Psychosomatics 46:529-539, November-December 2005
doi: 10.1176/appi.psy.46.6.529
© 2005 Academy of Psychosomatic Medicine
Prevalence and Correlates of Illness Worry in the General Population
Russell Noyes, Jr., M.D.,
Caroline P. Carney, M.D., M.S.,
Stephen L. Hillis, Ph.D.,
Laura E. Jones, M.S., and
Douglas R. Langbehn, M.D., Ph.D.
Received March 22, 2004; revision received Dec. 13, 2004; accepted Feb. 2, 2005. From the Department of Psychiatry, University of Iowa Roy J. and Lucille A. Carver College of Medicine, Iowa City, IA; the Departments of Internal Medicine and Psychiatry and the Regenstrief Institute, Indiana University School of Medicine, Indianapolis; the Department of Statistics and Actuarial Science, Iowa City Veterans Affairs Medical Center, Iowa City, IA; and the Department of Biostatistics, College of Public Health, University of Iowa, Iowa City, IA. Address correspondence and reprint requests to Dr. Noyes, Psychiatry Research, Medical Education Building, Iowa City, IA 52242-1000; russell-noyes{at}uiowa.edu (e-mail).
|
ABSTRACT
|
The prevalence and correlates of illness worry in the general population were investigated in a representative sample. The authors screened residents of the United States by telephone, and more detailed interviews were conducted with 123 respondents who reported at least 1 month of worry about serious illness in the past 12 months and an equal number of randomly selected persons without such worry. Data on demographic characteristics, medical and psychiatric conditions, functional impairment, and health care utilization were collected. At least 1 month of worry was endorsed by 13.1% of the screened population. Correlates of worry included a cluster of psychiatric conditions (major depressive episode, panic attacks, and generalized anxiety disorder) and three clusters of physical conditions (heart disease, cancer, and other diseases). Worry about serious illness was associated with functional impairment and health care utilization.
Key Words: Hypochondriasis • Prevalence • Correlates • Illness Worry • General Population
|
INTRODUCTION
|
Hypochondriasis is characterized by excessive or unrealistic worry about serious illness that persists for 6 months or more.1 Worry of this kind affects an estimated 2%–7% of primary care outpatients, among whom it is associated with impairment in functioning and increased utilization of health services.2 It is a significant problem, yet relatively little is known about its epidemiology. Few community surveys have been done, and most information about prevalence, risk factors, coexisting disorders, and morbidity has come from general medical and psychiatric populations. Data from these sources are confounded by treatment-seeking and may reflect greater severity of illness worry. We undertook a survey of the general population to determine the prevalence, correlates, and morbidity associated with this worry.
Existing surveys suggest that DSM-IV hypochondriasis may be infrequent in the general population.3–6 However, significant illness or hypochondriacal worry appears to be more common; a frequency of 6% was reported in one survey and 10% in another.6,7 Many researchers recognize a continuum, with normal illness worry at one end and hypochondriasis at the other. When a dimensional measure reflecting such a continuum is used, many persons in the community score high. In one study, 8% of respondents scored  8 on the 14-item Whiteley Index of Hypochondriasis.7 Another study found that a score of 5 discriminated between medical outpatients with and without hypochondriasis.8 Yet, no diagnostic threshold for the disorder has been established.
Relatively little information about risk factors has emerged from community surveys. One study observed more physical conditions among persons with illness worry.6 Investigations in primary care settings have similarly found more physical illness among patients with hypochondriasis.9–12 Consistent with this finding, other studies have found that persons with illness worry and hypochondriasis tended to be older.4,6,7,13 Although one survey found that more women had health worry,7 others have not found a clear gender difference.4,6,13,14 Two studies found that hypochondriasis was associated with less education,4,13 and one found that recent adverse life events were more frequent among those with illness worry.6
Community surveys have also shown that persons with illness worry and hypochondriasis have elevated rates of psychiatric disorders.3,4,6 In general medical populations, extensive comorbidity with depressive, anxiety, and somatoform disorders has been found.15,16 Greater morbidity and care-seeking among persons with hypochondriasis has been found in community surveys as well. For instance, Noyes et al.4 observed greater physical impairment and diminished work performance among persons with hypochondriasis, compared to those without the disorder. Similarly, greater disability was observed among persons with illness worry, even after medical morbidity had been taken into account.6 Increased utilization of medical services was also observed in these studies.
