| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Section for Clinical Psychology and Psychotherapy, Philipps-University of Marburg, Marburg, Germany (A.M.); and the Institute of Clinical Psychology and Psychotherapy, Technische Universität Dresden, Dresden, Germany (F.J.).
Address correspondence and reprint requests to Alexandra Martin, PhD, Philipps-University, Section for Clinical Psychology and Psychotherapy, Gutenbergstr. 18, D-35032 Marburg, Germany. E-mail: martin{at}staff.uni-marburg.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODRESULTSDISCUSSIONAppendix 1NOTESREFERENCES |
|---|
Methods: Analyses of the present study are based on the German Health Interview and Examination Survey–Mental Health Supplement (N = 4181, representative for the German population from 18–65 years). The assessment included interviews for somatic conditions and mental disorders and self-report ratings on health-related quality of life, healthcare utilization, disability days, and physical activity.
Results: Only three cases (0.05%) were identified as meeting full criteria of Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) hypochondriasis. The prevalence rate of the less restrictively defined form of hypochondriasis, ("subthreshold hypochondriasis") was 0.58% and an additional 2.12% reported having had illness worries for at least 6 months but did not meet further hypochondriasis criteria. The two subthreshold diagnostic groups provided strong evidence of difference from the nonhypochondriac controls: comorbidity with psychiatric and medical disorders and healthcare utilization were higher, and quality of life was markedly reduced.
Conclusions: The results provide additional support to not only consider "full" DSM-IV hypochondriasis, which is a very rare disorder in the general population, but also to include less restrictive hypochondriac conditions—associated with a clinically relevant degree of psychological and physical impairment—into clinical and scientific considerations.
Key Words: hypochondriasis • illness worry • epidemiology • psychiatric disorders • somatoform • DSM-IV
Abbreviations: DSM-IV = Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; GHS = The German National Health Interview and Examination Survey; GHS-CS = core survey; GHS-MHS = mental health supplement; H0 = no illness worry group; H1 = unrealistic illness worry group; H2 = subthreshold hypochondriasis; M-CIDI = Munich Composite International Diagnostic Interview.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODRESULTSDISCUSSIONAppendix 1NOTESREFERENCES |
|---|
Hypochondriasis is a highly disabling condition. Patients with hypochondriasis were found to view their health as worse than and to exhibit more severe psychiatric symptoms than age- and sex-matched control subjects (3). The association between health-related anxiety and higher medical care utilization persists even after adjusting for differences in sociodemographic characteristics and medical morbidity (4).
Relatively little is known about the epidemiology of hypochondriasis, especially in the general population. The few population-based studies using interviews consistently reveal no or only very few hypochondriasis cases (5–7). Only studies based on less restrictive inclusion criteria or studies using self-report questionnaires revealed higher case numbers. These prevalence rates varied from 4% to 10% in population-based samples (6,8,9).
Hypochondriasis has been investigated more often in medical patients. In general medical practice, the prevalence of hypochondriasis varied between 0.8% and 9% (10–12), and a median prevalence of 4.2% in seven primary care samples has been reported (13). Most of these studies indicate that hypochondriac concerns appear quite often in medical settings. However, clinical samples may be influenced by selection bias. Thus, estimates of prevalence still vary as a result of the definition of hypochondriasis, instruments, and sample characteristics.
It has been questioned whether to classify hypochondriasis as a discrete diagnostic entity. Some conceptualize hypochondriasis dimensional rather than as a category, underlying many different mental as well as physical conditions. The current DSM-IV hypochondriasis diagnosis has been questioned because it does not satisfy either clinical or nosologic validity requirements (14). Gureje and colleagues (10) have criticized the current definition of hypochondriasis as too restrictive for use. They outline the relevance of "abridged hypochondriasis" consisting of the triad of illness worry associated with distress or interference with functioning and medical help-seeking (ICD-10 F.45.2, criteria A and B). The full criteria of hypochondriasis, including the criterion that patients refuse to accept medical reassurance that symptoms had no physical causation, were identified only in 0.8% of cases in primary care (in 41 of 5447 patients across 14 countries); the prevalence of "abridged hypochondriasis" was 2.2%. Patients with abridged hypochondriasis were different from nonhypochondriac patients but highly similar to patients meeting the full ICD-10 criteria. Looper and Kirmayer (6) have also investigated the concept of the previously mentioned "abridged hypochondriasis" as well as the even broader concept of illness worry (Composite International Diagnostic Interview [CIDI] probe question for hypochondriasis) in a community population. They found a rate of 1.3% for abridged hypochondriasis, which is lower than in primary care. However, 6.2% of the sample reported illness worry during a period of 6 months, and illness worry was an independent predictor of increased levels of distress and of healthcare utilization.
