemiology and Clinical Research Unit, McGill University Health Centre, Montreal, QC, Canada
Brett D. Thombs*
Affiliation:
Lady Davis Institute for Medical Research, Jewish General Hospital, Montreal, QC, Canada Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, QC, Canada Department of Psychiatry, McGill University, Montreal, QC, Canada Department of Medicine, McGill University, Montreal, QC, Canada
*
Corresponding author: Brett D. Thombs; Email: brett.thombs@mcgill.ca
Rights & Permissions [Opens in a new window]

Abstract

Aims

Assessing depression symptoms in people with a chronic illness is challenging due to possible bias from overlapping somatic symptoms associated with both depression and chronic illnesses. Previous studies, however, have found that people with a chronic illness do not report more somatic symptoms on depression measures than people without a chronic illness with similar levels of mood and cognitive symptoms. The reason for this surprising finding is unknown. Our primary objective was to evaluate differences in mean sum scores of Patient Health Questionnaire-8 (PHQ-8) somatic symptom items (sleep disturbances, fatigue, appetite changes) in people with a chronic illness when the items were administered outside the context of a depression questionnaire versus as part of the PHQ-8. Secondary objectives were to evaluate individual somatic item scores. We hypothesised that people who completed somatic items outside of a depression assessment would have significantly higher scores than those who completed items as part of a depression assessment.

Methods

We conducted a randomised controlled experiment within the Scleroderma Patient-centred Intervention Network (SPIN) Cohort, a multinational cohort of people with systemic sclerosis. SPIN Cohort participants were randomly allocated to complete the PHQ-8 with somatic items (sleep disturbances, fatigue, appetite changes) presented separately from psychological items and without any indication that they were part of a depression questionnaire (Reordered Items arm) or in standard format (Standard PHQ-8 arm). Participants were automatically randomised when they logged into the SPIN Cohort platform to complete routine research assessments. The primary outcome was the mean sum score of PHQ-8 somatic items. Secondary outcomes were the mean scores of individual somatic items. Differences were assessed using between-groups t-tests.

Results

In total, 851 participants were included (N = 428 in Reordered Items arm, N = 423 in Standard PHQ-8 arm). Mean (SD) PHQ-8 score was 6.0 (5.3) for all participants. We found no statistically significant differences in PHQ-8 somatic item sum scores (0.05 points; 95% confidence interval [CI]: −0.29 to 0.38) or in mean scores for item 3 (sleep disturbances; 0.04 points; 95% CI: −0.09 to 0.19), item 4 (fatigue; 0.03 points; 95% CI: −0.11 to 0.16) and item 5 (appetite changes; −0.03 points; 95% CI: −0.15 to 0.10).

Conclusions

We did not find evidence that responses to PHQ-8 somatic items were influenced by whether participants were aware they were responding to items about depression. This finding supports the validity of self-reported questionnaires for depression symptom assessment in people with chronic medical conditions.

Keywords

assessmentdepressionrandomised controlled trialsomatic symptomssystemic sclerosis

Information

Type
Original Article
Information
Epidemiology and Psychiatric Sciences , Volume 35 , 2026 , e11
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2026. Published by Cambridge University Press.

Introduction

Major depression symptoms, as described in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), include cognitive and mood-related psychological symptoms as well as somatic symptoms, including fatigue, poor appetite or overeating and insomnia or hypersomnia (American Psychiatric Association, 2013). Most self-report depression symptom questionnaires include items that reflect both psychological and somatic symptoms (Sakakibara et al., Reference Sakakibara, Miller, Orenczuk and Wolfe2009; Wakefield et al., Reference Wakefield, Butow, Aaronson, Hack, Hulbert-Williams and Jacobsen2015). Scores obtained using such questionnaires, however, may be artefactually inflated in people with a chronic illness due to the overlap of somatic symptoms related to depression and those stemming from their physical illness (von Ammon Cavanaugh, Reference von Ammon Cavanaugh1995; Koenig et al., Reference Koenig, George, Peterson and Pieper1997; Sørensenf et al., Reference Sørensenf, Friis-Hasché, Haghfelt and Bech2005).

The 9-item Patient Health Questionnaire-9 (PHQ-9) (Kroenke et al., Reference Kroenke, Spitzer and Williams2001) and its 8-item version (PHQ-8) (Wu et al., Reference Wu, Levis, Riehm, Saadat, Levis, Azar, Rice, Boruff, Cuijpers, Gilbody, Ioannidis, Kloda, McMillan, Patten, Shrier, Ziegelstein, Akena, Arroll, Ayalon, Baradaran, Baron, Bombardier, Butterworth, Carter, Chagas, Chan, Cholera, Conwell, de Man-van Ginkel, Fann, Fischer, Fung, Gelaye, Goodyear-Smith, Greeno, Hall, Harrison, Härter, Hegerl, Hides, Hobfoll, Hudson, Hyphantis, Inagaki, Jetté, Khamseh, Kiely, Kwan, Lamers, Liu, Lotrakul, Loureiro, Löwe, McGuire, Mohd-Sidik, Munhoz, Muramatsu, Osório, Patel, Pence, Persoons, Picardi, Reuter, Rooney, Santos, Shaaban, Sidebottom, Simning, Stafford, Sung, Tan, Turner, van Weert, White, Whooley, Winkley, Yamada, Benedetti and Thombs2020) are commonly used for identifying people who may have depression (Moriarty et al., Reference Moriarty, Gilbody, McMillan and Manea2015; Levis et al., Reference Levis, Benedetti and Thombs2019) and as continuous measures to assess depression symptom severity (Kroenke et al., Reference Kroenke, Spitzer and Williams2001). The 9 items of the PHQ-9 align with the 9 DSM-5 symptom criteria for a major depressive episode (American Psychiatric Association, 2013), and the PHQ-8 includes all PHQ-9 items except an item on thoughts of suicide or self-harm (Wu et al., Reference Wu, Levis, Riehm, Saadat, Levis, Azar, Rice, Boruff, Cuijpers, Gilbody, Ioannidis, Kloda, McMillan, Patten, Shrier, Ziegelstein, Akena, Arroll, Ayalon, Baradaran, Baron, Bombardier, Butterworth, Carter, Chagas, Chan, Cholera, Conwell, de Man-van Ginkel, Fann, Fischer, Fung, Gelaye, Goodyear-Smith, Greeno, Hall, Harrison, Härter, Hegerl, Hides, Hobfoll, Hudson, Hyphantis, Inagaki, Jetté, Khamseh, Kiely, Kwan, Lamers, Liu, Lotrakul, Loureiro, Löwe, McGuire, Mohd-Sidik, Munhoz, Muramatsu, Osório, Patel, Pence, Persoons, Picardi, Reuter, Rooney, Santos, Shaaban, Sidebottom, Simning, Stafford, Sung, Tan, Turner, van Weert, White, Whooley, Winkley, Yamada, Benedetti and Thombs2020). In contrast to what might be expected, existing research suggests that having a medical illness and, among people with a medical illness, condition severity are not associated with higher PHQ-9 somatic symptom item scores (Leavens et al., Reference Leavens, Patten, Hudson, Baron and Thombs2012; Jones et al., Reference Jones, Ludman, McCorkle, Reid, Bowles, Penfold and Wagner2015; Cook et al., Reference Cook, Kallen, Bombardier, Bamer, Choi, Kim, Salem and Amtmann2017; Hu and Ward, Reference Hu and Ward2017; Marrie et al., Reference Marrie, Lix, Bolton, Fisk, Fitzgerald, Graff, Hitchon, Kowalec, Marriott, Patten, Salter and Bernstein2023). Differential item functioning (DIF) analyses assess the degree to which questionnaire items are likely to measure the intended construct and not be influenced by other external factors (Walker, Reference Walker2011). If DIF based on having a chronic illness were present, people with an illness would score higher on PHQ-9 somatic symptom items than people without an illness with similar psychological symptom item scores, presumably due to overlap with their medical symptoms.