This study examined the prevalence of and risk factors for illness worry in the United States. The aim was to replicate findings from a study in Montreal that was based upon a large representative sample.6 A nationwide survey was conducted so that findings could be generalized to the overall population and multivariate analyses could be used to identify important related factors. The survey also examined level of functioning and service utilization associated with illness worry. We hypothesized that illness worry would be associated with physical and psychiatric conditions and with impaired functioning and greater service utilization.
|
METHOD
|
Procedures
The telephone survey involved residents of the continental United States between age 18 and 65 years. Sampling was conducted so that the distribution of respondents was proportional to the distribution of the population by state and county.
The survey was conducted by the Iowa Social Science Institute by means of random-digit dialing of households. Randomly generated telephone numbers across the 48 states were obtained from Survey Sampling, Inc. (Fairfield, CT) The sample was prepared by first identifying working exchanges (i.e., the first three digits of the telephone number), then randomly generating the last four digits. The resulting numbers were then matched against census data to remove problem numbers (e.g., telephone numbers of businesses).
All persons who gave an affirmative answer to a screening question about illness worry and an equal number of randomly selected individuals who gave a negative answer were interviewed. The 15-minute interview was conducted by trained lay interviewers who used a computer-assisted telephone interview system. The interviewers were supervised by Institute staff. In their initial contact with households, interviewers asked to speak with the individual age 18–65 years who had had the most recent birthday. They identified themselves as conducting interviews on behalf of the University of Iowa College of Medicine to learn about attitudes toward illness and medical care within the United States. No substitutions were made within a household; if the identified respondent within a household was not available, an effort was made to call back at an acceptable time. A total of 10 callbacks were made before a person was considered unreachable.
Measures
The interview began with the hypochondriasis module of the Composite International Diagnostic Interview (CIDI).17 The stem question for this module was used to screen subjects. The original question was modified to cover a shorter interval so that thresholds for significant worry might be examined. The modified question was: "During the past 12 months, have you had a period of 1 month or more when you worried most of the time about having a serious illness?" Subjects were then asked what illness they worried about and whether they had seen a doctor. If so, they were asked if they had told the doctor about their worry, if the doctor had said they had the illness they worried about, and if seeing the doctor had reassured them. Subjects were asked how much they were bothered or distressed by their worry (not at all, a little, some, or a lot) and how much worry had interfered with their life or activities (not at all, a little, some, or a lot). Finally, subjects were asked whether their longest period of worry had been at least 6 months.
Subjects were then asked to complete the Illness Worry Scale,18 a nine-item scale that measures the tendency to worry about being ill, to be convinced that one is ill, and to feel more vulnerable to illness than others. Each item calls for a response of yes or no. A score of 3 was used by Robbins and Kirmayer19 to identify hypochondriacal somatizers. The measure has adequate psychometric properties.18
Modules from the CIDI Short-Form20 were used to screen for major depressive episode, panic attacks, and generalized anxiety disorder in the past 12 months. These major disorders were selected for their likely relationship with illness worry. The Short-Form was derived from the full CIDI and contains subsets of screening questions that come closest to duplicating the criteria for the diagnoses in the full CIDI. Classification accuracy, based on data from the National Comorbidity Survey, was 93% or better for the three conditions.20
Traumatic events were assessed with the Childhood Traumatic Events Scale.21 Respondents were asked whether certain traumatic events had occurred before age 17 years and, if so, how traumatic each one had been (1, not at all, to 7, extremely). The events included "death of a very close friend or family member," "major upheaval between parents" (e.g., separation or divorce), "traumatic sexual experience," "victim of violence," "extreme illness or injury," and "other major upheaval." To these we added "extreme illness or injury of a very close friend or family member" and "problem with alcohol or drugs in a parent or family member." The additional events distinguished patients with hypochondriasis from those without in an earlier study.22 The original instrument has proven reliability and validity.21
The SF-12 Health Survey23 was used to assess perceived health status. This measure is a brief alternative to the Medical Outcomes Study Short-Form Health Survey (SF-36).24 The 12-item measure includes a global rating of health (excellent, very good, good, fair, or poor) and subscales for physical and mental health. These subscales are highly correlated with, and share the psychometric properties of, the SF-36. The instrument has been reliably administered by telephone.23
Subjects were asked about demographic characteristics and were asked to report whether a doctor had told them they had any of eight physical conditions. These conditions included high blood pressure, heart disease, emphysema or lung disease, diabetes, cancer, stroke or neurological disease, arthritis or bone or joint disease, and other serious illness. The reliability of such self-reporting was previously proven to be satisfactory.25 In certain analyses, the number of conditions was used as an approximate measure of aggregate physical morbidity.26,27
Analyses
The subjects with illness worry and the comparison subjects without illness worry were first compared descriptively with respect to demographic characteristics, physical conditions, psychiatric conditions, and childhood traumatic events. Differences between groups were examined by using t tests for continuous variables and chi-square (or Fisher’s exact) tests for categorical variables.
