| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From Semmelweis University, Faculty of Medicine, Institute of Behavioural Sciences, Budapest, Hungary
Address reprint requests to: Adriennne Stauder, MD, Semmelweis University, Faculty of Medicine, Institute of Behavioural Sciences, Nagyvarad ter 4, Budapest 1089, Hungary. E-mail: staadr{at}net.sote.hu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
METHODS: Six hundred forty-six (646) consecutive patients with rhinoconjunctivitis (59.3%), asthma (26.8%), or "other" allergy (13.9%), aged 16 to 65 years, completed self-administered questionnaires in six outpatient allergy clinics; 60 of the respondents also participated in structured psychiatric interviews. Anxiety was measured with the Spielberger State-Trait Anxiety.
RESULTS: According to the interviews, STAI-T > 52 predicted with 86% accuracy a current psychiatric diagnosis, without differentiating between anxiety and depression. Using this threshold, the rate of anxiety and/or depressive disorders is estimated as 19% (95% CI: 15.9–22.1) in our unselected allergic outpatient sample; 46% of these patients never received any psychopharmacological treatment, indicating that anxiety related disorders are underdiagnosed and undertreated. Risk indicators were female gender; asthma; perennial symptoms; sleep problems; nonspecific allergy triggers like strong emotions; stressful situations; and considerable limitation in everyday activities attributed to the allergic symptoms.
CONCLUSIONS: Our findings confirm a high rate of anxiety and/or depressive disorders in patients visiting the allergy clinic. Self-administered questionnaires such as STAI-T provide reliable help for the identification of these frequent psychiatric problems.
Key Words: anxiety, • allergy, • rhinitis, • asthma, • Spielberger State-Trait Anxiety Inventory, • depression.
Abbreviations: DSM-IV = Diagnostic and Statistical Manual of Mental Disorders fourth edition; STAI-T = trait anxiety and STAI-S: state anxiety subscales of the Spielberger State-Trait Anxiety Inventory
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
The most common psychiatric disorder associated with allergic diseases is anxiety, the second most common being depression. Anxiety is characterized by exaggerated worry and autonomic hyperarousal together with distorted symptom perception, which often lead to more severe complaints (5, 6), avoidant behavior (3, 7, 8), and overuse of medications (8, 9). Severe asthma patients with psychiatric morbidity had more frequent exacerbations of the symptoms, more frequent visits to the GP, and increased risk for emergency visits and hospitalizations (10); panic-fear and low self-efficacy were the best predictors of high- frequency emergency room visits (11). An association between psychological disturbances and asthma death has also been observed (12–14). The overall quality of life of patients with anxiety is worse than would be expected from their somatic symptoms (15–17).
Despite these connections between anxiety and allergic symptoms, we found relatively few reports in the literature on the prevalence of anxiety disorders in adult allergic patients. Of those we found, the majority focused on asthma, and their data were sometimes contradictory. In different samples of asthma outpatients the prevalence of any anxiety disorder varied from 20% to 53% (10, 3, 18) , that of panic disorder from 6.5% to 17% (18, 19–22) and that of agoraphobia from 13.1% to 24% (18, 19).
In allergic rhinitis the rate of anxiety and/or depressive disorders was estimated to 15% in a health economic analysis (23). In a self-reported community survey 18% of those having hay fever reported panic attacks (24). In young adults with allergic rhinitis, a higher rate of extreme shyness (social anxiety) was observed (25); further analysis revealed that hay fever was directly connected with temperamental inhibition and panic attacks, but not with anxiety or anxiety sensitivity (26). On the other hand very high prevalence (70%) of allergies has been observed in panic patients (27, 28).
Some anxiety disorders were diagnosed in 35% of 100 subjects with type I atopy (allergy not specified) (28). In multiple chemical sensitivity the occurrence of psychiatric morbidity is especially high (29, 30), up to 44% in a controlled study (31). Higher psychiatric symptom scores were reported also in patients with chronic urticaria, generalized pruritus, and atopic dermatitis than in healthy controls (32, 33), but similar to other dermatology outpatients (34), no prevalence data were provided. In a recent Hungarian representative survey allergic people reached higher depression and neurosis scores than nonallergic ones, but anxiety symptoms were not studied specifically (35).
Despite the fact that the impact of psychiatric morbidity on the management of allergic diseases has been recognized, and that high prevalence of anxiety and depression has been demonstrated in several selected study groups, these findings are scarcely applied to everyday practice.