We identified 5 studies on DIF in the PHQ-9 (Leavens et al., Reference Leavens, Patten, Hudson, Baron and Thombs2012; Jones et al., Reference Jones, Ludman, McCorkle, Reid, Bowles, Penfold and Wagner2015; Cook et al., Reference Cook, Kallen, Bombardier, Bamer, Choi, Kim, Salem and Amtmann2017; Hu and Ward, Reference Hu and Ward2017; Marrie et al., Reference Marrie, Lix, Bolton, Fisk, Fitzgerald, Graff, Hitchon, Kowalec, Marriott, Patten, Salter and Bernstein2023), including 4 that compared people with and without a medical illness (Leavens et al., Reference Leavens, Patten, Hudson, Baron and Thombs2012; Cook et al., Reference Cook, Kallen, Bombardier, Bamer, Choi, Kim, Salem and Amtmann2017; Hu and Ward, Reference Hu and Ward2017; Marrie et al., Reference Marrie, Lix, Bolton, Fisk, Fitzgerald, Graff, Hitchon, Kowalec, Marriott, Patten, Salter and Bernstein2023) and one that examined whether having more cancer-related somatic symptoms was associated with PHQ-9 somatic item responses (Jones et al., Reference Jones, Ludman, McCorkle, Reid, Bowles, Penfold and Wagner2015). None of the studies found detectable DIF on any somatic items that meaningfully influenced total measure scores. One of the studies (Leavens et al., Reference Leavens, Patten, Hudson, Baron and Thombs2012) was conducted among people with systemic sclerosis (SSc; also known as scleroderma), a complex, rare, chronic, autoimmune disease characterised by microvascular damage and fibrosis of the skin and other organs, including the lungs, gastrointestinal tract, kidneys and heart (Allanore et al., Reference Allanore, Simms, Distler, Trojanowska, Pope, Denton and Varga2015; Denton and Khanna, Reference Denton and Khanna2017; Volkmann et al., Reference Volkmann, Andréasson and Smith2023). Disease onset typically peaks around age 50, and over 80% of people with SSc are women (Allanore et al., Reference Allanore, Simms, Distler, Trojanowska, Pope, Denton and Varga2015). Common symptoms include skin thickening, Raynaud’s phenomenon, difficulty breathing, limitations in hand mobility, and 3 symptoms measured by PHQ-9 items: fatigue, sleep problems and gastrointestinal issues that can lead to changes in appetite (Allanore et al., Reference Allanore, Simms, Distler, Trojanowska, Pope, Denton and Varga2015; Denton and Khanna, Reference Denton and Khanna2017; Volkmann et al., Reference Volkmann, Andréasson and Smith2023).

It is possible that people with somatic symptoms from medical conditions do not score higher on somatic symptom items than people without medical conditions due to item-order or context effects. Item-order or context effects occur in questionnaires when the order in which items are presented provides a context that affects how people interpret and respond to items (Bowling and Windsor, Reference Bowling and Windsor2008; Lietz, Reference Lietz2010; Thau et al., Reference Thau, Mikkelsen, Hjortskov and Pedersen2021; Lee et al., Reference Lee, Krishnamurty, Van Horn, Cooper, Coutanche, McMullen, Panter, Rindskopf and Sher2023). If people with a chronic illness with substantial somatic symptom burden are aware that they are answering questions about depression, they might not report overlapping somatic symptoms because they attribute them entirely to their physical illness and not to depression, which is more often associated with cognitive and mood-related symptoms. If this were the case, these people would theoretically score higher on items measuring somatic symptoms of depression if these items were removed from the context of a depression questionnaire.

We did not identify any studies that have examined whether similar reporting of somatic symptoms on depression symptom questionnaires between people with and without medical conditions may be due to context effects and the awareness that depression symptoms are being assessed. Our primary objective was to evaluate differences in mean PHQ-8 somatic item sum scores (item 3 = sleep disturbances, item 4 = fatigue, item 5 = appetite changes) in people with SSc when the items were administered outside the context of a depression questionnaire versus as part of the PHQ-8. Our secondary objective was to evaluate differences in mean scores for each individual PHQ-8 somatic item when administered outside of or as part of the PHQ-8. We hypothesised that people who completed somatic items outside the context of a depression assessment would have significantly higher scores than those who completed the same items as part of a depression assessment.

Methods

Study design

We conducted a two-arm parallel superiority randomised controlled experiment with a 1:1 allocation ratio within the Scleroderma Patient-centred Intervention Network (SPIN) Cohort. The SPIN Cohort is a large multinational cohort that collects longitudinal data on patient-reported outcomes from people living with SSc (Scleroderma Patient-centered Intervention Network, 2024). Cohort participants complete a series of patient-reported outcome measures online every 3 months. In this experiment, when participants logged in to complete their routine assessment, they were randomly assigned to Reordered Items or Standard PHQ-8 arms. Participants in the Reordered Items arm completed a reordered version of the PHQ-8 in which the 3 PHQ-8 somatic items (item 3 = sleep disturbances, item 4 = fatigue and item 5 = appetite changes) were presented first, without any indication that they were part of a depression questionnaire, and the 5 psychological items were presented later in the assessment protocol.

The SPIN Cohort study was approved by the Research Ethics Committee of the Centre intégré universitaire de santé et de services sociaux du Centre-Ouest-de-l’Île-de-Montréal (#MP-05-2013-150) and by the ethics committees of all recruiting sites. Since the only change to routine cohort assessments was changing the order of item presentation, the Research Ethics Committee of the Centre intégré universitaire de santé et de services sociaux du Centre-Ouest-de-l’Île-de-Montréal determined that no additional ethics approval was needed and that SPIN Cohort participants did not need to be notified about the experiment.

We registered the experiment (ClinicalTrials.gov, NCT06772896) and posted a protocol on the Open Science Framework (https://osf.io/qkf8g/files/5kn49) prior to initiation. We reported results consistent with the Consolidated Standards of Reporting Trials statement (Moher et al., Reference Moher, Hopewell, Schulz, Montori, Gøtzsche, Devereaux, Elbourne, Egger and Altman2010). Studies that use SPIN Cohort data have similar methods. Thus, the results from our experiment were reported as consistent with guidance from the Text Recycling Research Project (Hall et al., Reference Hall, Moskovitz and Pemberton2021).