A "best" set of factors for discriminating between subjects with illness worry and those without illness worry was selected by using stepwise discriminant regression analysis. Before this analysis was done, candidate predictor variables were reduced in number by using variable-cluster analysis. Because overfitting and uncertain interpretation may occur with conventional model-building procedures that involve many correlated predictor variables,28 we grouped independent nondemographic variables using a clustering procedure, VARCLUS, in SAS version 8.2 (SAS, Cary, NC). VARCLUS is an oblique component analysis related to multiple group factor analyses.29 It clusters variables and should be distinguished from procedures that cluster subjects. Results from the variable-cluster analysis are displayed in Table 1. This analysis showed that the candidate predictor psychiatric conditions comprised one cluster. Physical conditions made up three clusters: heart disease (heart disease, high blood pressure, diabetes), other disease (lung disease, neurological disease, bone or joint disease, other serious diseases), and cancer. Childhood traumatic events made up two clusters: illness or injury (death of a close friend or family member, extreme illness or injury of a close friend or family member) and other traumatic events. For the subgroup analysis limited to subjects with illness worry, the number of conditions divided by the number of possible conditions was used as a summary measure. Also, the number of demographic variables was reduced by dichotomizing race (white versus nonwhite), trichotomizing marital status (single versus married versus separated, divorced or widowed), and assigning integers to increasing levels of education and income.
View this table:
[in this window]
[in a new window]
|
TABLE 1. Candidate Predictors of Illness Worry in the General Population and Summary Measures for Variable Clusters Constituting the Predictors
|
After taking these data reduction steps, we used the SAS procedure MI30 to estimate covariance matrices for independent variables separately for the subjects with illness worry and the comparison subjects. This procedure estimates the covariance matrix by the maximum likelihood method and uses the data from all subjects, not just those without missing values. We then used the estimated covariance matrices as input for a forward stepwise discriminant regression analysis. In this fashion, we were able to identify important predictors using all the data. A logistic regression model was estimated by using age, gender, and the predictors identified in the stepwise discriminant procedure. This logistic regression analysis was based on data from subjects with complete data; few subjects had missing values.
Functional status and health care utilization of the subjects with illness worry and the comparison subjects were compared with adjustment for gender and discriminating risk variables. Ordinary least squares regression or logistic regression analysis was used to assess the relationship between outcome and these predictors, as well as between outcome and group status. In these analyses, the number of physical conditions was used in place of the clusters of conditions used earlier. For health care utilization, it was necessary to transform the distributions of certain variables that were highly skewed. Hospitalizations and emergency room visits were, for this reason, dichotomized. To reduce the influence of outliers, high scores were collapsed into single values for number of physician visits (>7 = 7), number of physicians (>5 = 5), and number of medications (>5 = 5).
Subjects with illness worry were then grouped by whether or not they had the illness that they worried about. To determine which factors were independent predictors of having the illness, a forward stepwise discriminant regression analysis based on covariance matrices was again used, and a logistic regression model was estimated by using age, gender, and the predictors identified in the stepwise procedure. Functional status and health care utilization were also compared for subjects with and without the illness worried about by using procedures described earlier. All analyses were preformed in SAS version 8.2; a test was deemed significant if p was <0.05, and a p value of 0.05 was required for entry and removal for the stepwise procedures.
|
RESULTS
|
Within a sample frame of 5,323 telephone numbers, 1,932 were problem telephone numbers (e.g., business telephones, numbers not in service), 1,210 were not reachable (e.g., unable to contact household, potential subject not available), and 484 were ineligible (e.g., due to age, language). There were 1,697 eligible subjects, of whom 937 (55.2%) participated in the survey. Of these, 123 (13.1%) (95% confidence interval [CI]=11.0–15.5) answered the screening question affirmatively, and 814 (86.9%) answered the question negatively. From those who answered negatively, 123 were randomly selected as comparison subjects.