In designing the present study we had three main objectives: 1) to estimate the prevalence of anxiety disorders in an unselected sample of patients attending outpatient allergy clinics; 2) to identify eventual risk factors and diagnostic cues by comparing patient subgroups; 3) to test the usefulness of self-administered questionnaires for identification of anxiety disorders. Our overall intention was to draw attention to the clinical importance of this problem, and to give cues for the diagnosis to enable better recognition and better management of anxiety disorders in this patient group.
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Subjects
Consecutive patients at the selected allergy clinics were asked to fill in the "ALLEPSY" questionnaire during a 1-week period chosen by the treating physician between June 1 and July 31, 1998. The overall compliance rate was over 50% in the six settings, similar to the compliance rate in other self-reported questionnaire surveys (36). A total of 849 questionnaires were returned. We had to exclude 70 respondents because their age was not between 16 and 65 years; 60 cases had no allergy diagnosis; 73 persons had missing answers for basic variables (gender, STAI-T). Finally data from 646 persons were included in the analysis.
A subsample of 60 persons (9.3% of the sample) took part in the psychiatric interview. The interviewees were recruited in two centers (Budapest, Debrecen) from those who signed the consent form attached to the self-administered questionnaire. We sent to 180 such people a postal letter requesting their participation in a personal interview. We included a response envelope and a form on which they could mark the suitable appointment dates from those proposed. Participation was voluntary and we could not offer any financial or material incentive; the main selection criteria were the willingness to participate and time management.
Questionnaires
We compiled a battery of self-administered questionnaires we called "ALLEPSY" (this acronym refers to our interdisciplinary project on allergy and psychiatry). Items on: 1) allergic symptoms (type, onset, perceived severity, seasonality, triggers); 2) general health state, medications, lifestyle; 3) sociodemographic background; and 4) validated psychodiagnostic questionnaires; and 5) informed consent form were included.
Allergy groups were determined on the basis of the treatment diagnosis: I) "RHINITIS": patients with allergic rhinoconjunctivitis; II) "ASTHMA": patients with asthma symptoms (even if they have rhinoconjunctivitis or other allergy too); III) "OTHER": subjects presenting with eczema, chronic urticaria, contact allergy, food allergy, or insect venom allergy.
Anxiety was assessed with the Hungarian version (37) of the Spielberger State-Trait Anxiety Inventory (38). We have chosen this scale because it focuses mainly on the cognitive symptoms of anxiety to avoid confounding with somatic symptoms related to the allergic (or other chronic) disease.
Further items included in the analysis: Seasonality: "In which season(s) do you experience allergic symptoms? Choose spring/summer/autumn/winter/the whole year/changing." The perceived severity of the allergy: "During your last allergic period, to what extent your allergic symptoms limited your everyday activities?", answer options on a visual analogue scale going from 0 ("not at all") to 10 ("limit all my activities"). An 18-item list of common triggering factors was also included: "Put a cross to any of the listed factors that trigger, or worsen your allergic symptoms". Psychiatric medication: "Have you ever used anxiolytics or sedatives? Choose never/sometimes/often/regularly." "Have you ever used antidepressives? Choose never/in the past/currently."; Family history of psychiatric disorder: "In your family, relatives was there any psychical or psychiatric disease?"; items from the Juhasz Neurosis Questionnaire (available in Hungarian) also appeared: "Do you ever feel unjustified fear, or anxiety?" (answers: no, seldom, often); "Do you ever experience mood changes like depression lasting 2 to 4 weeks?" (yes/no). Insomnia score (0–8) was calculated as the sum of the answers to four questions: difficulty falling asleep, sleep quality, early awakening, morning tiredness.
Psychiatric Interview
Authors personally conducted the psychiatric interviews (30–30 interviews); they were masked to the questionnaire scores at the interview. The validated Hungarian version of the M.I.N.I. International Neuropsychiatric Interview Plus (M.I.N.I. Plus) was used to determine the psychiatric diagnoses (39, 40). This relatively simple structured interview is designed to establish current and past psychiatric diagnoses for nonpsychotic psychiatric disorders according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) (41).