Eligibility criteria

Eligible SPIN Cohort participants must be classified as having SSc based on 2013 American College of Rheumatology/European League Against Rheumatism criteria (van den Hoogen et al., Reference van den Hoogen, Khanna, Fransen, Johnson, Baron, Tyndall, Matucci-Cerinic, Naden, Medsger, Carreira, Riemekasten, Clements, Denton, Distler, Allanore, Furst, Gabrielli, Mayes, van Laar, Seibold, Czirjak, Steen, Inanc, Kowal-Bielecka, Müller-Ladner, Valentini, Veale, Vonk, Walker, Chung, Collier, Csuka, Fessler, Guiducci, Herrick, Hsu, Jimenez, Kahaleh, Merkel, Sierakowski, Silver, Simms, Varga and Pope2013) confirmed by a SPIN physician; aged ≥18 years; fluent in English, French, or Spanish; and have access to a computer or tablet with internet access. The SPIN Cohort is a convenience sample. Cohort participants are recruited during regular medical visits at SPIN recruitment sites (Scleroderma Patient-centered Intervention Network, 2024), and written informed consent is obtained. A medical form is submitted online by site personnel to enrol participants. Once online registration is completed, participants are sent an automated welcome email with instructions on how to activate their SPIN account and complete SPIN Cohort measures online. Cohort participants complete routine online assessments that last approximately 20 minutes upon enrolment and at 3-month intervals. All SPIN Cohort participants who logged in to the SPIN Cohort platform to complete their routine 3-month assessment during the period when the experiment was conducted were included.

Experiment arms

Participants assigned to the Reordered Items arm received an assessment protocol in which the 3 PHQ-8 somatic items were presented separately from and prior to the 5 PHQ-8 psychological items without any indication that they were part of a depression symptom questionnaire. PHQ-8 psychological items were presented (1) later in the assessment and (2) with at least 1 other patient-reported outcome measure between the somatic items and psychological items. Apart from these 2 rules, as routinely done in SPIN Cohort assessments, all measures were presented in random order.

Participants assigned to the Standard PHQ-8 arm served as the control arm and completed the standard version of the PHQ-8, with the PHQ-8 and all other measures administered in random order. The title of the PHQ-8 was not presented to participants in either experiment arm; the PHQ-8 does not include instructions to respondents. PHQ-8 items were presented on 2 separate webpages in the Reordered Items arm and on a single webpage in the Standard PHQ-8 arm. The routine online assessment also contained 9 other measures, including the Health Literacy Survey Questionnaire (Finbråten et al., Reference Finbråten, Wilde-Larsson, Nordström, Pettersen, Trollvik and Guttersrud2018) and the Patient-Reported Outcomes Measurement Information System (Hays et al., Reference Hays, Spritzer, Schalet and Cella2018), which included 8 measures.

Randomisation and blinding

The SPIN Cohort platform was programmed to randomise each participant via simple randomisation in a 1:1 ratio to either the Reordered Items or Standard PHQ-8 arm. Randomisation occurred automatically and immediately when participants logged in to complete their routine assessment. Thus, study investigators were fully blind to participants’ assigned experiment arm. Since participants were not notified about the experiment, they were fully blind to the study objectives and their assigned experiment arm.

Participant characteristics and experiment outcomes

SPIN physicians provided age, sex and medical information upon enrolment of participants in the SPIN Cohort. Participants reported sociodemographic data, including race or ethnicity, education level and marital status.

The primary outcome analysis compared the mean sum score of the 3 PHQ-8 somatic items (item 3 = sleep disturbances, item 4 = fatigue and item 5 = appetite changes) between participants in the two experimental arms. Secondary outcome variables were individual mean scores of the 3 PHQ-8 somatic items.

The PHQ-8 consists of 8 items that measure depression symptoms over the last 2 weeks. Items are scored on a 4-point scale ranging from 0 (not at all) to 3 (nearly every day), with higher scores (range 0–24) indicating more depression symptoms. The PHQ-8 has been shown to be equivalent to the PHQ-9 (Wu et al., Reference Wu, Levis, Riehm, Saadat, Levis, Azar, Rice, Boruff, Cuijpers, Gilbody, Ioannidis, Kloda, McMillan, Patten, Shrier, Ziegelstein, Akena, Arroll, Ayalon, Baradaran, Baron, Bombardier, Butterworth, Carter, Chagas, Chan, Cholera, Conwell, de Man-van Ginkel, Fann, Fischer, Fung, Gelaye, Goodyear-Smith, Greeno, Hall, Harrison, Härter, Hegerl, Hides, Hobfoll, Hudson, Hyphantis, Inagaki, Jetté, Khamseh, Kiely, Kwan, Lamers, Liu, Lotrakul, Loureiro, Löwe, McGuire, Mohd-Sidik, Munhoz, Muramatsu, Osório, Patel, Pence, Persoons, Picardi, Reuter, Rooney, Santos, Shaaban, Sidebottom, Simning, Stafford, Sung, Tan, Turner, van Weert, White, Whooley, Winkley, Yamada, Benedetti and Thombs2020), which is a valid measure of depression symptoms in SSc (Milette et al., Reference Milette, Hudson and Thombs2010). The PHQ-8 is available in English, French and Spanish (Arthurs et al., Reference Arthurs, Steele, Hudson, Baron and Thombs2012; Gómez-Gómez et al., Reference Gómez-Gómez, Benítez, Bellón, Moreno-Peral, Oliván-Blázquez, Clavería, Zabaleta-Del-Olmo, Llobera, Serrano-Ripoll, Tamayo-Morales and Motrico2023). Only 2 studies have examined the minimal important difference of the PHQ-9 using anchor-based approaches, and they estimated a minimal important difference of between 2.0 and 3.7 points (Bauer-Staeb et al., Reference Bauer-Staeb, Kounali, Welton, Griffith, Wiles, Lewis, Faraway and Button2021; Kounali et al., Reference Kounali, Button, Lewis, Gilbody, Kessler, Araya, Duffy, Lanham, Peters, Wiles and Lewis2022). Many studies on the factor structure of the PHQ-9 have found that a one-factor model adequately explains item variance (Lamela et al., Reference Lamela, Soreira, Matos and Morais2020). However, other studies have suggested that a two-factor model, consisting of somatic and psychological latent factors, provides a better fit (Chilcot et al., Reference Chilcot, Rayner, Lee, Price, Goodwin, Monroe, Sykes, Hansford and Hotopf2013). Somatic latent factors in two-factor models of the PHQ-9 include 3–5 items (Lamela et al., Reference Lamela, Soreira, Matos and Morais2020). The 3-item version consists of item 3 (sleep disturbance), item 4 (fatigue) and item 5 (appetite changes) (Chilcot et al., Reference Chilcot, Rayner, Lee, Price, Goodwin, Monroe, Sykes, Hansford and Hotopf2013; Lamela et al., Reference Lamela, Soreira, Matos and Morais2020), which are commonly experienced in SSc. We performed a two-factor confirmatory factor analysis (CFA) to verify the 3-item somatic latent factor using SPIN Cohort data and found that the model fit well (Comparative Fit Index = 0.997; Tucker–Lewis Index = 0.995; Root Mean Square Error of Approximation = 0.058; Standardised Root Mean Square Residual = 0.036). See Appendix 1 for CFA methods and results.

