Study of Rates and Factors Associated to Psychosomatic Syndromes Assessed Using the Diagnostic Criteria for Psychosomatic Research across Different Clinical Settings Open Access

In the past 10 years, psychosomatic diseases have received increased attention as a leading public health problem. Somatization affects up to 30–40% of internal medicine patients, likely making it the most costly comorbidity [1]. However, their appraisal seems masked by a limited capacity to catch the diagnoses by the standard nosography [2]. Traditional psychiatric classifications such as the DSM have failed to adequately explain the complex interplay between biological and psychological factors in functional somatic syndromes and organic diseases [3, 4]. Although the criteria under “somatic symptom and related disorders” in DSM-5 have undergone significant changes compared to DSM-IV, many psychosomatic symptoms, such as subclinical affective disturbances [5], personality, and behavioral factors, are still not included [6‒8]. On the other hand, the new DSM-5 diagnosis of psychological factors affecting other medical conditions remains vague and lacks specific criteria [9]. DSM is also not well suited to assess the wide range of health behaviors, i.e., how individuals experience, perceive, evaluate, and react to their health status [10‒13].

The Diagnostic Criteria for Psychosomatic Research (DCPR) were proposed by an international group of investigators in 1995 as a complementary tool of assessment with the aim of expanding the traditional domains of the disease model by translating psychosocial variables derived from psychosomatic research into operational tools [14]. DCPR were recently revised [15], with the updated version including also allostatic overload and hypochondriasis. In a variety of clinical settings, DCPR diagnoses were 3.6 times more common than DSM diagnoses [3, 16], and DCPR have proven to be clinically useful in somatic disorders defined by the PFAMC category [17]. The DCPR system has been applied in gastroenterology [18], neurology [19], endocrinology [20], oncology [21], rheumatology [22, 23], cardiology [24], and consultation-liaison psychiatry [25]. They allowed to identify high levels of somatization [26, 27], demoralization [28, 29], allostatic overload [30‒32], and abnormal illness behavior [33], which negatively impact treatment outcomes and health-related quality of life. Alexithymia and persistent somatization, as assessed through the DCPR model, were found as predictors of irritable bowel syndrome (IBS) severity, accounting for 18.5% of the variance [34]. Demoralization identified an additional 20.3% of eating disorder patients not captured by DSM criteria [35]. Additionally, DCPR have more incremental validity than the DSM-5 in explaining an individualʼs psychological states, including well-being, stress, maladaptive illness behavior, and psychological distress (2–24%) [36, 37]. Consequently, there is a necessity to expand the assessment targets of psychiatric evaluation in medically ill patients within the DSM-5 framework [38, 39].

Within this framework, investigating the prevalence rates of DCPR is crucial for obtaining a comprehensive understanding of their occurrence across various clinical settings. Thus, the present research extends this body of work by assessing prevalence in a larger sample. In addition, we examined the multifaceted psychosocial factors influencing DCPR syndromes, an area of research underdeveloped even though it holds significant potential for informing preventive strategies. Patients with fibromyalgia (FM), IBS, migraines, coronary heart disease (CHD), or type 2 diabetes (T2D) were identified as clinical populations in need of been studied since they are often recognized as having psychosomatic components and pose a substantial burden on public health and healthcare systems. We examined DCPR syndromes among subjects with such chronic diseases (i.e., FM, IBS, migraine, CHD, and T2D) to explore (1) DCPR diagnoses prevalence across the five clinical populations, (2) factors associated to a high psychosomatic load (that is, having 3 or more DCPR diagnoses), and (3) psychosocial variables associated with having 3 or more DCPR diagnoses.