Given the fact that there is a lack of surveys to assess hypochondriasis in the adult general population based on clinical diagnostic interviews, the first aim of the present study was to estimate 12-month prevalence rates of DSM-IV hypochondriasis and of subthreshold hypochondriac conditions in the general population.
The second target of this study was to investigate external validity of the criteria used to define hypochondriasis in DSM-IV. This question was addressed with the comparison of three diagnostic groups fulfilling criteria of hypochondriasis to different degrees—"subthreshold hypochondriasis," "unrealistic illness worry," and "no illness worry"—with regard to some relevant indicators such as physical and mental health, quality of life, and healthcare utilization.
|
METHOD |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODRESULTSDISCUSSIONAppendix 1NOTESREFERENCES |
|---|
Assessment
Hypochondriasis/Mental Disorders
Psychiatric diagnoses were assessed by the computer-assisted version of the Munich Composite International Diagnostic Interview (DIA-X/M-CIDI) (18). This fully structured interview is a modified version of the World Health Organization CIDI, version 1.2 (19) supplemented by questions to cover DSM-IV criteria of mental disorders. Psychometric properties and feasibility of the CIDI somatoform section are regarded as good (20). Interviews were conducted at the respondent’s homes by clinically trained interviewers (psychologists and MDs) because the use of lay interviewers has been discussed as a limitation of CIDI studies in the past. The following diagnostic groups were defined with increasing fulfillment of DSM-IV criteria of hypochondriasis (for the detailed coding algorithm, see the Appendix):
Given the low rates of threshold DSM-IV-hypochondriasis in the general population sample, further analyses only included comparisons among the groups "no illness worries" (H0), "unrealistic illness worry" (H1), and "subthreshold hypochondriasis" (H2). The subjects fulfilling all criteria of hypochondriasis were included in the group "subthreshold hypochondriasis."
Somatic Conditions
A structured computer-assisted clinical medical interview (CAPI) was conducted by study doctors to assess the presence of a wide range of acute and chronic somatic conditions (e.g., cardiac diseases, ulcers, gastritis, diabetes, cancer, neurologic diseases, musculoskeletal diseases, allergies). In this interview, stem questions ("Were you ever assigned the diagnosis of ... by a physician?") were supplemented by additional disease-specific probe questions (16).
Health-Related Quality of Life
The German version of the 36-Item Short Form Health Survey SF-36 (21,22) is a multidimensional questionnaire measuring health-related quality of life based on the World Health Organization definition of health. The SF-36 includes eight subscales (ranging from 2–10 items): general health, physical functioning, bodily pain, vitality, mental health, social functioning, physical role limitation, emotional role limitation. Subscale scores were transformed into a 0 to 100 scale, higher scores representing better health.
Health Service Utilization and Illness Behavior
Subjects were asked to indicate whether they had seen medical specialists from a list of 18 doctor subspecialties during the previous 12 months and, if so, how often they had consulted them. The total number of healthcare visits during the past 12 months was counted as an indicator of healthcare utilization. All subjects were asked to report the number of disability days (not being able to carry out usual activities as a result of feeling sick) during the past 12-month period. In addition, subjects were asked to indicate the number of hours per week spent on physical/sporting activity (considering the previous 3 months); based on these data, the number of subjects with less than 1 weekly hour of physical activity was calculated.
General Sociodemographic Information
Data on age, marital status, education, income, and work situation were collected by questionnaire. The "index of social class" was derived from information on educational level, income, and current job position (23).