Statistical Analysis
For statistical analysis the SPSS for Windows (Statistical Package for Social Sciences) was used (42). We used ANOVA to compare means, with Tukey post hoc test if more than two groups were compared. To measure association between variables Pearson correlation, linear regression, and partial correlation were calculated. We excluded from statistical analysis the cases with missing values pairwise; in the case of STAI-T subjects with more than two missing answers were excluded from the whole analysis. If only one or two answers were missing the score was calculated by replacing the missing value by the mean value of the 18 or 19 valid answers.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
Anxiety Scores
For the whole allergic sample base-rate STAI-T was 42.9 ± 10.7; STAI-S mean 40.6 ± 11.0. There was significant (p < .001) gender difference in the means: STAI-T men: 40.4 ± 9.9; STAI-T women: 44.9 ± 10.9; STAI-S men: 38.8 ± 9.9; STAI-S women: 42.1 ± 11.7. STAI scores increased with age (Pearson r = 0.21 for STAI-T and r = 0.19 for STAI-S; p < .001). There was also a strong linear correlation between STAI-T and STAI-S (r = 0.75).
According to allergy type, STAI-T mean was higher in asthma than in rhinitis and other allergies (45.6 vs. 41.7 and 43.2; p < .01); STAI-S was higher in asthma and other allergies than in rhinitis (43.3 vs. 38.8; p < .01). Higher anxiety score was found in patients with perennial symptoms than in those with seasonal allergy (46.6 vs. 41.7; p < .001) (Table 1). Perceived severity of the allergic symptoms ("Limitation caused by my allergic symptoms") was positively correlated to anxiety scores (r = 0.27 for STAI-T and r = 0.22 for STAI-S).
Comparison of STAI-T Scores with the Psychiatric Diagnoses
The Spielberger Anxiety Inventory is widely used in nonclinical and clinical trials to measure level of anxiety; however no ranges were defined for STAI-T to differentiate between psychiatric cases and noncases. We compared the questionnaire scores with clinical diagnoses established by structured psychiatric interviews (M.I.N.I. Plus) to determine such threshold scores for STAI-T.
From the 58 interviewees 29 persons (50%) met one or more current DSM-IV diagnosis (selection bias: those with psychiatric problems were more motivated to participate at the interview). In 7 cases only anxiety disorder, in 8 cases only depressive disorder, in 12 cases both depressive and anxiety disorder, and in 1 case somatoform disorder was diagnosed according to the M.I.N.I. Plus. The main diagnoses were: major depressive episode (10 cases), dysthymia (2), panic disorder/attacks (7 cases), agoraphobia (4), social phobia (2), generalized anxiety (2), simple phobia (4), adjustment disorder with depressive and/or anxiety symptoms (7), mixed anxiety and depression/MAD (4), and somatoform disorder (3). In 19 cases a single diagnosis and in 10 cases a multiple diagnosis was established (in some categories such as MAD or adjustment disorder both anxiety and depressive symptoms are included in a single diagnosis). We did not consider as psychiatric cases three young men who reported occasional fainting at blood-taking.
We summarized individual STAI-T scores and psychiatric diagnoses in Figure 1. Diagnoses were grouped with focus on anxiety disorders: I) no current psychiatric problem; II) anxiety disorder (also comorbid); III.) only depression or other disorder. All subjects with STAI-T > 52 received some psychiatric diagnosis; however, with six persons only depression or somatoform disorder was diagnosed. It suggests that high STAI-T scores are connected not only with anxiety, but with all clinically significant psychiatric disorders.
|
|
In Table 3 we summarized the distribution of the sample by gender and allergy type according to these anxiety categories. Though we can estimate as 19.0% (95% CI:15.9–22.1) the rate of current psychiatric disorders in our allergic sample, a further 13.2% (95% CI: 10.5–15.9) have mild or subclinical disorders, and 67.8% (95% CI: 64.1–71.5%) have probably no psychiatric problem. The rate of hiANX was significantly higher in women then in men (23.5% vs. 13.3%; p < .001), and significantly higher in asthma than in rhinitis or other allergies (30.1% vs. 14.6% and 16.7%). In seasonal allergy the prevalence of hiANX is significantly lower (16.2%) than in the patients with perennial or varying symptoms (27.9% and 22.8%).