One hundred seventy-five general hospitals across China recruited participants aged 18 to 75 years from May 1, 2022, to December 31, 2023 (online suppl. Fig. S1; for all online suppl. material, see https://doi.org/10.1159/000541404). The sample comprised both outpatients (46.6%) and inpatients (53.4%). The five working groups had a diagnosis of FM, IBS, migraine, CHD, or T2D, respectively. Participants were included based on specific diagnostic criteria. FM had to be diagnosed according to the 2016 revisions to the 2010/2011 fibromyalgia diagnostic criteria [40]; IBS according to the Rome IV criteria [41]; migraines by the use of the International Classification of Headache Disorders, 3rd edition (ICHD-3) [42]; the T2D according to the guideline for the prevention and treatment of T2D mellitus (2020 edition) [43]. As what concerns CHD, one of the following diagnoses was made: myocardial infarction; typical angina accompanied by ST-T changes on electrocardiogram during episodes of angina; coronary angiography or coronary CTA demonstrating coronary stenosis >50% or moderate to severe stenosis; coronary stent or balloon dilation and coronary artery bypass grafting. The study exclusion criteria were presence of malignant cancer, occurrence of mental disorders, pregnancy, lactation, cognitive impairment, inability to complete the questionnaires proposed. This nationwide research project received ethical approval from the central institution (Zhongda Hospital, Southeast University, 2021ZDSYLL349-P02). Each participant provided a written consent. All procedures followed the Declaration of Helsinki. The multicenter collaboration employs a hub-and-spoke model.

Two senior internists reached a consensus on the diagnoses of the potentially eligible patients according to the established diagnostic criteria for the five diseases. All participants completed a comprehensive questionnaire with the assistance of trained physicians, nurses, or medical students, documenting sociodemographic characteristics (e.g., age, sex, weight, height, marital status, education, working activity, annual family income, smoking, drinking, and living area) and personal clinical history. Subsequently, psychosomatic assessment was conducted using the Semi-Structured Interview for Diagnostic Criteria in Psychosomatic Research-revised [39]. This is a clinimetric tool [33] focusing on the previous 6–12 months and having 79 items with a yes/no answer. It assesses 14 psychosomatic syndromes (i.e., allostatic overload, type A behavior, alexithymia, hypochondriasis, disease phobia, thanatophobia, health anxiety, persistent somatization, conversion symptoms, anniversary reaction, illness denial, demoralization, irritable mood, secondary somatic symptoms) through four diagnostic modules (i.e., stress, personality, illness behavior, psychological manifestation). Since mental disorders were excluded at enrollment, somatic symptoms secondary to a mental disorder were not assessed. The assessment was completed using self-report scales, as mentioned below. The PsychoSocial Index (PSI), a 55-item sensitive index assessing stress and related psychological distress [44], has 12 items that investigate sociodemographic and clinical data and 43 items exploring 5 domains, i.e., stress (items 13–20 and 22–30), well-being (items 31–36), psychological distress (items 37–51), abnormal illness behavior (items 52–54), and quality of life (item 55). Well-being and quality of life can be merged into a global well-being score. Anxiety was measured with the Generalized Anxiety Disorder 7-Item Scale (GAD-7) [45] within the past 2 weeks. Response options for each item range from 0 (“not at all bothered by the problem”) to 3 (“bothered nearly every day”). Depression was measured with the Patient Health Questionnaire (PHQ-9) [46], a nine-item scale designed to evaluate depressive symptoms within the past 2 weeks. Response options for each item range from 0 (“not at all bothered by the problem”) to 3 (“bothered nearly every day”). Subjective well-being was measured via the World Health Organization-5 Well-Being Index (WHO-5) [47], a short, global rating scale comprising five items that encompass fundamental aspects of an individualʼs perception of overall state of well-being. The Chinese versions of DCPR [48], PSI [44], PHQ-9 [46, 49], GAD-7 [50], and WHO-5 [51] were used.