Statistical Analyses
Prevalence estimates were weighted according to the sampling scheme and to adjust the net sample to age, sex, and community distributions in Germany (corresponding to the national administrative statistics of December 1997). The mental health supplement sample was considered to provide a good to almost perfect representation of the German population in terms of the selected biosocial variables (16). Calculations of 12-month prevalence rates and 95% confidence intervals were performed using the Stata software package, version 7.0 (24).
Further analyses were based on nonweighted data. Thus, all subjects fulfilling the previously mentioned criteria were included, and group sizes differed slightly from weighted data. To assess differences between the three diagnostic groups, mean scores of criteria variables were compared by the WELCH test (asymptotic F distribution) (25). This procedure is comparable to analysis of variances but takes into account that population variances differ significantly, and testing is more conservative (26). If the overall group effect was significant, multiple paired post hoc comparisons of groups were computed with the test of GAMES-HOWELL to control for the probability of increasing type I error. This test does not need homogeneity of variances and is recommended in cases of unequal sample sizes (27,28). Effect sizes (Cohen’s d) of the paired comparisons were computed. Health-related quality of life was assessed with the eight subscales of the SF-36. To reduce risk of type I error, group differences were first analyzed with multivariate analysis of variances as a global test. All analyses on group differences were conducted with SPSS statistical package (version 12.0).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODRESULTSDISCUSSIONAppendix 1NOTESREFERENCES |
|---|
|
Comparison of Diagnostic Groups
The subjects were assigned to three different groups according to the degree of criteria fulfilled for DSM-IV hypochondriasis. The three study groups "subthreshold hypochondriasis" (N = 33, including the three threshold cases), "unrealistic illness worry" (N = 101, with subthreshold hypochondriasis not included), and the control group "no illness worry" (N = 3698) were compared on sociodemographic variables, medical and psychiatric conditions, psychological variables, and healthcare utilization.
Sociodemographic Variables
In the subthreshold hypochondriasis group, 63.4% were women; in the illness worry group, 52.5% were women; and in controls with no illness worries, 53.9% were women (not significant; Table 2). Subjects experiencing illness worries were significantly older than controls (mean age, 45.2 versus 41.5 years), whereas subjects in the subthreshold group were not. Being single was more frequent in control subjects (25.2% versus 12.1/14.3%). The groups did not differ with respect to their subjects’ socioeconomic status (index of social class).
|
Physical and Psychiatric Comorbidity
The rate of subjects with at least one physical condition during the last 12 months was generally high in all groups (Table 2). At least one physical diagnosis was assigned to 69.4% of the controls (H0), and even more subjects in the two hypochondriasis groups were affected by at least one physical condition (90.9% in H2, 87.1% in H1; these physical conditions were not rated as explaining the subjects’ illness worry). The number of additional physical conditions was highest in subthreshold hypochondriac subjects followed by subjects with unrealistic illness worry and lowest in controls without illness worry.
The base rate of having any mental disorder in the total sample was 31.1% (29). Most of the subjects in H2 and H1 met the DSM-IV criteria of at least one mental disorder (81.8% in "subthreshold hypochondriasis," 74.3% in "illness worry" group), whereas subjects without illness worries (H0) were not affected to the same degree by any mental disorder (30.6%). Likewise, the average number of comorbid mental disorders was highest in "subthreshold hypochondriasis" (mean, 3.3) followed by the "illness worry group" (mean, 1.8) and lowest in control subjects (mean, 0.5).
Health-Related Quality of Life
Table 3 presents the mean scores of the SF-36 dimensions, which cover different aspects of health-related quality of life. First, a multivariate analysis of variance was computed. The overall effect of the group factor on all eight subscales was highly significant (
= 0.93, F(16,7350) = 16.4; p < .001). Both illness worry groups perceived all included aspects of health-related quality of life significantly worse than control subjects without illness worries. "Subthreshold hypochondriasis" and "unrealistic illness worry" groups differed in five subscales (reduced quality of life in the "subthreshold hypochondriasis" group with regard to physical role functioning, general health, vitality, emotional role functioning, and mental health).