|
Factors Connected with Anxiety Disorders
To identify eventual risk factors and diagnostic cues of psychiatric disorders, we calculated the odds ratio (OR) for a number of variables comparing the loANX group to the hiANX group (Table 4); the OR values for the miANX group were in between, so we do not list them here. We found approximately a two-fold risk of psychiatric disorders in women compared with men (OR = 2.3; CI: 1.5–3.5); in asthma compared with rhinitis (OR = 2.4; CI:1.6–3.8); in case of perennial or changing symptoms compared with seasonal symptoms (OR = 2.0; CI: 1.3–3.0); and in patients with history of psychiatric problem in the family (OR = 2.1; CI: 1.2–3.7). Positive answers to the following questions can serve as diagnostic cues: "Allergic symptoms cause considerable limitations in everyday activities": score > 5 on a 0 to 10 scale (OR = 2.9; CI: 1.9–4.4); problems with sleep: insomnia score > 3 (OR = 7.5; CI: 4.8–11.8); "Did you used to feel unjustified fear, anxiety?": the answer yes indicated 10-fold risk (OR: 11.1; CI: 6.7–18.4), the answer "often feels anxiety" (OR = 63.6; CI: 26.4–153.3); positive answer to "Have you had mood changes lasting 2 to 4 weeks" indicated 15-fold risk (OR:15.0; CI: 9.3–24.1). The rate of suicide attempts was much higher in the hiANX group than in the loANX group ((8.1% vs. 2.1%) (OR = 4.2; CI: 1.7–10.6); and hiANX patients reported more often that "strong emotions" (OR = 6.4; CI: 3.5–11.5) or "stressful situations" (OR = 4.2; CI:2.6–6.7) trigger or worsen their allergic symptoms.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Mean STAI-T score (42.9 ± 10.7) in our unselected allergic outpatient sample was similar to the mean (42.9 ± 8.4) established for the Hungarian population (37); however, higher than American population means (STAI-T < 39) (38). Concerning clinical groups, in a Hungarian comparative study STAI-T, the mean was 65.3 in generalized anxiety and 55.8 in panic disorder compared with 38.2 in the control group (43). Another study on construct validity of STAI found mean scores ranging from 52 to 56 in anxiety groups and 33.4 in nonclinical controls (44). These means are consistent with the threshold scores we determined for STAI-T, scores > 52 indicating clinically significant anxiety (or depressive) disorders and scores between 48 to 52 indicating mild or subclinical disorders; at STAI-T < 48 the probability of any clinically significant disorder is very low. However this good accuracy at the screening (> 85%) was reached without discriminating between anxiety and depressive disorders as most cases with depression also reached high STAI-T scores.
Some other researchers have also reported that STAI-T scores were strongly correlated not only with anxiety diagnoses, but also with depression (45, 46). An exploratory analysis of STAI-T revealed that certain items measure primarily negative affect; therefore they might reflect depression rather than anxiety (44). Most screening instruments assessing anxiety and depression tend to be highly correlated; this overlap is difficult to avoid given that these states have a lot of common features like dysfunctional thinking, cognitive distortions, and negative affect (eg, distress, irritability, tension) with a few distinctive features such as physiological hyperarousal in anxiety or low levels of positive affect in depression. Besides this common problem of symptom overlap in questionnaire screening, the "true" comorbidity of anxiety and depression is very common, estimated to 50% in population studies (47, 48), and even in cases without multiple diagnosis mixed symptoms can be present. This can be especially true for allergic patients; Kennedy et al. (27) found that psychiatric patients treated either for depression or panic disorder with a history of allergies reached significantly higher anxiety scores than nonallergic ones.
With these stipulations, we concluded that STAI-T had acceptable screening properties in our sample, which enabled us to make an estimation of the prevalence of anxiety and/or depressive disorders in our sample of unselected allergic outpatients. The rate of significant disorders was 19%, with a further 13.2% of the sample having mild or subclinical symptoms. This rate is significantly higher than observed in a recent Hungarian study, where the 1-month prevalence of some anxiety and/or affective disorders was 12.9% in a comparable sample of 3000 persons attending primary care services (49). In that study depression without anxiety was diagnosed only in 1.6% of the cases, and similarly to our survey women had a 2.5-fold risk of a positive psychiatric diagnosis.
Concerning the type of the allergy, we found the highest prevalence of anxiety and/or depressive disorders in asthma: 30.1% with further 9.2% of subclinical cases. Our data are comparable to the findings of Yellowlees and Kalucy (3) reporting some anxiety disorder in 34% of patients with asthma; this rate is higher than the 20% found by ten Brinke et al. (10) in severe asthma but significantly lower than the 52.3% reported by Nascimento et al. (18).
The 14.6% prevalence of psychiatric disorders in allergic rhinitis in our sample is concordant with the 15% reported on the basis of a health economic analysis; however they estimated the rate of depression as 12% and that of anxiety as 3% (23). This higher proportion of depression can be explained by the methodology used: diagnoses were based on anxiolytic and/or antidepressive drug prescriptions over a 1-year period; however antidepressives are also the first-choice treatment for most anxiety disorders like panic disorder, agoraphobia, or social phobia. Although the rate of anxiety problems in allergic rhinconjunctivitis is only slightly higher than in the general population, and much lower than in asthma, the importance of the problem in this patient group should not be underestimated. Nearly half of all patients with high anxiety were treated for rhinitis at the allergy clinics.