Upon initial contact through general hospitals or telephone, interested participants received an information package encompassing general project details, informed consent forms, and questionnaires. The minimum requirement for participation entailed the willingness of individuals to complete the questionnaires. During the screening process, participants were examined for malignant tumors, mental disorders, pregnancy, or lactation to exclude them. This information was collected by self-reporting or medical records. Thereafter, clinical interviews of eligible patients were conducted via an online assessment system. If participants’ age or level of education precluded using electronic devices for evaluation, paper-based instruments were employed and the data were digitized upon completion.

Nationwide, 175 branch center leaders and research assistants underwent training for the project. This training encompassed practicing with the online assessment system and monitoring data quality. For the DCPR, branch center leaders and research assistants were required to attend interview sessions led by trained interviewers. Regarding quality control, the primary center conducts weekly reviews to ensure the accuracy and authenticity of data entry, such as examining inconsistencies in responses or missing scales, regularly monitoring the frequency and patterns of missing data, and addressing potential issues that may lead to incomplete or omitted data. Every 1–2 months, meetings were held to discuss contentious cases or summarize research while maintaining consensus on DCPR diagnoses. Furthermore, to standardize DCPR assessment as much as possible, the main center developed operation manuals and demonstration videos. Consistency evaluations were conducted using simulated patient videos (including a male or a female subject), with inter-rater agreement demonstrated by intraclass correlation coefficients ranging from 0.907 to 1.000.

Collaboratively developed with software engineering and computer science experts, the online assessment system incorporates the input of five working group leaders and users’ feedback to ensure compatibility with authentic clinical settings. Accessible via electronic devices by scanning QR codes or visiting designated URLs, the platform grants various access levels based on user roles, such as project leaders, working group heads, and researchers. This hierarchical access structure safeguards sensitive data by restricting access to authorized personnel only. The system comprises distinct interfaces for clinicians/researchers and patients, allowing patient data to be input through the patient interface. Simultaneously, each participant is assigned an anonymous identification code. The clinician/researcher interface allows working group leaders to access comprehensive participant information within their specific subcenters. Moreover, the platform implements data validation rules to preclude the entry of logically implausible or improbable values. Digital data were securely uploaded to a central storage server, backed up to tape daily, and archived in at least two locations.

Continuous data are presented as means ± SDs Categorical variables are presented as rates. Multivariable binary logistic regressions were used to evaluate associations of the condition of having 3 or more DCPR among the following psychosocial variables: PHQ-9, GAD-7, WHO-5, PSI. The aim is to compare those with higher psychosomatic load with those with lower one as done in other studies [52]. The analyses were adjusted for age, sex, education, BMI, marital status, working activity, annual family income, smoking status, drinking status, rural residential area. Subsequently, a binary logistic regression model was established for all patients, incorporating the above covariates along with disease type, categorized into five classes: FM, IBS, migraine, CHD, and T2D. Dummy variables were created for each disease type and included in the logistic regression analysis. The advantage of this approach is that multiple disease types and covariates can be considered simultaneously, providing a more comprehensive risk assessment. Collinearity for each covariate in the logistic models was evaluated using the variance inflation factors; there was no evidence of collinearity. The Hosmer-Lemeshow goodness-of-fit test in the multiple logistic regression was performed to assess the model calibration by R package ResourceSelection (version 0.3-6), and calibration curves were plotted.

To examine a possible nonlinear relation between DCPR ≥3 diagnoses and PSI, PHQ-9, or GAD-7 scores, adjusting for age, sex, education, BMI, marital status, working activity, annual family income, smoking status, drinking status, rural residential area, we used restricted cubic splines regression with 4 knots using the rms package (version 6.8-0, https://hbiostat.org/R/rms/) and ggrcs package (version 0.3.8) [53]. When nonlinearity was detected, a recursion algorithm was used to calculate the inflection point of the relationship between the respective variables and DCPR ≥3 diagnoses, respectively, and a two-segment binary logistic regression model was performed on both sides of the inflection point.

Analyses were performed with R, version 4.3.3. p values were two-sided, and a p value <0.05 was considered significant.