|
Healthcare Utilization and Illness Behavior
The number of healthcare visits during the previous 12 months served as an indicator of healthcare utilization (Table 4). The mean overall number of healthcare visits was almost twice as high in the illness worry groups (in H2: mean, 20; in H1: mean, 18) than in control subjects without illness worry (mean, 10). Accordingly, results of the paired comparisons showed the number of doctor visits to be significantly higher in both illness worry groups than in control subjects; healthcare utilization did not differ significantly between subthreshold hypochondriasis and unrealistic illness worry group.
|
Most of the healthcare visits were dedicated to general medical doctors in all groups, and only a few consultations with psychiatrists and psychotherapists were reported. The average number of mental health visits did not vary between the diagnostic groups. However, with regard to the proportion of subjects consulting a mental health professional at least once during the previous year, a slightly different picture emerged: 28.1% of the "subthreshold hypochondriasis" group and 18.8% of the "unrealistic illness worry" group had seen a psychiatrist or psychotherapist at least once during the previous year. At a significantly lower rate, subjects in the "no illness worry" group (7.6%) consulted mental health professionals.
The total number of disability days during the previous 12 months was higher in the "subthreshold hypochondriasis" (mean, 57.3) and "unrealistic illness worry" (mean, 29.3) groups compared with control subjects (mean, 11.4); the comparison between the "subthreshold hypochondriasis" and "unrealistic illness worry" groups was not significant.
Physical activity was another aspect of illness behavior investigated in this survey. Approximately 75% of subjects in the two hypochondriasis groups reported less than 1 hour weekly spent on physical activity, whereas the proportion of physically inactive subjects was smaller in the control group (60%).
Reanalysis to Control for Physical Comorbidity and Age
In a second step, a reanalysis of the main variables was conducted controlling for the number of comorbid physical diagnoses and for age (ANCOVAs). Although the effects of the covariates were significant, the group differences for healthcare utilization, disability days, and health-related quality of life were still highly significant. Thus, the identified group differences in these features cannot be solely explained by the differences in somatic health status or age.
Reanalysis to Control for Potential Influence of the Hypochondriasis Cases
Because of the small number of hypochondriasis cases, the data of the three cases were analyzed together with the 30 subthreshold hypochondriasis cases in comparison of the diagnostic groups. To rule out a possible bias resulting from these cases, we reanalyzed the data without the hypochondriasis cases, showing the same result pattern as described. In addition, descriptive data of the three hypochondriasis cases suggest mean values to be comparable to subthreshold cases.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODRESULTSDISCUSSIONAppendix 1NOTESREFERENCES |
|---|
Prevalence in the General Population
The GHS-MHS was the first representative nationwide mental health survey in Germany covering somatic and mental disorders simultaneously within one study. Threshold hypochondriasis was found to be very rare: only three (weighted: two) of 4181 subjects were identified, resulting in a 12-month prevalence estimate of 0.05%. This result confirms prior findings that hypochondriasis according to its classification in DSM-IV and ICD-10 is extremely rare in the general population as well as in primary care surveys (6,10).
It has been argued to define hypochondriasis by using less restrictive criteria, because the estimated prevalence in subthreshold conditions such as "abridged hypochondriasis" might depict the clinical reality of the disorder more appropriately (10). Following this approach, we defined a subthreshold group of hypochondriasis by omitting the criterion B of DSM-IV ("the preoccupation persists despite appropriate medical evaluation and reassurance"), which might be a less reliable criterion. This condition was characterized by hypochondriac illness worries lasting for at least 6 months and the preoccupation resulting in clinically significant distress or impairment in important areas of functioning. The weighted 12-month prevalence rate of subthreshold hypochondriasis was 0.58%. As expected, this is still a lower rate than in primary care. Furthermore, an additional 2.12% of the subjects reported persistent illness worries without evidence of the conditions the subjects were worrying about.
Although psychometric properties and feasibility of the somatoform section of the standardized CIDI have been proven acceptable at least in clinical populations (20), one could question whether such an interview-based survey can properly assess the phenomenon of hypochondriasis. Affected subjects are probably willing to report illness worries but might be less likely to disclose their difficulties to accept medical reassurance and the severity of the problem, leading to low prevalence rates for hypochondriasis.