Statistical analysis

In August 2024, prior to initiating the experiment, we determined that 903, 910, 851 and 798 SPIN Cohort participants completed all measures in their scheduled assessments during the previous 4 3-month periods (August 2023–October 2023, November 2023–January 2024, February 2024–April 2024 and May 2024–July 2024). Assuming that at least 798 participants would log in to the SPIN Cohort platform to complete an assessment and be randomised, using a two-tailed test with a significance level of 0.05, we calculated that we would be able to detect an estimated effect size of 0.20 standardised mean difference, which is considered a small difference (Cohen, Reference Cohen1988), with 80% power.

Two-tailed between-groups t-tests with 95% confidence intervals (CIs) were used to compare the mean sum score of the 3 PHQ-8 somatic items (item 3 = sleep disturbances, item 4 = fatigue and item 5 = appetite changes) and mean scores of individual somatic items between participants in the Reordered Items and Standard PHQ-8 arms. We determined pre-experiment that analyses would be conducted as complete case analyses because we did not expect significant missing data. However, per our protocol, if >10% of randomised participants had not completed all PHQ-8 items, we would have conducted intent-to-treat analyses with missing data handled using multiple imputations by chained equations (Rubin, Reference Rubin1987; van Buuren and Groothuis-Oudshoorn, Reference van Buuren and Groothuis-Oudshoorn2011). Analyses were conducted using the statistical software R (R Core Team, n.d.).

Post hoc analysis

We conducted post hoc subgroup analyses at the request of a peer reviewer to explore the possible effects of sex (male and female), age (≤60 years and >60 years), SSc subtype (diffuse and limited) and language (English, French, Spanish) on our results. We used linear regression models to examine the interaction between each subgroup variable and our experiment arms.

Results

The experiment was conducted during participant assessments from 17 January to 17 April 2025.

Participants

Of 881 SPIN Cohort participants who were randomised, 30 (3%) did not complete all PHQ-8 items and were excluded. Thus, 851 (97%) participants with complete PHQ-8 data were included in the analyses in the Reordered Items (N = 428) and Standard PHQ-8 (N = 423) arms. There were no substantive differences between participants who completed all PHQ-8 items and those who did not complete all items. See Appendix 2 for a comparison of characteristics between included and excluded participants. The flow of participants is presented in Figure 1.

Figure 1. Participant flow diagram.

Sociodemographic variables and disease characteristics are presented in Table 1. Mean (SD) age was 62.3 (11.8) years, 752 participants were female (88%), 703 participants identified as White (83%), 291 participants were classified as having diffuse SSc (34%) and mean (SD) time since onset of first non-Raynaud’s disease manifestation was 16.9 (9.8) years. Participants were from France (N = 326, 38%), the United States (N = 218, 26%), Canada (N = 197, 23%), the United Kingdom (N = 75, 9%) and Australia, Spain, or Mexico (N = 35, 4%). Characteristics of participants in the Reordered Items and Standard PHQ-8 arms were similar.

Table 1. Participant sociodemographic and disease characteristics at baseline

a Due to missing data, total sample size N < 851 for some characteristics.

b Race or ethnicity data were self-reported in each country using standard categories used in that country. Therefore, categories differed between countries.

Outcomes

The mean (SD) PHQ-8 score was 6.0 (5.3) for the full sample, 6.0 (5.4) for Reordered Items arm participants and 6.0 (5.2) for Standard PHQ-8 arm participants. Outcome comparisons are presented in Table 2. The difference in PHQ-8 somatic item sum scores was not statistically significant (0.05 points; 95% CI −0.29 to 0.38). Differences in mean item scores were 0.04 points (95% CI −0.09 to 0.19) for item 3 (sleep disturbances), 0.03 points (95% CI −0.11 to 0.16) for item 4 (fatigue) and −0.03 points (95% CI −0.15 to 0.10) for item 5 (appetite changes).

Table 2. Differences in mean Patient Health Questionnaire-8 somatic item scores between the Reordered Items and Standard PHQ-8 groups

CI, confidence interval.

a Differences between groups are calculated as Standard PHQ-8 – Reordered Items.

See Appendix 3 for post hoc analysis results. We did not find any significant subgroup interactions.

Discussion

We examined whether administering PHQ-8 somatic items outside the context of a depression questionnaire would influence item scores among people with SSc, a chronic condition with substantial somatic symptom burden, including symptoms that overlap with PHQ-8 somatic items (sleep disturbance, fatigue, appetite changes). We did not find a statistically significant or substantive difference in mean PHQ-8 somatic item sum scores between participants who completed these items outside the context of a depression symptom assessment or as part of the standard PHQ-8. Our estimated difference of 0.05 points was 1.4%–2.5% of the minimal important difference (Bauer-Staeb et al., Reference Bauer-Staeb, Kounali, Welton, Griffith, Wiles, Lewis, Faraway and Button2021; Kounali et al., Reference Kounali, Button, Lewis, Gilbody, Kessler, Araya, Duffy, Lanham, Peters, Wiles and Lewis2022). There were no statistically significant or substantive differences in individual somatic item scores.

To the best of our knowledge, this is the first study to test the hypothesis that context effects may explain why people with chronic illnesses with substantial somatic symptom burden do not report more somatic items on self-report depression assessments like the PHQ-8 than people with similar psychological symptom scores but without a chronic illness. Our results suggest that findings from previous studies that did not find DIF in PHQ-9 scores (Leavens et al., Reference Leavens, Patten, Hudson, Baron and Thombs2012; Jones et al., Reference Jones, Ludman, McCorkle, Reid, Bowles, Penfold and Wagner2015; Cook et al., Reference Cook, Kallen, Bombardier, Bamer, Choi, Kim, Salem and Amtmann2017; Hu and Ward, Reference Hu and Ward2017; Marrie et al., Reference Marrie, Lix, Bolton, Fisk, Fitzgerald, Graff, Hitchon, Kowalec, Marriott, Patten, Salter and Bernstein2023) are not due to context effects.

Over 85% of SPIN Cohort participants have gastrointestinal symptoms. Previous studies have found that approximately 90% of people with SSc experience fatigue and that more than 75% have sleep difficulties (Bassel et al., Reference Bassel, Hudson, Taillefer, Schieir, Baron and Thombs2011; Willems et al., Reference Willems, Kwakkenbos, Leite, Thombs, van den Hoogen, Maia, Vliet Vlieland and van den Ende2014). It is possible that people with SSc may become accustomed to these symptoms over time, which would affect how they view their severity. In this case, they might view the same symptom presentation as less severe compared to someone without a chronic illness. Another possibility is that our findings reflect shared pathways in physical and mental disorders. Large bodies of evidence have linked depression to inflammation and immune symptom dysfunction, both of which are core elements to many chronic conditions (Pasco et al., Reference Pasco, Moylan, Allen, Stuart, Hayley, Byrne and Maes2013; Miller and Raison, Reference Miller and Raison2016; Beurel et al., Reference Beurel, Toups and Nemeroff2020; Troubat et al., Reference Troubat, Barone, Leman, Desmidt, Cressant, Atanasova, Brizard, El Hage, Surget, Belzung and Camus2021), including SSc (Allanore et al., Reference Allanore, Simms, Distler, Trojanowska, Pope, Denton and Varga2015). It is possible that somatic symptoms of depression cannot be easily separated from similar somatic symptoms in SSc.

