Integrated Care for Psychosomatic Symptoms: An Illustrative Analysis Using Synthetic Outpatient Data

Full Text

Psychosomatic disorders represent an important area of clinical medicine in which physical symptoms, psychological processes, and social circumstances interact to influence health and functioning. These interactions are relevant across specialties because patients frequently present with persistent symptoms that cannot be adequately understood through structural pathology alone. However, psychosomatic illness should not be interpreted as imaginary disease or as evidence that physical complaints lack biological foundations. Rather, a biopsychosocial perspective recognizes that bodily experiences may emerge from multiple interacting mechanisms, including established medical conditions, altered physiological regulation, emotional distress, and behavioral responses to symptoms.

The terminology used to describe these presentations remains heterogeneous. Psychosomatic disorders, somatic symptom disorder, functional somatic syndromes, and persistent physical symptoms refer to overlapping but nonidentical concepts. Some classifications emphasize symptom patterns, whereas others focus on distress, functional impairment, or disproportionate symptom-related thoughts and behaviors. Importantly, the presence of medically explained symptoms does not exclude clinically significant psychological contributions, and the absence of an identified organic cause does not establish a psychiatric diagnosis. Clear operational definitions are therefore essential when investigating psychosomatic presentations and interpreting their clinical significance.

Several pathways may contribute to the relationship between psychological distress and physical symptom burden. Sustained stress can influence autonomic activity, neuroendocrine regulation, sleep, and pain processing. Anxiety may increase attention to bodily sensations and threat-related interpretations, while depressive symptoms may affect motivation, activity levels, and perceived capacity to manage illness. These processes can coexist with inflammation, tissue injury, or other disease-specific mechanisms. Nevertheless, their relative importance likely differs between individuals, and observational associations cannot determine whether psychological symptoms precede, follow, or reciprocally reinforce physical complaints.

The clinical consequences of persistent psychosomatic symptoms extend beyond symptom intensity. Patients may experience limitations in work, family responsibilities, social participation, and everyday activities. Repeated consultations and investigations can occur when symptoms remain distressing or explanations are fragmented. Conversely, premature attribution of symptoms to psychological causes may undermine trust and delay recognition of treatable medical disease. Effective assessment therefore requires a balanced approach that evaluates relevant physical pathology while also considering emotional distress, sleep disturbance, coping patterns, and the patient’s understanding of illness.

Primary care offers a particularly relevant setting for studying these relationships because clinicians encounter undifferentiated symptoms, chronic medical conditions, and psychological concerns within the same consultation. However, time constraints and variable access to mental health services can limit comprehensive assessment. Patients may also be reluctant to discuss emotional difficulties because of stigma or concern that their symptoms will be dismissed. Standardized patient-reported measures may help characterize symptom burden and associated distress, although screening instruments cannot replace clinical evaluation or establish the cause of an individual patient’s symptoms.

Despite growing recognition of biopsychosocial care, uncertainty remains regarding how psychological distress relates to physical symptom severity and functional impairment in routine outpatient populations. Interpretation is complicated by differences in diagnostic definitions, recruitment settings, and measurement approaches. Concurrent medical illness, demographic characteristics, sleep quality, and socioeconomic circumstances may also influence observed relationships. Studies that assess these domains together can provide a clearer description of clinically relevant patterns without assuming that psychological factors are the sole explanation for persistent symptoms.

The present cross-sectional study aims to examine associations between psychological distress, somatic symptom burden, and functional impairment among adults attending primary care. Anxiety, depressive symptoms, perceived stress, and sleep quality will be considered alongside demographic characteristics and documented chronic medical conditions. We hypothesize that greater psychological distress will be associated with higher somatic symptom burden and poorer functioning after adjustment for selected covariates. By identifying patterns of co-occurring difficulties, the study seeks to inform more comprehensive clinical assessment while acknowledging that its observational design cannot establish causality or determine the direction of these relationships.

2. Materials and Methods

A prospective observational cohort study will investigate associations between psychological distress, somatic symptom burden, and functional impairment among adults attending primary care and general internal medicine clinics. Participants will be assessed at enrollment, three months, and six months. The study will use a biopsychosocial framework, recognizing that psychological processes may influence physical symptoms without implying that symptoms are fabricated or lack biological contributors. The protocol will require institutional ethics approval before recruitment, and all participants will provide written informed consent. Reporting will follow established guidance for observational cohort studies.