Of those who reported illness worry, 80.5% (N=99) had seen a physician and 74.0% (N=91) had told that physician about their worry. Most reported that they had been reassured as a result of seeing a physician; only 24.2% (N=24) of those who had seen a physician claimed "little" or "no" reassurance. A total of 70.7% (N=87) of those with illness worry claimed at least "some" distress related to their worry, and 53.7% (N=66) reported that worry about serious illness interfered with their life at least "some." The subjects who reported illness worry scored higher on the Illness Worry Scale (mean=2.8 [SD=2.3] versus mean=1.0 [SD=1.4] for the comparison subjects; t=–7.43, p<0.0001). Relative to the comparison subjects, more subjects with illness worry scored 3 on the Illness Worry Scale (48.8% versus 12.2%;  2=21.43, df=1, p<0.0001).
The demographic characteristics of the subjects with illness worry and the comparison subjects are summarized in Table 2. The subjects with worry had lower household income and were more likely to be women and to be separated, divorced, or widowed. More than one-half of the subjects with worry (54.6%, N=67) had the illness they worried about. As a group, the subjects with worry reported having more physical conditions than did the comparison subjects (mean=1.8 [SD=1.4] versus mean=0.8 [SD=1.0]; t=–6.33, p<0.0001). Specifically, more subjects with illness worry reported having joint disease (48.0% versus 25.2%; 2=13.7, df=1, p=0.0002), high blood pressure (32.5% versus 15.5%; 2=9.8, df=1, p<0.002), lung disease (24.4% versus 11.4%; 2=7.1, df=1, p<0.008), other serious conditions (23.6% versus 9.8%; 2=8.5, df=1, p<0.004), diabetes (14.6% versus 4.1%; 2=8.1, df=1, p<0.005), heart disease (13.8% versus 4.9%; 2=5.8, df=1, p<0.02), cancer (13.0% versus 2.4%; 2=9.6, df=1, p<0.002), and neurological disease (7.3% versus 2.4%; 2=3.2, df=1, p<0.08).
View this table:
[in this window]
[in a new window]
|
TABLE 2. Demographic and Clinical Characteristics of Subjects With Illness Worry and Comparison Subjects Without Illness Worry
|
Table 2 also shows the percentages of subjects with illness worry and comparison subjects who reported childhood traumatic events. The differences between groups were significant for three variables: "extreme illness or injury of a close friend or family member," "problem with alcohol or drugs in a parent or family member," and "victim of violence." When the proportions of subjects who had experienced events (regardless of how traumatic) was examined, "extreme illness or injury of self" was also more likely to be reported by the subjects with illness worry than by the comparison subjects (20.3% versus 10.6%; 2=4.5, df=1, p<0.04).
Table 2 shows the percentages of illness worry subjects and comparison subjects with psychiatric conditions. All of the psychiatric conditions examined—major depressive episode, panic attacks, and generalized anxiety disorder—were significantly more frequent among the subjects with illness worry. Likewise, affirmation of stem (screening) questions about these disorders was more frequent among subjects with worry (major depressive episode 28.5% versus 11.4%; 2=11.2, df=1, p=0.0008; panic attacks 43.9% versus 20.3%; 2=15.7, df=1, p<0.0001; generalized anxiety disorder 28.5% versus 8.1%; 2=17.0, df=1, p<0.0001). In addition, relative to the comparison subjects, the subjects with worry were more likely to have at least one psychiatric condition (38.2% versus 13.8%; 2=19.1, df=1, p<0.0001) and had a greater overall number of conditions (mean=0.6 [SD=1.0] versus mean=0.2 [SD=0.5]; t=4.79, p< 0.0001).
Table 3 shows results of a comparison of subjects with more or less than 6 months of illness worry. Those with a duration of worry of 6 months or more had an earlier age at onset of illness worry and more severe worry. They also reported greater distress and impairment because of illness worry and greater utilization of medical care.
The best predictors of illness worry, as identified by stepwise discriminant regression analysis, were the cluster of psychiatric conditions and the three clusters of physical conditions (i.e., heart disease, other disease, and cancer). The results of a logistic regression analysis of these variables, as well as age and gender, are shown in Table 4. In the table, the odds ratios reported for clusters other than cancer represent the presence of all components of the cluster; if only some components were present, values were correspondingly lower. The c-statistic for the model was 0.77, indicating 77% accuracy in discriminating between the subjects with illness worry and the comparison subjects.