Our study has several strengths. We were able to include approximately 850 people, which provided sufficient power to detect even very small differences if they were present. Since we conducted our experiment as part of routine SPIN Cohort assessments, participants were not aware that an experiment was being done and were, thus, fully blinded to study conditions and our hypothesis. There are also limitations to consider. First, the SPIN Cohort is a convenience sample, and the outcome measure was completed online, which may reduce generalisability. However, participant characteristics from the SPIN Cohort have been shown to be comparable to those of other large SSc cohorts (Dougherty et al., Reference Dougherty, Kwakkenbos, Carrier, Salazar, Assassi, Baron, Bartlett, Furst, Gottesman, van den Hoogen, Malcarne, Mouthon, Nielson, Poiraudeau, Sauvé, Boire, Bruns, Chung, Denton, Dunne, Fortin, Frech, Gill, Gordon, Herrick, Hinchcliff, Hudson, Johnson, Jones, Kafaja, Larché, Manning, Pope, Spiera, Steen, Sutton, Thorne, Wilcox, Thombs and Mayes2018). Furthermore, it is unlikely that differences in sociodemographic or SSc disease characteristics would alter the degree to which context effects might play a role in depression symptom assessment. Second, it is possible that some SPIN Cohort participants may have recognised PHQ-8 somatic items even when they were administered separately and without reference to the PHQ-8. We believe that this is unlikely since the PHQ-8 had not been administered as part of SPIN Cohort routine assessments since October 2020, over 4 years before our experiment. Furthermore, the individual PHQ-8 items are written in a way that does not give any indication that they refer to somatic symptoms related to depression or any other cause.

In summary, we conducted a randomised experiment to test whether context effects may explain why people with chronic illnesses do not appear to report more somatic symptoms on depression self-report measures compared to others without chronic illnesses with similar psychological symptom levels. Our results showed no difference in PHQ-8 somatic item scores when these were administered separately from cognitive and mood-related items versus as part of the standard PHQ-8. We did not find evidence that people’s knowledge that depression is being assessed influences how they report somatic symptoms of depression that may overlap with symptoms of physical illnesses.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/S204579602610047X.

Availability of data and materials

De-identified participant data reported in this study will be made available upon request to the corresponding author.

Acknowledgements

None.

Financial support

The Scleroderma Patient-centred Intervention Network Cohort has received funding from the Canadian Institutes of Health Research (CIHR); the Arthritis Society; the Lady Davis Institute for Medical Research of the Jewish General Hospital, Montreal, Quebec, Canada; the Jewish General Hospital Foundation, Montreal, Quebec, Canada; McGill University, Montreal, Quebec, Canada; the Scleroderma Society of Ontario; Scleroderma Canada; Sclérodermie Québec; Scleroderma Manitoba; Scleroderma Atlantic; the Scleroderma Association of BC; Scleroderma SASK; Scleroderma Australia; Scleroderma New South Wales; Scleroderma Victoria and the Scleroderma Foundation of California. SH was supported by a CIHR Canadian Graduate Scholarship – Master’s award, and MCG and BDT by Canada Research Chairs, all outside of the present work.

Open access funding provided by Radboud University Nijmegen. Open access funding provided by Radboud University Medical Center.

Competing interests

The authors have no conflicts of interest to declare.

Ethical standards

The SPIN Cohort study was approved by the Research Ethics Committee of the Centre intégré universitaire de santé et de services sociaux du Centre-Ouest-de-l’Île-de-Montréal (#MP-05-2013-150) and by the ethics committees of all recruiting sites. No additional ethics approval was needed for the present study.