Because of the low number of cases with full hypochondriasis, it was not possible to compare this group directly with subthreshold conditions. It is therefore difficult to draw any conclusion about the relevance of this group. Instead, we considered three diagnostic groups ("subthreshold hypochondriasis," "unrealistic illness worry," and "no illness worry").
The two hypochondriac groups did not differ from nonhypochondriac subjects with respect to most sociodemographic variables. Women did not appear to be affected more often by hypochondriac conditions than men. Comparably, Creed and Barsky (13) identified only one of 11 studies that showed a significant association with female sex. Subjects experiencing illness worries were slightly older than control subjects, but subjects with the more severe form of subthreshold hypochondriasis were not older than the two other groups. The majority of previous population-based studies (eight of 11 studies) did not report a significant association between hypochondriasis and older age (13). On the other hand, general health anxiety increases probably with age as physical problems do as well (9).
Comparisons Between Hypochondriac Groups and Control Subjects
In all included health-related variables, the two hypochondriac groups differed from the nonhypochondriac control subjects: comorbidity with mental disorders and medical conditions were higher, health-related quality of life was markedly reduced (in all SF-36 dimensions), and healthcare utilization and illness behavior (disability days, lack of physical activity) were exceeding the one of control subjects. The effect sizes of paired group comparisons were medium to large.
Subjects with subthreshold hypochondriasis as well as those with unrealistic illness worries reported almost twice the number of healthcare visits in the previous year than control subjects. It has to be considered that the data on healthcare utilization were based on self-report, and assessment through medical or health insurance record review would be preferable but was not possible within the limits of a representative population survey. Still, the positive significant relationship between self-reported and dossier data on doctor visits has been reported previously (30). Furthermore, the result of this study is compatible with many previous findings of an association between hypochondriasis and somatization with increased outpatient utilization (4,31,32). The results on higher rates of general doctor visits in subthreshold hypochondriasis and the unrealistic illness worry group persisted even if controlled for higher physical morbidity and age. Thus, the degree of illness worry and hypochondriac preoccupation seem to be important in addition to possibly existing somatic conditions. This confirms prior findings of higher physical comorbidity being associated with illness worry but illness worry contributing to increased levels of healthcare utilization as an independent factor (6,7). Of course, our cross-sectional assessment does not allow answering the question of causality.
Compared with attendance rates in general medical care, psychotherapists or psychiatrists were rarely consulted. Although the number of subjects reporting at least one mental healthcare visit was higher in both hypochondriac subgroups than in control subjects, the mean number of visits at psychotherapists and psychiatrists was low, and as a result of these low rates and variability, no significant differences between groups were found. Against the background of considerable psychiatric comorbidity, this is especially striking in hypochondriac groups and suggests once more that people with fears of having a physical disease prefer to consult general medical services rather than mental health professionals. The low mental health consultation rates may also result from not being referred to mental health services, because these problems often lack recognition in general medical settings (33).
With regard to all these clinical features, the results indicate increased levels of impairment not only in the group of subthreshold hypochondriasis, but already in the group characterized by persistent illness worries. Comparing subthreshold hypochondriasis with the unrealistic illness worry group demonstrates more severe disability in the former group. In the group with subthreshold hypochondriasis, comorbidity with physical and mental disorders was highest and quality of life lowest in some important areas (physical and emotional role functioning, perception of the general and mental health and vitality). Interestingly, the two hypochondriac groups did not differ with respect to their healthcare utilization and the other aspects of illness behavior. These results confirm findings from studies using abridged hypochondriasis or significant illness worry definitions for hypochondriasis (6,10).