Eligible participants will be adults aged 18–65 years who report one or more distressing physical symptoms persisting or recurring for at least three months. Presentations may include gastrointestinal discomfort, headache, musculoskeletal pain, palpitations, fatigue, or dizziness. Symptoms will not need to be medically unexplained, and participants with stable chronic medical conditions will remain eligible. Exclusion criteria will include acute medical instability, delirium, cognitive impairment preventing informed consent, current psychosis or mania, and inability to complete assessments in an available validated language. Pregnancy-related symptoms will be excluded because their trajectories may require a separate analytical framework.

Consecutive potentially eligible patients will be approached during designated recruitment sessions to reduce discretionary selection. Screening logs will record eligibility, participation decisions, and reasons for exclusion without retaining unnecessary identifiers. A target of 240 participants will allow for approximately 20% attrition, leaving about 192 participants with six-month data. This target is intended to support estimation of a modest association between baseline psychological distress and subsequent symptom burden using a parsimonious adjusted model. Because effect-size assumptions are uncertain, confidence intervals and estimation precision will receive greater emphasis than statistical significance.

At enrollment, a study clinician will review presenting symptoms, medical history, medication use, previous investigations, and relevant examination findings. Additional investigations will be ordered only when clinically indicated, rather than to establish that symptoms are unexplained. Red-flag findings, including unexplained weight loss, gastrointestinal bleeding, focal neurological deficits, or exertional chest pain, will prompt appropriate clinical evaluation. Participation will not replace usual care. Physical disease and psychological distress will be documented separately, and neither will be presumed to exclude the other.

Somatic symptom burden will be measured using the Patient Health Questionnaire–15. Psychological distress will be assessed using the Hospital Anxiety and Depression Scale, selected to limit overlap with physical symptom items. Its anxiety and depression subscales will be retained separately in secondary analyses. Excessive symptom-related thoughts, feelings, and behaviors will be assessed with the Somatic Symptom Disorder–B Criteria Scale. Questionnaire thresholds will indicate elevated symptoms, not establish psychiatric diagnoses. A trained clinician will conduct a standardized diagnostic interview when somatic symptom disorder is suspected, applying diagnostic criteria without requiring the absence of medical disease.

Functional impairment will be evaluated using the 12-item World Health Organization Disability Assessment Schedule. Participants will also report sleep duration, smoking, alcohol use, physical activity, employment, educational attainment, and major stressful events during the preceding six months. Medical comorbidity will be summarized from documented diagnoses, while medication exposure will include analgesics, antidepressants, anxiolytics, and relevant disease-specific treatments. Healthcare utilization will be recorded as consultations, emergency visits, and hospital admissions, verified against available records with consent. Only validated language versions of instruments will be administered.

Follow-up assessments will repeat symptom, distress, and disability measures using the same administration mode whenever feasible. Treatment changes, new diagnoses, and significant life events will be recorded. Research staff will receive standardized training in neutral interviewing and questionnaire administration. Data will be entered into a secure electronic database with range checks, coded identifiers, and restricted access. A random 10% of records will undergo source verification. Participants reporting severe distress or safety concerns will receive timely clinical assessment according to a predefined referral pathway.

The primary outcome will be somatic symptom burden at six months, analyzed in relation to baseline psychological distress. A linear regression model will adjust for baseline somatic symptom burden, age, sex, medical comorbidity, and baseline psychotropic medication use. Covariates will be selected before analysis on clinical grounds rather than univariable significance. Secondary analyses will examine anxiety and depression separately, six-month disability, and symptom trajectories using linear mixed-effects models with participant-specific random intercepts. These models will treat time categorically and include a baseline distress-by-time interaction. Residual distributions, influential observations, and collinearity will be examined; robust standard errors will be used when appropriate.

Missing questionnaire items will be handled according to instrument scoring instructions. Missing outcome and covariate data will be described by extent and pattern. If appropriate, multiple imputation by chained equations will incorporate baseline variables, repeated outcomes, and predictors of loss to follow-up. Complete-case analyses will provide a sensitivity comparison. Additional sensitivity analyses will exclude participants receiving new major medical diagnoses during follow-up. Estimates will be reported with 95% confidence intervals and two-sided P values, with secondary findings interpreted as exploratory. Given the observational design, associations will not be interpreted as evidence that psychological distress causes physical symptoms.