Footnotes

SPIN Investigators: Claire E. Adams, Lady Davis Institute for Medical Research, Montreal, Quebec, Canada; Marie Hudson, Department of Medicine, McGill University, Montreal, Quebec, Canada; Maureen D. Mayes, University of Texas McGovern School of Medicine, Houston, Texas, USA; James Stempel, Scleroderma Chicago, Chicago, Illinois, USA; Robyn K. Wojeck, Amgen Inc., Thousand Oaks, California, USA; Christian Agard, Centre Hospitalier Universitaire – Hôtel-Dieu de Nantes, Nantes, France; Laurent Alric, CHU Rangueil, Toulouse, France; Marc André, Centre Hospitalier Universitaire Gabriel-Montpied, Clermont-Ferrand, France; Floryan Beaslay, CHU La Réunion, Saint-Denis, La Réunion, France; Elana J. Bernstein, Columbia University, New York, New York, USA; Sabine Berthier, Centre Hospitalier Universitaire Dijon Bourgogne, Dijon, France; Lyne Bissonnette, Université de Sherbrooke, Sherbrooke, Quebec, Canada; Sophie Blaise, CHU Grenoble Alpes, Grenoble, France; Eva Bories, CHU Rangueil, Toulouse, France; Alessandra Bruns, Université de Sherbrooke, Sherbrooke, Quebec, Canada; Carlotta Cacciatore, Assistance Publique – Hôpitaux de Paris, Hôpital St-Louis, Paris, France; Patricia Carreira, Servicio de Reumatologia del Hospital 12 de Octubre, Madrid, Spain; Marion Casadevall, Assistance Publique – Hôpitaux de Paris, Hôpital Cochin, Paris, France; Benjamin Chaigne, Assistance Publique – Hôpitaux de Paris, Hôpital Cochin, Paris, France; Lorinda Chung, Stanford University, Stanford, California, USA; Benjamin Crichi, Assistance Publique – Hôpitaux de Paris, Hôpital St-Louis, Paris, France; Thylbert Deltombe, CHU La Réunion, Saint-Denis, La Réunion, France; Christopher P. Denton, Royal Free London Hospital, London, UK; Tannvir Desroche, CHU La Réunion, Saint-Denis, La Réunion, France; Robyn Domsic, University of Pittsburgh, Pittsburgh, Pennsylvania, USA; James V. Dunne, St. Paul’s Hospital and University of British Columbia, Vancouver, British Columbia, Canada; Bertrand Dunogue, Assistance Publique – Hôpitaux de Paris, Hôpital Cochin, Paris, France; Regina Fare, Servicio de Reumatologia del Hospital 12 de Octubre, Madrid, Spain; Dominique Farge-Bancel, Assistance Publique – Hôpitaux de Paris, Hôpital St-Louis, Paris, France; Paul R. Fortin, CHU de Québec – Université Laval, Quebec, Quebec, Canada; Tracy Frech, Vanderbilt University, Nashville, Tennessee, USA; Loraine Gauzère, CHU La Réunion, Saint-Denis, La Réunion, France; Anne Gerber, CHU La Réunion, Saint-Denis, La Réunion, France; Jessica K. Gordon, Hospital for Special Surgery, New York City, New York, USA; Brigitte Granel-Rey, Université, and Assistance Publique – Hôpitaux de Marseille, Hôpital Nord, Marseille, France; Aurélien Guffroy, Les Hôpitaux Universitaires de Strasbourg, Nouvel Hôpital Civil, Strasbourg, France; Geneviève Gyger, Jewish General Hospital and McGill University, Montreal, Quebec, Canada; Eric Hachulla, Centre Hospitalier Régional Universitaire de Lille, Hôpital Claude Huriez, Lille, France; Daphna Harel, New York University, New York, New York, USA; Monique Hinchcliff, Yale School of Medicine, New Haven, Connecticut, USA; Sabrina Hoa, Centre hospitalier de l’Université de Montréal – CHUM, Montreal, Quebec, Canada; Michael Hugues, Salford Royal NHS Foundation Trust, Salford, UK; Alena Ikic, CHU de Québec – Université Laval, Quebec, Quebec; Sindhu R. Johnson, Toronto Scleroderma Program, Mount Sinai Hospital, Toronto Western Hospital, and University of Toronto, Toronto, Ontario, Canada; Nader Khalidi, McMaster University, Hamilton, Ontario, Canada; Kimberly S. Lakin, Hospital for Special Surgery, New York City, New York, USA; Marc Lambert, Centre Hospitalier Régional Universitaire de Lille, Hôpital Claude Huriez, Lille, France; Maggie Larche, University of Calgary, Calgary, Alberta, Canada; David Launay, Centre Hospitalier Régional Universitaire de Lille, Hôpital Claude Huriez, Lille, France; Yvonne C. Lee, Northwestern University, Chicago, Illinois, USA; Paul Legendre, Centre Hospitalier du Mans, Le Mans, France; Catarina Leite, University of Minho, Braga, Portugal; Hélène Maillard, Centre Hospitalier Régional Universitaire de Lille, Hôpital Claude Huriez, Lille, France; Nancy Maltez, University of Ottawa, Ottawa, Ontario, Canada; Joanne Manning, Salford Royal NHS Foundation Trust, Salford, UK; Isabelle Marie, CHU Rouen, Hôpital de Bois-Guillaume, Rouen, France; Maria Martin Lopez, Servicio de Reumatologia del Hospital 12 de Octubre, Madrid, Spain; Thierry Martin, Les Hôpitaux Universitaires de Strasbourg, Nouvel Hôpital Civil, Strasbourg, France; Ariel Masetto, Université de Sherbrooke, Sherbrooke, Quebec, Canada; Arsène Mekinian, Assistance Publique – Hôpitaux de Paris, Hôpital St-Antoine, Paris, France; Sheila Melchor Díaz, Servicio de Reumatologia del Hospital 12 de Octubre, Madrid, Spain; Morgane Mourguet, CHU Rangueil, Toulouse, France; Christelle Nguyen, Université Paris Descartes, Université de Paris, Paris, France, and Assistance Publique – Hôpitaux de Paris, Paris, France; Karen Nielsen, Scleroderma Society of Ontario, Hamilton, Ontario, Canada; Mandana Nikpour, St Vincent’s Hospital and University of Melbourne, Melbourne, Victoria, Australia; Louis Olagne, Centre Hospitalier Universitaire Gabriel-Montpied, Clermont-Ferrand, France; Vincent Poindron, Les Hôpitaux Universitaires de Strasbourg, Nouvel Hôpital Civil, Strasbourg, France; Janet Pope, University of Western Ontario, London, Ontario, Canada; Susanna Proudman, Royal Adelaide Hospital and University of Adelaide, Adelaide, South Australia, Australia; Grégory Pugnet, CHU Rangueil, Toulouse, France; Loïc Raffray, CHU La Réunion, Saint-Denis, La Réunion, France; François Rannou, Université Paris Descartes, Université de Paris, Paris, France, and Assistance Publique – Hôpitaux de Paris, Paris, France; Alexis Régent, Assistance Publique – Hôpitaux de Paris, Hôpital Cochin, Paris, France; Frederic Renou, CHU La Réunion, Saint-Denis, La Réunion, France; Sébastien Rivière, Assistance Publique – Hôpitaux de Paris, Hôpital St-Antoine, Paris, France; David Robinson, University of Manitoba, Winnipeg, Manitoba, Canada; Esther Rodríguez Almazar, Servicio de Reumatologia del Hospital 12 de Octubre, Madrid, Spain; Tatiana Sofia Rodríguez-Reyna, Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán, Mexico City, Mexico; Sophie Roux, Université de Sherbrooke, Sherbrooke, Quebec, Canada; Perrine Smets, Centre Hospitalier Universitaire Gabriel-Montpied, Clermont-Ferrand, France; Vincent Sobanski, Centre Hospitalier Régional Universitaire de Lille, Hôpital Claude Huriez, Lille, France; Robert F. Spiera, Hospital for Special Surgery, New York City, New York, USA; Virginia Steen, Georgetown University, Washington, DC, USA; Evelyn Sutton, Dalhousie University, Halifax, Nova Scotia, Canada; Carter Thorne, Southlake Regional Health Centre, Newmarket, Ontario, Canada; Damien Vagner, CHU La Réunion, Saint-Denis, La Réunion, France; John Varga, University of Michigan, Ann Arbor, Michigan, USA; Pearce Wilcox, St. Paul’s Hospital and University of British Columbia, Vancouver, British Columbia, Canada; Vanessa Cook, Jewish General Hospital, Montreal, Quebec, Canada; Cassidy Dal Santo, Jewish General Hospital, Montreal, Quebec, Canada; Monica D’Onofrio, Jewish General Hospital, Montreal, Quebec, Canada; and Elsa-Lynn Nassar, Jewish General Hospital, Montreal, Quebec, Canada