Table 5 shows measures of unadjusted functional status and health care utilization for the subjects with illness worry and the comparison subjects. Because of skewed distributions, nonparametric Wilcoxon rank-sum tests were used for the analysis. The subjects with worry rated their health and physical and mental functioning as worse than did the comparison subjects. Also, the subjects with worry reported more utilization of health care in the past 12 months than did the comparison subjects. All differences in health care utilization, except for the difference in number of medications, remained significant after adjustment for the variables that discriminated between the worry subjects and the comparison subjects.
View this table:
[in this window]
[in a new window]
|
TABLE 5. Impairment in Functioning and Health Care Utilization of Subjects With Illness Worry and Comparison Subjects Without Illness Worry
|
Among the subjects with worry, ratings of functional status were negatively correlated with scores on the Illness Worry Scale. Partial correlations, with adjustment for the number of physical conditions, were r=–0.49 (p<0.0001) for overall rating of health, r=–0.31 (p=0.0005) for physical health, and r=–0.46 (p< 0.0001) for mental health.
When subjects who did and did not have the illness they were worried about were compared, those with the illness worried about had more physical conditions than those without the illness (mean=2.3 [SD=1.5] versus mean=1.1 [SD=1.0]; t=5.47, p<0.0001). Also, the subjects who had the illness they were worried about tended to be older (mean=49.6 years [SD=11.8] versus mean=43.8 years [SD=13.6]; p<0.05) and were more likely to be separated, divorced, or widowed (2=6.06, df=1, p<0.05). In addition, fewer subjects with the illness worried about were worried about cancer, compared to those without the illness worried about (16.4% versus 55.4%; 2=20.57, df=1, p<0.0001). More of the subjects who had the illness they were worried about had major depressive episodes, but the subgroups did not differ with respect to the number of psychiatric conditions (mean=0.8 [SD=1.0] versus mean=0.5 [SD=0.9]; t=1.71, p= 0.09).
Stepwise discriminant regression analysis showed that the presence of the heart disease cluster was the only predictor of having the illness worried about. Logistic regression analysis, with heart disease, age, and gender as independent variables, showed an odds ratio of 2.16n, where n was the number of heart disease cluster components present (95% CI=1.18–3.97, p<0.02). For this model, the c-statistic was 0.64, indicating 64% accuracy in discriminating between those with and without the illness worried about.
The same stepwise discriminant regression procedure was repeated, substituting a measure of aggregate physical morbidity—the total number of physical conditions—for the three physical condition summary variables. Because by definition subjects with the illness worried about had at least one physical condition, we subtracted one from the number of conditions each subject had. At this point, only marital status entered into the stepwise procedure. The logistic regression model for age, gender, and marital status showed a nonsignificant effect of marital status (odds ratio=2.20, 95% CI=0.92–5.25, p<0.08).
Finally, we compared functional status and health care utilization of illness worry subjects with and without the illness they worried about after adjustment for age, gender, and number of physical conditions. Subjects who had the illness they were worried about rated their physical health as worse than did those without the illness they were worried about (mean=39.2 versus mean=44.5; p<0.008). Also, those with the illness they were worried about reported having seen more doctors (mean=3.0 versus mean=2.1; p<0.009), making more doctor visits (mean=5.3 versus mean=3.7; p=0.0009), and taking more medications (mean=3.2 versus mean=2.1; p= 0.0005). After adjustment for age, gender, and number of physical conditions, those who had the illness they were worried about also had more emergency visits (odds ratio=2.51, 95% CI=1.10–5.73, p<0.03) and were more likely to have been hospitalized, although this result only approached significance (odds ratio=2.61, 95% CI= 0.99–6.88, p=0.053).
|
DISCUSSION
|
We identified worry about serious illness lasting at least 1 month in 13.1% of the surveyed population; 6.9% of the population had illness worry that persisted for 6 months or more, 7.0% reported significant distress or impairment related to illness worry, and 6.4% scored 3 on the Illness Worry Scale, a minimal threshold for caseness. Our prevalence estimate for worry lasting 6 months or more is similar to the estimates of 6% and 10% found in previous surveys.6,7 Almost none of the worried subjects in this study met the full criteria for hypochondriasis as assessed by the CIDI. The CIDI criteria for hypochondriasis require a subject to have seen a physician, been told that he or she does not have the illness that is the focus of the worry, and failed to respond to reassurance. In our study, most of the subjects with worry had communicated their worry to a physician, but many had the illness they were worried about and had received some reassurance from the physician. However, even though the subjects in our study did not meet the diagnostic criteria, they showed functional impairment and increased utilization of care, indicating that their worry was a significant problem.