The sample of 7,775 participants included 6,647 (85.5%) with complete data, covering five diseases (online suppl. Table S1). Missing data were sporadic and evenly distributed across populations. The most common diagnoses were alexithymia (64.47%), irritable mood (20.55%), and demoralization (15.60%). The FM group showed the highest rate of demoralization at 49.02%, whereas alexithymia was more prevalent in other clinical groups and the entire sample, with rates between 52.48% and 70.47% (Table 1; online suppl. Fig. S2).

The multivariable logistic regression analyses revealed that higher anxiety or abnormal illness behavior, and poorer global well-being were associated with having at least three DCPR diagnoses in patients with FM, IBS, migraine, CHD, or T2D. Unique risk factors varied among these groups. Stress (FM: OR 1.24, 95% CI: 1.06–1.46, p = 0.009; T2D: OR 1.26, 95% CI: 1.18–1.36, p < 0.001) was significantly associated to DCPR diagnoses in FM and T2Ds (Table 2). Shared factors, including anxiety (GAD-7 scores: OR, 1.14, 95% CI: 1.11–1.17; p < 0.01), abnormal illness behavior (OR: 1.27, 95% CI: 1.20–1.35; p < 0.01), and global well-being (OR: 0.7, 95% CI: 0.66–0.74; p < 0.01) in the whole sample were observed (Table 3). Patients with FM, IBS, and migraine showed to be more prone to have DCPR diagnoses than those with CHD or T2D (online suppl. Table S2). The calibration curves for various patient groups are depicted in online supplementary Figure S3.

In online supplementary Figure S4, we used restricted cubic splines to flexibly model and visualize the relationship between various psychosocial scores and the prevalence of having at least three DCPR diagnoses. From online supplementary Figure S4A, as stress scores increase, the risk of having at least three DCPR diagnoses rises (OR = 1.73 for stress ≤5, p for nonlinearity <0.001, 95% CI: 1.65–1.82); however, no significant correlation is observed when the stress scores reach 5 (p = 0.456). The risk increases steadily with higher PHQ-9 scores, peaking at scores below 11 (OR = 1.39 for PHQ-9 <11, 95% CI: 1.35–1.43) and moderates slightly at higher scores (OR = 1.16 for PHQ-9 ≥11, p for nonlinearity <0.001, 95% CI: 1.12–1.20) (online suppl. Fig. S4B). A similar pattern is observed with GAD-7 scores (online suppl. Fig. S4C), where the inflection point is at 7 (OR = 1.53 for GAD-7 <11, p for nonlinearity <0.001, 95% CI: 1.48–1.60). A significant increase in risk is noted for abnormal illness behavior scores <3 (OR = 2.30, p for nonlinearity <0.001, 95% CI: 2.16–2.50), with a diminishing effect as scores exceed this threshold (p for nonlinearity <0.001) (online suppl. Fig. S4D).

A significant psychosomatic burden was found across five clinical populations affected by chronic conditions. Alexithymia, irritable mood, and demoralization were the most prevalent psychosomatic syndromes, with demoralization being most common in FM. Factors specifically associated to DCPR diagnoses were higher anxiety or abnormal illness behavior and poorer global well-being, with distinctive factors such as stress in FM and T2D. Patients with FM, IBS, and migraine were more prone to have DCPR diagnoses compared to those with CHD or T2D. Increasing stress, depression, anxiety, and abnormal illness behavior were significantly associated with higher odds of DCPR risk, with specific thresholds where the effects plateaued or diminished.