An important limitation of our study is the relatively low response rate to the screening and to the interviews. It has probably minimal effect on the reliability analysis of the STAI-T questionnaire or on the analysis of the risk factors; however it might influence our epidemiological findings. Our experience was that because no incentive was offered, those with psychiatric problems were more motivated to participate in the interviews; in spite of this, the subsample turned out well representing the population studied regarding most of the demographic variables. Concerning the questionnaire survey where data were collected during the visit at the allergy clinic, indirect evidence suggests that no such selection bias aroused: mean STAI-T score of the sample was similar to the Hungarian standard (while it was significantly higher in the interview sample). The 95% confidence interval values give an estimate of the possible effect sizes of the sample selection bias and indicate the range of reliability of the results.
Clinical Implications
Anxiety and depression is more prevalent in chronically ill than in "general" populations. In a general medical outpatient sample for example the prevalence of mental disorders was 27% (50). The rate of psychiatric morbidity is especially high in patients with unexplained somatic complaints, who are often referred to an allergy clinic, to exclude (or confirm) the eventual allergic background of their unusual symptoms (51). Although these patients seek help primarily for their somatic symptoms, identification and adequate treatment of the underlying or comorbid psychiatric disorders is especially important. In our sample only half of those with high STAI-T scores ever had any psychopharmacological treatment, suggesting that a considerable proportion of the psychiatric problems have never been dealt with.
The diagnosis of anxiety disorders might be especially difficult in patients with allergic diseases because of the similarities of certain symptoms such as disturbed sleep, daytime tiredness, difficulty in concentrating, hyperventilation, or dyspnea (21, 52) . Anxiety attacks might appear either between or in combination with asthma attacks; constant worry, negative affect, or low mood might seem to be "normal" in patients with chronic disease. Psychiatric diagnostic criteria are based on the frequency, magnitude, and duration of these symptoms together with the subjective suffering and everyday interference caused by them.
Considering the high prevalence of these problems, and the overload of the specialized clinics, diagnostic and treatment effectiveness might be improved by routine use of self-administered questionnaires, as the collection of structured information before the visits allows focused discussion of problem areas during the available visit time (53–55).
In our study STAI-T in itself was a reliable indicator of anxiety and/or depressive disorders, but its combination with a depression questionnaire would give more specific results. A number of other validated screening instruments are also available. None of them can replace the psychiatric interview, but can contribute to it, and may help decisions about specific treatments or referral to a mental health specialist.
If routine employment of screening questionnaires is not adopted, direct diagnostic questions such as "How is your sleep?," "Do you ever feel unjustified fear, or anxiety?," "Have you experienced mood changes lasting 2 to 4 weeks?," or "Do you know about any psychological or psychiatric problem in the family?" can be very informative. We also identified a number of factors associated with increased risk of psychiatric problems in allergic patients such as female gender, older age, asthma, perennial symptoms, nonspecific triggering factors such as stress, strong emotions, extreme cold or warm, physical load, and considerable limitation in everyday activities attributed to the allergic disease.
We concluded from our study that the prevalence of anxiety and/or depression in allergic outpatients is high, and these symptoms often remain undiagnosed and untreated. Screening questionnaires may effectively contribute to the identification and management of psychiatric disorders in these patients.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
This study was realized as PhD work of the authors. The research was supported by the Hungarian National Research Found (OTKA) Grant F029857 and TS-040889, by the NKFP-01/002/2001 project and by the PhD School of the Semmelweis University Budapest.
Received for publication September 3, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  C. Hausteiner, S. Bornschein, E. Bubel, S. Groben, C. Lahmann, M. Grosber, B. Lowe, F. Eyer, B. Eberlein, H. Behrendt, et al. Psychobehavioral Predictors of Somatoform Disorders in Patients With Suspected Allergies Psychosom Med, November 1, 2009; 71(9): 1004 - 1011. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Alati, M. O'Callaghan, J. M. Najman, G. M. Williams, W. Bor, and D. A. Lawlor Asthma and Internalizing Behavior Problems in Adolescence: A Longitudinal Study Psychosom Med, May 1, 2005; 67(3): 462 - 470. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