In this explicitly simulated prospective cohort, 240 adults with persistent somatic symptoms associated with psychological distress were followed for 12 weeks. The numerical findings represent a hypothetical dataset rather than observations from an actual clinical investigation. Participants received either integrated psychosomatic care, combining routine medical management with structured psychological support, or routine medical care alone, with 120 participants in each group. Treatment allocation was not randomized. Of 286 individuals assessed for eligibility, 29 did not meet the predefined symptom and duration criteria, 11 declined participation, and six could not complete the baseline assessment. Eligibility required clinically significant somatic symptoms lasting at least six months, associated functional impairment, and a medical assessment addressing potential underlying disease.

The mean participant age was 39.7 years (standard deviation [SD], 11.8), and 156 participants (65.0%) were women. Median symptom duration was 19 months (interquartile range [IQR], 10–36). The most frequently reported complaints were fatigue in 169 participants (70.4%), musculoskeletal pain in 151 (62.9%), gastrointestinal discomfort in 128 (53.3%), and palpitations or chest discomfort in 91 (37.9%). Symptoms commonly overlapped, with 172 participants (71.7%) reporting complaints involving at least two bodily systems. These symptom categories were not interpreted as evidence that psychological processes were the sole cause of the reported physical experiences.

Baseline demographic and clinical characteristics were broadly similar between care groups, although potentially relevant differences remained. Participants receiving integrated care were slightly younger than those receiving routine care, with mean ages of 38.9 and 40.5 years, respectively. Women accounted for 67.5% and 62.5% of the groups. Median symptom duration was 18 months in the integrated-care group and 20 months in the routine-care group. Baseline somatic symptom severity, anxiety, depressive symptoms, and functional impairment were modestly higher in the integrated-care group, as shown in Table 1. These differences were retained as potential confounders rather than dismissed on the basis of baseline significance testing.

At enrollment, 74 participants (30.8%) had a documented stable medical comorbidity, most commonly treated hypertension, thyroid disease, or asthma. The proportions were 29.2% in integrated care and 32.5% in routine care. Ninety-three participants (38.8%) were taking an antidepressant, without a substantial difference between groups. A baseline Generalized Anxiety Disorder-7 score of at least 10 was recorded in 139 participants (57.9%), while 127 (52.9%) had a Patient Health Questionnaire-9 score of at least 10. These screening thresholds indicated elevated symptom burden and were not treated as equivalent to independently confirmed psychiatric diagnoses.

Week-12 assessments were available for 213 participants (88.8%): 108 receiving integrated care and 105 receiving routine care. Among the 27 participants without follow-up assessments, 15 could not be contacted, eight reported scheduling difficulties, and four withdrew consent. Participants without follow-up had slightly higher baseline somatic symptom scores than completers, averaging 15.5 versus 14.5, respectively. Missingness therefore could not be assumed to be entirely unrelated to symptom burden. Within integrated care, participants attended a median of six psychological support sessions (IQR, 4–8), and 91 of 120 (75.8%) attended at least four sessions.

Somatic symptom scores decreased in both groups, with a larger observed reduction among participants receiving integrated care. Table 1 presents descriptive baseline and week-12 scores; baseline summaries include all enrolled participants, whereas week-12 summaries include those assessed at follow-up.

In the primary regression analysis, the adjusted week-12 Patient Health Questionnaire-15 score was 2.4 points lower in integrated care than in routine care (95% confidence interval [CI], −3.4 to −1.4). The model included baseline symptom severity, age, sex, symptom duration, stable medical comorbidity, and baseline antidepressant use. Multiple imputation was used for missing follow-up outcomes. The complete-case estimate was similar at −2.5 points (95% CI, −3.6 to −1.4). In a sensitivity analysis assigning imputed integrated-care outcomes an additional two symptom points, the estimated difference was attenuated to −2.2 points (95% CI, −3.3 to −1.1).

Among participants with follow-up data, 49 of 108 receiving integrated care (45.4%) and 25 of 105 receiving routine care (23.8%) achieved a prespecified reduction of at least five points on the somatic symptom scale. The unadjusted absolute difference was 21.6 percentage points (95% CI, 9.1–34.0). This responder definition provided a complementary description of improvement but did not establish symptom resolution or recovery. Substantial residual symptoms remained in both groups at week 12.