With respect to related factors, illness worry was associated with current physical and psychiatric conditions. Looper and Kirmayer6 reported similar findings in their survey of hypochondriacal concerns in a Canadian population, and Roht et al.31 observed a relationship between chronic medical conditions and similar concerns among persons living near hazardous waste sites. Increased rates of medical illness have also been found among primary care outpatients with hypochondriasis.9–12 Several explanations for the association between illness worry and physical conditions are possible. Some subjects may have responded affirmatively to our screening question on the sole basis of having an illness they viewed as serious. Another possibility is that medical illness is an important contributor to illness worry. In support of this interpretation, several studies have found that serious illness or injury in childhood is associated with hypochondriasis or somatization in adults.22,32,33 Exposure to illness in friends or family members may contribute to similar outcomes.22,34,35 Exposures of this kind may give rise to dysfunctional beliefs and attitudes, including worry about illness.36,37 Still another possibility is that illness worry and physical illness are independent consequences of individual characteristics such as neuroticism and less education.38
Psychiatric conditions were included in the final model of illness worry. In this survey, these disturbances were interrelated, meaning that subjects with one psychiatric condition were likely to have another. Although a minority of subjects with illness worry had such conditions, many more of them endorsed stem questions indicating the presence of some symptoms. Looper and Kirmayer6 also observed depressive and anxiety symptoms in a high percentage of illness worry subjects, and Noyes et al.4 found hypochondriasis associated with psychiatric symptoms and disorders in a nonclinical population. Also, greater proportions of hypochondriacal patients in primary care have depressive, anxiety, and other somatoform disorders.15,16 The relationship between hypochondriacal concerns and psychiatric symptoms has been the subject of speculation and may fit more than one model. However, strong evidence exists that persons with high levels of neuroticism or negative affectivity experience psychiatric distress of various kinds, including hypochondriasis.39
We showed that illness worry is associated with impairment in health status and functioning beyond that explained by physical conditions. This finding was true for subjects who had the illness they were worried about as well as for those who did not. Thus, even at a threshold of 1 month, such worry appears to have a significant effect on perceived health and functioning. Looper and Kirmayer6 also reported greater disability among persons with illness worry, compared to those without such worry. They observed that worried subjects had more days when they were unable to perform usual activities because of somatic symptoms. Similarly, Noyes et al.4 observed poorer physical functioning and work performance among hypochondriacal family members. Level of impairment has a bearing on how hypochondriasis should be defined. A 6-month duration of illness worry adds substantially to impairment, and this duration appears to be an appropriate threshold for the disorder.
Illness worry was associated with increased health care utilization after adjustment for physical conditions. This increased utilization was reflected in the higher numbers of physicians, physician visits, emergency visits, and hospitalizations among subjects with illness worry. Looper and Kirmayer6 also showed increased help-seeking related to illness worry. Likewise, Noyes et al.4 observed that hypochondriacal family members used more health services than did nonhypochondriacal family members. Similarly, most studies of hypochondriasis in clinical populations have documented increased use of medical services.2 Some of this help-seeking is aimed at reassurance, which, according to DSM-IV, the patient with hypochondriasis fails to secure. However, in our survey, most worried subjects received some reassurance. This finding suggests that resistance to reassurance is not a consistent characteristic of illness worry. Gureje et al.12 came to a similar conclusion with respect to hypochondriasis. They found that primary care patients who met the criteria for this disorder except for "persistent refusal to accept medical reassurance" differed little from those who met the full criteria.
More than one-half of the subjects with illness worry in our study had the illness they were worried about. Those with the illness had more physical conditions and were older. Also, their worry was more severe and persistent, even though it was less likely to be focused on cancer. By virtue of having the illness they worried about, they had more physical conditions than the remaining subjects. However, after adjustment for the number of such conditions, age, and gender, no important differences in related factors remained. Consequently, we found no clear basis on which to distinguish these subgroups. Of course, some subjects who had the illness they were worried about may have had normal worry. At present no satisfactory definition of excessive illness worry exists. Nevertheless, those who had the illness they were worried about appeared to have had a more persistent and unavoidable stimulus for worry, which may have contributed to the greater severity and chronicity observed. Regardless, subjects who focused on existing illness appeared to have significant worry associated with excess morbidity and health services utilization.