Definition of Hypochondriasis and the Criterion of Reassurance
Most definitions of hypochondriasis include "the persistent preoccupation with illness worry or conviction" as a core criterion. The criterion of "reassurance" has been frequently criticized. Gureje and colleagues (10) argued that the ICD-10 criterion "persistent refusal to accept medical reassurance" might constitute a "bottleneck" in the diagnosis of hypochondriasis, because it has a disproportionate effect on prevalence. Others criticized its current conceptualization in DSM-IV because a precise definition on the "appropriateness" of medical reassurance is not provided. Patients with hypochondriasis are generally regarded as resistant to routine, ordinarily given medical reassurance, but not necessarily to types of reassurance that are more adapted to the particular situation of the hypochondriac subject (34). There also might be physicians who do not provide an appropriate and effective reassurance for patients with hypochondriac concerns or people consider not being reassured the same as not being heard enough. Based on these objections, further specification or modification of reassurance criterion seems to be useful. Fink et al. (14) did not include the reassurance criterion in their new empirically derived set of diagnostic criteria for hypochondriasis any more, because it lacked discrimination between different somatoform disorders.
The present study revealed a very low prevalence of the hypochondriasis (fulfilling all criteria), which is not sufficient evidence to abandon the criterion of reassurance as has been suggested previously (10). However, results showed that already subjects with a syndrome consisting of persistent and interfering hypochondriac illness worry show features that are generally associated with full hypochondriasis. The present study addressed only some relevant external criteria variables such as sociodemographic characteristics, healthcare utilization, and other aspects of illness behavior and health-related quality of life. Further studies on the validity of the diagnostic criteria should consider additional external criteria variables (e.g., cognitive–perceptual and biologic processes), but also predisposing factors and aspects of predictive validity (stability of the syndrome; responsiveness to treatment).
Several authors suggested modifications of the current classification of somatoform disorders: Rief and Hiller (35) proposed three different somatoform diagnostic categories characterized by multiple somatization, specific somatoform symptoms, or health anxiety. Fink et al. (14) suggested a new set of empiric diagnostic criteria for hypochondriasis with obsessive rumination about illness as the central feature. In agreement with these authors, we recommend the adoption of a severity specifier to distinguish clinically relevant subgroups.
Clinical Perspectives
The importance of early intervention in hypochondriasis and somatization has been emphasized because the condition becomes more refractory once the patient develops a stable adaptation of the sick role (4), and there is the risk of iatrogenic harm that can result from somatization (36,37). Chronic somatoform disorders are usually associated not only with substantial individual burden for the patient, but also with high direct healthcare costs (as a result of inappropriate or overutilization of medical services) and high indirect costs (disability and work loss resulting from sickness behavior) (38). A number of interventions have been developed and the effectiveness of cognitive–behavioral therapy has been demonstrated (39–41). Early access to appropriate interventions relies on the early identification of the disorder. This could be facilitated by focusing on central features such as preoccupation with health and illness worry. The present study has provided further evidence on the clinical relevance of hypochondriac conditions, which are still labeled as "subthreshold" by current diagnostic classification systems.
|
Appendix 1 |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODRESULTSDISCUSSIONAppendix 1NOTESREFERENCES |
|---|
|
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODRESULTSDISCUSSIONAppendix 1NOTESREFERENCES |
|---|
Data of this study are available as Public Use File from the second author (manual and variable description in German): Dr. Frank Jacobi, Institute of Clinical Psychology and Psychotherapy, Chemnitzer Str. 46, D-01187 Dresden. E-mail: jacobi{at}psychologie.tu-dresden.de. For further information about the Core Survey (GHS-CS) and its public use file, contact the Robert Koch-Institute, Mrs. Ross, Nordufer 20, D-13353 Berlin. E-mail: rossp{at}rki.de
Received for publication September 5, 2005; revision received April 7, 2006.
DOI:10.1097/01.psy.0000238213.04984.b0
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODRESULTSDISCUSSIONAppendix 1NOTESREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. Lieb, G. Meinlschmidt, and R. Araya Epidemiology of the Association Between Somatoform Disorders and Anxiety and Depressive Disorders: An Update Psychosom Med, November 1, 2007; 69(9): 860 - 863. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Rief and G. Rojas Stability of Somatoform Symptoms Implications for Classification Psychosom Med, November 1, 2007; 69(9): 864 - 869. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Kroenke, M. Sharpe, and R. Sykes Revising the Classification of Somatoform Disorders: Key Questions and Preliminary Recommendations Psychosomatics, August 1, 2007; 48(4): 277 - 285. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