Secondary outcomes showed a similar direction of association. Adjusted between-group differences favored integrated care for anxiety symptoms by 1.6 points (95% CI, −2.7 to −0.5), depressive symptoms by 1.5 points (95% CI, −2.7 to −0.3), and functional impairment by 2.7 points (95% CI, −4.2 to −1.2). These secondary comparisons were exploratory and were not corrected for multiple testing. Change in somatic symptom severity correlated with change in anxiety severity among completers (Pearson r = 0.46; 95% CI, 0.35–0.56), without establishing temporal direction or mediation.

Exploratory interaction analyses did not provide clear evidence that the association between care pathway and somatic symptom improvement differed by sex or the presence of stable medical comorbidity. Estimates within these subgroups were imprecise. Within integrated care, attending at least four sessions was associated with greater improvement, but this comparison remained susceptible to differences in motivation, access, and baseline prognosis.

No serious adverse event was recorded as attributable to psychological support. Seven integrated-care participants reported transient emotional discomfort during sessions; two discontinued psychological sessions while remaining in medical follow-up. Medication initiation or dose adjustment occurred in 18 integrated-care participants and 21 routine-care participants. Overall, the simulated findings showed greater short-term improvement associated with integrated care, while nonrandom allocation, incomplete follow-up, and concurrent treatment changes limited causal interpretation.

The clinical significance of psychosomatic disorders lies in the interaction between bodily symptoms, emotional distress, and impaired daily functioning. These relationships should be interpreted within a biopsychosocial framework rather than through a strict division between physical and psychological disease. Psychological distress may amplify symptom perception and disability, but its presence does not establish that symptoms are exclusively psychological in origin. Conversely, persistent physical symptoms can contribute to anxiety, low mood, and reduced social participation. This bidirectional relationship is central to interpreting associations between psychological measures and somatic symptom burden without assigning causal priority to either component.

An important distinction concerns symptom severity versus the processes that maintain symptom-related distress. Patients with similar physical complaints may differ substantially in their illness beliefs, coping strategies, sleep quality, and functional limitations. Consequently, symptom counts alone may provide an incomplete representation of clinical burden. Measures of interference with work, relationships, and routine activities can add clinically meaningful information. Interpretation also requires attention to diagnostic terminology: psychosomatic presentations are heterogeneous and should not automatically be equated with somatic symptom disorder, functional neurological disorder, or symptoms without an established medical explanation. Psychological and behavioral factors may influence illness experience even when an identifiable medical disease is present.

Several mechanisms could plausibly connect sustained emotional distress with physical symptom burden. Altered autonomic regulation, stress-related neuroendocrine activity, disturbed sleep, and heightened sensitivity to bodily sensations may contribute to symptom persistence in some patients. Cognitive and behavioral processes, including threat-focused interpretation, repeated checking, avoidance, and reassurance seeking, may further reinforce disability. Nevertheless, these pathways should be regarded as explanatory hypotheses unless directly assessed. Associations based on questionnaire data cannot demonstrate autonomic dysfunction, inflammatory activation, or altered central pain processing. Such mechanisms may coexist, differ across symptom groups, or reflect consequences rather than antecedents of chronic illness.

Sleep disturbance deserves particular attention because it can link emotional symptoms, pain, fatigue, and reduced coping capacity. Poor sleep may increase the salience of bodily discomfort, while persistent symptoms and illness-related worry may disrupt sleep initiation and continuity. This reciprocal pattern complicates the interpretation of sleep as an independent predictor. Similarly, anxiety and depressive symptoms frequently overlap with somatic questionnaire items, including fatigue, appetite change, and concentration difficulties. Apparent associations may therefore partly reflect shared measurement content rather than distinct clinical processes. Analyses separating overlapping items, where appropriate, would help clarify whether relationships extend beyond this source of correlation.

The social context of psychosomatic symptoms is also clinically relevant. Financial insecurity, occupational strain, caregiving responsibilities, and limited support may increase both distress and barriers to recovery. These exposures should not be treated merely as background characteristics when they plausibly influence symptom reporting, healthcare access, and treatment adherence. Cultural expectations can additionally shape how patients describe suffering and whether they consider psychological care acceptable. An emphasis on bodily complaints may represent a culturally familiar form of communicating distress rather than resistance to psychological explanation. Clinicians should therefore explore patients’ explanatory models respectfully and avoid interpreting communication differences as evidence of poor insight.