This nationwide study offered a unique dataset for examining the prevalence and clinical impact of DCPR syndromes. Some individuals develop alexithymia following traumatic events or physical illness, aligning with the DCPR view of alexithymia as a state condition. Extensive research has established a link between alexithymia and various physical illnesses. The prevalence of alexithymia estimated in our study was higher than that observed in other countries (FM: 33.7–38.2% [22, 54]; CHD: 37%) [55]. One explanation might be rooted in the cultural context of China. Traditional Chinese culture often emphasizes emotional restraint and collectivism, which can discourage expressing and recognizing of personal emotions [56, 57]. Another possible explanation may be the rapid socioeconomic changes in China over the past few decades that have introduced significant stressors, including urbanization, occupational pressures, lifestyle changes, which challenged mental health [58‒60]. Irritable mood was prevalent in our sample. Substantial evidence suggests a pathogenic role for anger, hostility, and irritable mood in physical diseases [61], particularly cardiovascular ones [62, 63]. Irritable mood is related to increased gray matter volume in the orbitofrontal cortex [64] and heightened activity in the amygdala and hypothalamus [65] as well as to increased monoamine transmission [66, 67], contributing to its connection with anxiety and depression. Using DCPR to identify irritable mood can assist in neurological research and treatment development, potentially predicting response to certain medications [68].

The distinctive clinical symptoms of FM differentiate it from other groups, where alexithymia is the predominant DCPR diagnosis. Consistently with previous research, about 50% of FM patients experienced demoralization [22]. The core symptom of FM, chronic widespread pain, often leads to frustration, helplessness, and hopelessness [69, 70]. The unpredictability of pain and the perceived lack of control over its management may also exacerbate these negative emotional states, resulting in demoralization. Subjective incompetence is considered a clinical hallmark of demoralization [29, 71], and when hopelessness and helplessness become pervasive and persistent, they may evolve into clinical depression that can, in turn, amplify the perception of pain [72]. As previous research has found, demoralization as a response to chronic stress is associated with various features, including allostatic overload [73]. It also diminishes quality of life, mediated by depressive symptoms [74], and is linked to poor euthymia [35], which include well-being [75].

The present findings showed a significant association between anxiety and abnormal illness behavior and an elevated risk of psychosomatic symptoms. These results build upon existing evidence that anxiety frequently results in heightened awareness and negative interpretation of bodily sensations [76] and that abnormal illness behavior intensifies frustration with medical experiences [77, 78], especially in patients presenting with three or more DCPR diagnoses (OR = 2.3). In contrast, well-being showed a protective role against DCPR diagnoses. Higher well-being is likely to bolster resilience to stress [79], thereby mitigating the risk of psychosomatic symptoms. This finding underscores the importance of incorporating the assessment of positive emotions and well-being into clinical evaluations, particularly under the light of the emerging evidence on interventions promoting well-being, such as well-being therapy [39, 80], which ultimately fosters a balance among psychic forces, resilience, and euthymia [75].

Distinctive factors associated with DCPR diagnoses revealed the complex interplay with chronic conditions. Prior research has noted a notable prevalence of psychosomatic syndromes in individuals with FM [22, 54]. Our study builds upon this by demonstrating that stress plays a significant role in the likelihood of individuals with FM receiving a DCPR diagnosis, particularly in those with three or more diagnoses compared to those with fewer. Chronic stress may worsen pain sensitivity and psychological symptoms by influencing the neuroendocrine system, such as the HPA axis, and immune responses [81]. Furthermore, individuals with FM demonstrate impairment in the default mode network and the amygdala [82], mirroring observations in posttraumatic stress disorder [83, 84]. These findings are consistent with our investigation, which revealed a 33% prevalence of allostatic overload, signifying that (chronic) stress may surpass the body ability to cope, exacerbating symptoms.

The large transdiagnostic sample findings showed a pronounced proneness to psychosomatic symptoms among patients with FM, IBS, or migraine compared to those with CHD or T2D. This phenomenon can be attributed to several underlying factors. First, the pathophysiological mechanisms of FM, IBS, and migraine are closely linked to central sensitization and heightened stress responses [85‒88]. Second, the chronic pain and discomfort characteristic of FM, IBS, and migraines can lead to heightened awareness and reporting of psychosomatic symptoms. This is consistent with earlier studies linking functional disorders to higher psychological distress [89, 90].