References

Allanore, Y, Simms, R, Distler, O, Trojanowska, M, Pope, J, Denton, CP and Varga, J (2015) Systemic sclerosis. Nature Reviews Disease Primers 1, 15002.CrossRefGoogle ScholarPubMed
American Psychiatric Association (2013) Diagnostic and Statistical Manual of Mental Disorders, 5th Edn. Washington, DC: American Psychiatric Publishing.Google Scholar
Arthurs, E, Steele, RJ, Hudson, M, Baron, M and Thombs, BD and Canadian Scleroderma Research Group (2012) Are scores on English and French versions of the PHQ-9 comparable? an assessment of differential item functioning. PLoS One 7(12), e52028.CrossRefGoogle ScholarPubMed
Bassel, M, Hudson, M, Taillefer, SS, Schieir, O, Baron, M and Thombs, BD (2011) Frequency and impact of symptoms experienced by patients with systemic sclerosis: results from a Canadian National Survey. Rheumatology 50(4), 762–767.CrossRefGoogle ScholarPubMed
Bauer-Staeb, C, Kounali, D-Z, Welton, NJ, Griffith, E, Wiles, NJ, Lewis, G, Faraway, JJ and Button, KS (2021) Effective dose 50 method as the minimal clinically important difference: evidence from depression trials. Journal of Clinical Epidemiology 137, 200–208.CrossRefGoogle Scholar
Beurel, E, Toups, M and Nemeroff, CB (2020) The bidirectional relationship of depression and inflammation: double trouble. Neuron 107(2), 234–256.CrossRefGoogle ScholarPubMed
Bowling, A and Windsor, J (2008) The effects of question order and response-choice on self-rated health status in the english longitudinal study of ageing (ELSA). Journal of Epidemiology & Community Health 62(1), 81–85.CrossRefGoogle ScholarPubMed
Chilcot, J, Rayner, L, Lee, W, Price, A, Goodwin, L, Monroe, B, Sykes, N, Hansford, P and Hotopf, M (2013) The factor structure of the PHQ-9 in palliative care. Journal of Psychosomatic Research 75(1), 60–64.CrossRefGoogle ScholarPubMed
Cohen, J (1988) Statistical Power Analysis for the Behavioral Sciences, 2nd Edn New York: Routledge.Google Scholar
Cook, KF, Kallen, MA, Bombardier, C, Bamer, AM, Choi, SW, Kim, J, Salem, R and Amtmann, D (2017) Do measures of depressive symptoms function differently in people with spinal cord injury versus primary care patients: the CES-D, PHQ-9, and PROMIS®-D. Quality of Life Research 26(1), 139–148.CrossRefGoogle Scholar
Denton, CP and Khanna, D (2017) Systemic sclerosis. The Lancet 390(10103), 1685–1699.CrossRefGoogle ScholarPubMed
Dougherty, DH, Kwakkenbos, L, Carrier, ME, Salazar, G, Assassi, S, Baron, M, Bartlett, SJ, Furst, DE, Gottesman, K, van den Hoogen, F, Malcarne, VL, Mouthon, L, Nielson, WR, Poiraudeau, S, Sauvé, M, Boire, G, Bruns, A, Chung, L, Denton, C, Dunne, JV, Fortin, P, Frech, T, Gill, A, Gordon, J, Herrick, AL, Hinchcliff, M, Hudson, M, Johnson, SR, Jones, N, Kafaja, S, Larché, M, Manning, J, Pope, J, Spiera, R, Steen, V, Sutton, E, Thorne, C, Wilcox, P, Thombs, BD, Mayes, MD and SPIN Investigators (2018) The scleroderma patient-centered intervention network cohort: baseline clinical features and comparison with other large scleroderma cohorts. Rheumatology 57(9), 1623–1631.CrossRefGoogle ScholarPubMed
Finbråten, HS, Wilde-Larsson, B, Nordström, G, Pettersen, KS, Trollvik, A and Guttersrud, Ø (2018) Establishing the HLS-Q12 short version of the European health literacy survey questionnaire: latent trait analyses applying rasch modelling and confirmatory factor analysis. BMC Health Services Research 18(1), 506.CrossRefGoogle ScholarPubMed
Gómez-Gómez, I, Benítez, I, Bellón, J, Moreno-Peral, P, Oliván-Blázquez, B, Clavería, A, Zabaleta-Del-Olmo, E, Llobera, J, Serrano-Ripoll, MJ, Tamayo-Morales, O and Motrico, E (2023) Utility of PHQ-2, PHQ-8 and PHQ-9 for detecting major depression in primary health care: a validation study in Spain. Psychological Medicine 53(12), 5625–5635.CrossRefGoogle ScholarPubMed
Hall, S, Moskovitz, C and Pemberton, M (2021) Best practices for researchers. https://textrecycling.org/resources/best-practices-for-researchers/ (accessed 9 December 2024).Google Scholar
Hays, RD, Spritzer, KL, Schalet, BD and Cella, D (2018) PROMIS®-29 v2.0 profile physical and mental health summary scores. Quality of Life Research 27(7), 1885–1891.CrossRefGoogle ScholarPubMed
Hu, J and Ward, MM (2017) Screening for depression in arthritis populations: an assessment of differential item functioning in three self-reported questionnaires. Quality of Life Research 26(9), 2507–2517.CrossRefGoogle ScholarPubMed
Jones, SMW, Ludman, EJ, McCorkle, R, Reid, R, Bowles, EJA, Penfold, R and Wagner, EH (2015) A differential item function analysis of somatic symptoms of depression in people with cancer. Journal of Affective Disorders 170, 131–137.CrossRefGoogle ScholarPubMed
Koenig, HG, George, LK, Peterson, BL and Pieper, CF (1997) Depression in medically ill hospitalized older adults: prevalence, characteristics, and course of symptoms according to six diagnostic schemes. The American Journal of Psychiatry 154(10), 1376–1383.Google ScholarPubMed
Kounali, D, Button, KS, Lewis, G, Gilbody, S, Kessler, D, Araya, R, Duffy, L, Lanham, P, Peters, TJ, Wiles, N and Lewis, G (2022) How much change is enough? evidence from a longitudinal study on depression in UK primary care. Psychological Medicine 52(10), 1875–1882.CrossRefGoogle ScholarPubMed
Kroenke, K, Spitzer, RL and Williams, JB (2001) The PHQ-9: validity of a brief depression severity measure. Journal of General Internal Medicine 16(9), 606–613.CrossRefGoogle ScholarPubMed
Lamela, D, Soreira, C, Matos, P and Morais, A (2020) Systematic review of the factor structure and measurement invariance of the patient health questionnaire-9 (PHQ-9) and validation of the Portuguese version in community settings. Journal of Affective Disorders 276, 220–233.CrossRefGoogle ScholarPubMed
Leavens, A, Patten, SB, Hudson, M, Baron, M and Thombs, BD and Canadian Scleroderma Research Group (2012) Influence of somatic symptoms on patient health questionnaire-9 depression scores among patients with systemic sclerosis compared to a healthy general population sample. Arthritis Care & Research 64(8), 1195–1201.CrossRefGoogle ScholarPubMed
Lee, L, Krishnamurty, P and Van Horn, S (2023) Question Order Effects. In Cooper, H, Coutanche, MN, McMullen, LM, Panter, AT, Rindskopf, D and Sher, KJ (edited by), APA Handbook of Research Methods in Psychology: Foundations, Planning, Measures, and Psychometrics, 2nd Edn. Washington, DC: American Psychological Association, pp. 277–296.Google Scholar
Levis, B, Benedetti, A and Thombs, BD and DEPRESsion Screening Data (DEPRESSD) Collaboration (2019) Accuracy of patient health questionnaire-9 (PHQ-9) for screening to detect major depression: individual participant data meta-analysis. BMJ 365, l1476.CrossRefGoogle ScholarPubMed