These considerations support a clinical approach that combines appropriate medical assessment with attention to emotional and behavioral contributors. Repeated investigation without a clear indication may expose patients to incidental findings, unnecessary procedures, and greater uncertainty; however, premature psychological attribution carries a different risk by delaying recognition of medical disease. A proportionate approach includes assessment for warning signs, review of existing diagnoses and medications, and planned reassessment when symptoms change. Validation of the patient’s experience is essential throughout this process. Explaining that symptoms are genuine and may be influenced by interacting physiological, psychological, and social factors can create a shared basis for care without implying that the symptoms are imagined or voluntarily produced.

The clinical implications of these findings should be considered within a biopsychosocial framework rather than interpreted as evidence that persistent physical symptoms are exclusively psychological. In psychosomatic presentations, emotional distress, physiological arousal, illness beliefs, and social circumstances may interact with established medical disease or contribute to symptoms without a fully explanatory structural abnormality. This distinction is important because attributing symptoms prematurely to stress can undermine trust and delay appropriate investigation. A clinically useful interpretation therefore recognizes the reality of patients’ symptoms while maintaining diagnostic openness and evaluating potentially reversible medical contributors.

An integrated assessment may be more informative than a narrow focus on either somatic complaints or psychological distress. Alongside symptom history and physical examination, clinicians can explore sleep quality, recent stressors, functional impairment, medication use, and the patient’s understanding of illness. Screening for anxiety and depressive symptoms may identify additional treatment needs, but screening results should not be treated as proof of a psychosomatic cause. Similarly, the presence of an identifiable medical condition does not exclude clinically meaningful psychological influences on symptom severity, coping, or recovery. Reassessment remains essential when symptoms change or new warning signs emerge.

Communication is likely to influence whether this approach is acceptable to patients. Explanations that connect stress responses with bodily processes may reduce the perceived divide between physical and mental health, provided that they are individualized and presented without certainty about mechanisms that were not directly measured. Shared decisions should emphasize achievable goals, such as improving daily functioning, sleep, and participation in valued activities, rather than promising complete symptom elimination. Where appropriate, coordinated care involving primary care clinicians, relevant medical specialists, and mental health professionals may reduce fragmented management. However, the present study cannot establish whether such coordination improves outcomes or reduces healthcare use.

Several limitations constrain interpretation. Associations between psychological measures and somatic symptom burden do not establish directionality: distress may intensify symptoms, persistent symptoms may increase distress, or both may reflect shared underlying factors. Self-reported measures can also introduce recall bias and shared-method effects, particularly when questionnaires contain overlapping physical and emotional items. Even with statistical adjustment, residual confounding by disease severity, sleep disturbance, socioeconomic adversity, treatment exposure, or unmeasured psychiatric conditions may remain. Consequently, observed relationships should not be interpreted as independent causal pathways without further evidence.

Diagnostic heterogeneity is another important consideration. The term psychosomatic encompasses diverse clinical presentations and does not describe a single mechanism or uniformly defined diagnostic category. Patients with predominantly pain-related symptoms may differ substantially from those presenting with gastrointestinal, cardiopulmonary, or fatigue-related complaints. Patterns observed in one clinical population may therefore not generalize to other settings or symptom groups. Cultural expectations, stigma surrounding mental illness, and access to specialist care can also affect symptom reporting and help-seeking. These influences warrant particular attention when applying findings beyond the population studied.

Future research should prioritize longitudinal designs with repeated assessments of symptoms, distress, functioning, and treatment exposure. Such studies could clarify temporal relationships and distinguish persistent vulnerability from responses to illness. Where feasible, combining patient-reported outcomes with clinical assessments and appropriately selected physiological measures may strengthen interpretation, although no single biomarker should be assumed to identify psychosomatic illness. Intervention studies should evaluate both benefits and potential harms, including diagnostic overshadowing, treatment burden, and patients’ perceptions of invalidation. Overall, the findings support attention to psychological and social dimensions of physical symptoms, but not the replacement of medical evaluation with psychological explanations. The most defensible clinical implication is a balanced, patient-centered approach that addresses distress and disability while preserving diagnostic vigilance and continuity of care.

The findings of this study suggest that psychosomatic symptom burden is associated with a complex interaction of psychological distress, sleep disturbance, and limitations in daily functioning. Higher levels of anxiety and depressive symptoms were accompanied by greater perceived physical symptom severity, supporting a biopsychosocial interpretation rather than a strict separation between physical and mental health. These associations should not be interpreted as evidence that symptoms are imagined, exclusively psychological, or independent of underlying medical disease. Instead, they indicate that emotional and behavioral factors may contribute to the experience and persistence of physical symptoms alongside biological and social influences.