Several studies demonstrated that stress, abnormal illness behavior, depression, and anxiety exacerbate psychosomatic symptoms in chronic disease patients. The present results extend beyond the evidence that increased stress significantly correlates with higher DCPR risk up to a score of 5 (OR = 1.73). Patients with at least three DCPR diagnoses showed to be 2.3 times more likely to exhibit abnormal illness behavior within the 0–3 score range than those with less than three DCPR diagnoses. These thresholds guide the creation of focused interventions that enhance resource allocation according to clinical attributes.

The mentioned results should be considered under the light of several limitations. The DCPR assessment was done at 175 subcenters. Therefore, the possibility of variation between subcenters exists. However, all participating subcenters completed a standardized assessment process before the study, and stringent quality control procedures were used during the study. Second, the patients were recruited at tertiary level clinics and may, therefore, represent the most severe groups. Female subjects in FM, IBS, and migraine were overrepresented, but such sex imbalance mirrors the clinical realm that is characterized by higher prevalence and incidence of these diseases among females than males [91‒93]. Additionally, exclusion criteria were identified via self-reports or medical records, potentially underestimating mental disorders and cognitive impairment. Finally, the cross-sectional design of the study makes any inference of a causal relationship between DCPR syndromes and predictors weak.

The present study proved the clinical utility of running an assessment via the use of DCPR in diverse clinical settings. In addition, it suggests that anxiety, abnormal illness behavior, and well-being should be recognized as factors associated with DCPR diagnoses. At a clinical level, identifying the DCPR diagnoses and associated factors, which may impact treatment outcomes, can enhance cliniciansʼ understanding of patientsʼ characteristics. This knowledge can aid in identifying care paths for patients who need complex interventions under the light of the biopsychosocial model [94]. At a research level, investigating analogous phenomena to prevent ambiguity and identifying key factors of DCPR diagnoses can facilitate the advancement of innovative interventions tailored to address the vulnerabilities associated with each psychosomatic condition.

We thank the Chinese Psychosomatic Medicine Association, the collaborators, and the participants for their time, support, and effort. Special thanks are extended to the DCPRs-C Working Group for their significant contributions to the investigation and data collection. A detailed list of its members is in the online supplementary material.

This nationwide research project received ethical approval from the central institution (Zhongda Hospital, Southeast University, 2021ZDSYLL349-P02). Each participant provided a written informed consent.

The authors have no conflicts of interest to declare.

This project was partially supported by the National Natural Science Foundation of China (Grant No. 81971277, Yonggui Yuan) and the Jiangsu Provincial Key Research and Development Program (Grant No. BE2019748, Yonggui Yuan).

W.X.: project administration, data curation, formal analysis, methodology, writing – original draft, and writing – review and editing; W.H.J.: project administration, formal analysis, methodology, and writing – review and editing; Y.Y.Y.: conceptualization, methodology, and resources; R.J.D., H.T., Y.Y.W., Y.P.T., D.F.L., Y.P.W., and M.W.W.: project administration, methodology, supervision, and investigation; B.W.C.: methodology and formal analysis; Y.Y.K.: resources and software; L.L.T. and L.Y.: investigation and data curation; F.C.: conceptualization, project administration, methodology, formal analysis, supervision, and writing – review and editing; Y.G.Y.: conceptualization, project administration, funding acquisition, methodology, supervision, and writing – review and editing; and on behalf of the DCPRs-C Working Group: investigation and data curation.

Wei Xu and Wenhao Jiang should be considered as the co-first authors to this work.Fiammetta Cosci and Yonggui Yuan should be considered as the co-senior authors to this work.

The datasets generated and/or analyzed for the current study contain clinical data and are not publicly available due to the protection of participants’ rights to privacy and data protection. Further inquiries can be directed to the corresponding author.