Lietz, P (2010) Research into questionnaire design: a summary of the literature. International Journal of Market Research 52(2), 249–272.CrossRefGoogle Scholar
Marrie, RA, Lix, LM, Bolton, JM, Fisk, JD, Fitzgerald, KC, Graff, LA, Hitchon, CA, Kowalec, K, Marriott, JJ, Patten, SB, Salter, A and Bernstein, CN and CIHR Team in Defining the Burden and Managing the Effects of Immune-mediated Inflammatory Disease (2023) Assessment of differential item functioning of the PHQ-9, Hads-D and PROMIS-depression scales in persons with and without multiple sclerosis. Journal of Psychosomatic Research 172, 111415.CrossRefGoogle ScholarPubMed
Milette, K, Hudson, M and Thombs, BD and Canadian Scleroderma Research Group (2010) Comparison of the PHQ-9 and CES-D depression scales in systemic sclerosis: internal consistency reliability, convergent validity and clinical correlates. Rheumatology 49(4), 789–796.CrossRefGoogle ScholarPubMed
Miller, AH and Raison, CL (2016) The role of inflammation in depression: from evolutionary imperative to modern treatment target. Nature Reviews Immunology 16(1), 22–34.CrossRefGoogle Scholar
Moher, D, Hopewell, S, Schulz, KF, Montori, V, Gøtzsche, PC, Devereaux, PJ, Elbourne, D, Egger, M and Altman, DG (2010) CONSORT 2010 explanation and elaboration: updated guidelines for reporting parallel group randomised trials. BMJ 340, c869.CrossRefGoogle ScholarPubMed
Moriarty, AS, Gilbody, S, McMillan, D and Manea, L (2015) Screening and case finding for major depressive disorder using the Patient Health Questionnaire (PHQ-9): a meta-analysis. General Hospital Psychiatry 37(6), 567–576.CrossRefGoogle ScholarPubMed
Pasco, JA, Moylan, S, Allen, NB, Stuart, AL, Hayley, AC, Byrne, ML and Maes, M (2013) So depression is an inflammatory disease, but where does the inflammation come from? BMC Medicine 11, 200.Google Scholar
R Core Team (n.d.) R: a language and environment for statistical computing. https://www.r-project.org/ (accessed 13 May 2025).Google Scholar
Rubin, DB (1987) Multiple Imputation for Nonresponse in Surveys. New York: John Wiley & Sons.CrossRefGoogle Scholar
Sakakibara, BM, Miller, WC, Orenczuk, SG and Wolfe, DL and the SCIRE Research Team (2009) A systematic review of depression and anxiety measures used with individuals with spinal cord injury. Spinal Cord 47(12), 841–851.CrossRefGoogle ScholarPubMed
Scleroderma Patient-centered Intervention Network (2024) About us. https://www.spinsclero.com/about (accessed 9 December 2024).Google Scholar
Sørensenf, C, Friis-Hasché, E, Haghfelt, T and Bech, P (2005) Postmyocardial infarction mortality in relation to depression: a systematic critical review. Psychotherapy and Psychosomatics 74(2), 69–80.CrossRefGoogle ScholarPubMed
Thau, M, Mikkelsen, MF, Hjortskov, M and Pedersen, MJ (2021) Question order bias revisited: a split-ballot experiment on satisfaction with public services among experienced and professional users. Public Administration 99, 189–204.CrossRefGoogle Scholar
Troubat, R, Barone, P, Leman, S, Desmidt, T, Cressant, A, Atanasova, B, Brizard, B, El Hage, W, Surget, A, Belzung, C and Camus, V (2021) Neuroinflammation and depression: a review. European Journal of Neuroscience 53(1), 151–171.CrossRefGoogle ScholarPubMed
van Buuren, S and Groothuis-Oudshoorn, K (2011) mice: Multivariate Imputation by chained equations in R. Journal of Statistical Software 45(3), 1–67.Google Scholar
van den Hoogen, F, Khanna, D, Fransen, J, Johnson, SR, Baron, M, Tyndall, A, Matucci-Cerinic, M, Naden, RP, Medsger, TA Jr, Carreira, PE, Riemekasten, G, Clements, PJ, Denton, CP, Distler, O, Allanore, Y, Furst, DE, Gabrielli, A, Mayes, MD, van Laar, JM, Seibold, JR, Czirjak, L, Steen, VD, Inanc, M, Kowal-Bielecka, O, Müller-Ladner, U, Valentini, G, Veale, DJ, Vonk, MC, Walker, UA, Chung, L, Collier, DH, Csuka, ME, Fessler, BJ, Guiducci, S, Herrick, A, Hsu, VM, Jimenez, S, Kahaleh, B, Merkel, PA, Sierakowski, S, Silver, RM, Simms, RW, Varga, J and Pope, JE (2013) 2013 classification criteria for systemic sclerosis: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Annals of the Rheumatic Diseases 72(11), 1747–1755.CrossRefGoogle ScholarPubMed
Volkmann, ER, Andréasson, K and Smith, V (2023) Systemic sclerosis. The Lancet 401(10373), 304–318.CrossRefGoogle ScholarPubMed
von Ammon Cavanaugh, S (1995) Depression in the medically ill. Critical issues in diagnostic assessment. Psychosomatics 36(1), 48–59.CrossRefGoogle ScholarPubMed
Wakefield, CE, Butow, PN, Aaronson, NA, Hack, TF, Hulbert-Williams, NJ and Jacobsen, PB and International Psycho-Oncology Society Research Committee (2015) Patient-reported depression measures in cancer: a meta-review. The Lancet Psychiatry 2(7), 635–647.CrossRefGoogle Scholar
Walker, CM (2011) What’s the DIF? Why differential item functioning analyses are an important part of instrument development and validation. Journal of Psychoeducational Assessment 29(4), 364–376.CrossRefGoogle Scholar
Willems, LM, Kwakkenbos, L, Leite, CC, Thombs, BD, van den Hoogen, FH, Maia, AC, Vliet Vlieland, TP and van den Ende, CH (2014) Frequency and impact of disease symptoms experienced by patients with systemic sclerosis from five European countries. Clinical and Experimental Rheumatology 32(6 Suppl 86), S–88–S–93.Google ScholarPubMed
Wu, Y, Levis, B, Riehm, KE, Saadat, N, Levis, AW, Azar, M, Rice, DB, Boruff, J, Cuijpers, P, Gilbody, S, Ioannidis, JPA, Kloda, LA, McMillan, D, Patten, SB, Shrier, I, Ziegelstein, RC, Akena, DH, Arroll, B, Ayalon, L, Baradaran, HR, Baron, M, Bombardier, CH, Butterworth, P, Carter, G, Chagas, MH, Chan, JCN, Cholera, R, Conwell, Y, de Man-van Ginkel, JM, Fann, JR, Fischer, FH, Fung, D, Gelaye, B, Goodyear-Smith, F, Greeno, CG, Hall, BJ, Harrison, PA, Härter, M, Hegerl, U, Hides, L, Hobfoll, SE, Hudson, M, Hyphantis, T, Inagaki, M, Jetté, N, Khamseh, ME, Kiely, KM, Kwan, Y, Lamers, F, Liu, SI, Lotrakul, M, Loureiro, SR, Löwe, B, McGuire, A, Mohd-Sidik, S, Munhoz, TN, Muramatsu, K, Osório, FL, Patel, V, Pence, BW, Persoons, P, Picardi, A, Reuter, K, Rooney, AG, Santos, IS, Shaaban, J, Sidebottom, A, Simning, A, Stafford, L, Sung, S, Tan, PLL, Turner, A, van Weert, HC, White, J, Whooley, MA, Winkley, K, Yamada, M, Benedetti, A and Thombs, BD (2020) Equivalency of the diagnostic accuracy of the PHQ-8 and PHQ-9: a systematic review and individual participant data meta-analysis. Psychological Medicine 50(8), 1368–1380.CrossRefGoogle ScholarPubMed
Figure 0

Figure 1. Participant flow diagram.

Figure 1

Table 1. Participant sociodemographic and disease characteristics at baseline

Figure 2

Table 2. Differences in mean Patient Health Questionnaire-8 somatic item scores between the Reordered Items and Standard PHQ-8 groups

Supplementary material: File

Hu et al. supplementary material

Hu et al. supplementary material
Download Hu et al. supplementary material(File)
File 135.2 KB