An important clinical implication is the value of assessing psychological well-being and functional impairment during the evaluation of persistent somatic complaints. A patient-centered approach should combine appropriate medical investigation with sensitive inquiry into stress, sleep, coping strategies, and relevant social circumstances. Such assessment may help identify potentially modifiable contributors to symptom burden while maintaining vigilance for coexisting or newly emerging organic pathology. Communication that validates patients’ experiences and explains the interaction between bodily processes and psychological stress may also support therapeutic engagement, although its effects on clinical outcomes require direct evaluation.

The observed relationships provide a rationale for considering coordinated care involving primary care clinicians, relevant medical specialists, and mental health professionals. However, the present findings do not establish that any specific intervention reduces symptoms or improves quality of life. Psychological therapies, sleep-focused interventions, and treatment of comorbid psychiatric conditions should be considered according to individual clinical indications and existing evidence, rather than inferred to be effective from associations alone. Care plans should also reflect patients’ preferences, cultural understandings of illness, and access to services.

Several limitations constrain the interpretation of these results. The observational design limits causal inference, and reliance on self-reported measures may introduce recall bias and shared measurement effects. Residual confounding, including differences in medical comorbidity, medication use, and socioeconomic circumstances, cannot be excluded. The extent to which these findings apply to other populations and healthcare settings also remains uncertain. Longitudinal studies with repeated clinical assessments are needed to clarify temporal relationships and distinguish factors associated with symptom onset from those contributing to persistence. Controlled trials should subsequently evaluate whether integrated management improves both symptom-related and functional outcomes.

Overall, this study supports a comprehensive, non-stigmatizing approach to psychosomatic presentations. Its principal contribution is to emphasize the interdependence of physical symptoms, emotional distress, and everyday functioning while preserving diagnostic caution and avoiding unsupported causal or therapeutic conclusions.

1. American Psychiatric Association. (2022). Diagnostic and statistical manual of mental disorders (5th ed., text rev.).

2. Engel, G. L. (1977). The need for a new medical model: A challenge for biomedicine. Science, 196(4286), 129–136.

3. Fava, G. A., & Sonino, N. (2000). Psychosomatic medicine: Emerging trends and perspectives. Psychotherapy and Psychosomatics, 69(4), 184–197.

4. Henningsen, P., Zimmermann, T., & Sattel, H. (2003). Medically unexplained physical symptoms, anxiety, and depression: A meta-analytic review. Psychosomatic Medicine, 65(4), 528–533.

5. Henningsen, P., Zipfel, S., & Herzog, W. (2007). Management of functional somatic syndromes. The Lancet, 369(9565), 946–955.

6. Henningsen, P. (2018). Management of somatic symptom disorder. Dialogues in Clinical Neuroscience, 20(1), 23–31.

7. Kroenke, K., Spitzer, R. L., & Williams, J. B. W. (2001). The PHQ-9: Validity of a brief depression severity measure. Journal of General Internal Medicine, 16(9), 606–613.

8. Kroenke, K., Spitzer, R. L., & Williams, J. B. W. (2002). The PHQ-15: Validity of a new measure for evaluating the severity of somatic symptoms. Psychosomatic Medicine, 64(2), 258–266.

9. Spitzer, R. L., Kroenke, K., Williams, J. B. W., & Löwe, B. (2006). A brief measure for assessing generalized anxiety disorder: The GAD-7. Archives of Internal Medicine, 166(10), 1092–1097.

10. Rief, W., & Martin, A. (2014). How to use the new DSM-5 somatic symptom disorder diagnosis in research and practice: A critical evaluation and a proposal for modifications. Annual Review of Clinical Psychology, 10, 339–367.

11. Barsky, A. J., & Borus, J. F. (1999). Functional somatic syndromes. Annals of Internal Medicine, 130(11), 910–921.

12. Wessely, S., Nimnuan, C., & Sharpe, M. (1999). Functional somatic syndromes: One or many? The Lancet, 354(9182), 936–939.

13. McEwen, B. S. (1998). Protective and damaging effects of stress mediators. The New England Journal of Medicine, 338(3), 171–179.

14. Kirmayer, L. J., & Young, A. (1998). Culture and somatization: Clinical, epidemiological, and ethnographic perspectives. Psychosomatic Medicine, 60(4), 420–430.