The strengths of this study include the use of a representative nationwide sample and examination of risk factors by using multivariate statistics. The limitations include a moderately high rate of refusals, reliance on a telephone interview, and use of an approximate measure of aggregate illness severity. Without information on nonparticipants, it is difficult to judge the representativeness of the sample, and without a higher rate of participation, generalization of findings to the entire population is uncertain. In addition, medical information obtained by telephone can be difficult to evaluate. It is not always clear, for example, how appropriate the evaluation of unexplained symptoms may have been. Finally, recall of illness worry in the past 12 months may have been influenced by current state and may have been inaccurate. Nevertheless, the study shows that persistent illness worry associated with impairment in functioning is common in the general population. Physical and psychiatric disorders appear to be important factors in the development and/or persistence of illness worry. Indeed, chronic physical illness appears to be a common focus of worry that may be severe and persistent in some persons. Prospective study of patients who develop particular illnesses should make it possible to determine the relative contribution of individual characteristics (e.g., personality), illness variables (e.g., severity), environmental factors (e.g., perceived stress, social support), and quality of health care (e.g., doctor-patient relationship) to the development of illness worry and hypochondriasis.
|
REFERENCES
|
- American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington, DC, APA, 1994
- Noyes R: Epidemiology of hypochondriasis, in Hypochondriasis: Modern Perspectives on an Ancient Malady. Edited by Starcevic V, Lipsitt DR. New York, Oxford University Press, 2001, pp 127–154
- Faravelli C, Salvatori S, Galassi F, Aiazzi L, Drei C, Cabras P: Epidemiology of somatoform disorders: a community survey in Florence. Soc Psychiatry Psychiatr Epidemiol 1997; 32:24–29[CrossRef][Medline]
- Noyes R, Happel RL, Yagla SJ: Correlates of hypochondriasis in a nonclinical population. Psychosomatics 1999; 40:461–469[Abstract/Free Full Text]
- Lieb R, Pfister H, Mastaler M, Wittchen H-U: Somatoform syndromes and disorders in a representative population sample of adolescents and young adults: prevalence, comorbidity, and impairments. Acta Psychiatr Scand 2000; 101:194–208[CrossRef][Medline]
- Looper KJ, Kirmayer LJ: Hypochondriacal concerns in a community population. Psychol Med 2001; 31:577–584[CrossRef][Medline]
- Reif W, Hessel A, Braehler E: Somatization symptoms and hypochondriacal features in the general population. Psychosom Med 2001; 63:595–602[Abstract/Free Full Text]
- Speckens AEM, van Hemert AM, Spinhoven P, Bolk JH: The diagnostic and prognostic significance of the Whiteley Index, the Illness Attitudes Scale, and the Somatosensory Amplification Scale. Psychol Med 1996; 26:1085–1090[Medline]
- Barsky AJ, Wyshak G, Klerman GL: Transient hypochondriasis. Arch Gen Psychiatry 1990; 47:746–752[Abstract/Free Full Text]
- Barsky AJ, Wyshak G, Latham KS, Klerman GL: The relationship between hypochondriasis and medical illness. Arch Intern Med 1991; 151:84–88[Abstract/Free Full Text]
- Robbins JM, Kirmayer LJ: Transient and persistent hypochondriacal worry in primary care. Psychol Med 1996; 26:575–589[Medline]
- Gureje O, Ustun TB, Simon GE: The syndrome of hypochondriasis: a cross-national study in primary care. Psychol Med 1997; 27:1001–1010[CrossRef][Medline]
- Altamura AC, Garta MG, Tacochini G, Musazzi A, Pioli MR, Italian Collaborative Group on Somatoform Disorders: Prevalence of somatoform disorders in a psychiatric population: an Italian nationwide survey. Eur Arch Psychiatry Clin Neurosci 1998; 248:267–271[CrossRef][Medline]
- Eminson M, Benjamin S, Shortall A, Woods T, Faragher B: Physical symptoms and illness attitudes in adolescents: an epidemiological study. J Child Psychol Psychiatry 1996; 37:519–528[Medline]
- Barsky AJ, Wyshak G, Klerman GL: Psychiatric comorbidity in DSM-III-R hypochondriasis. Arch Gen Psychiatry 1992; 49:101–108[Abstract/Free Full Text]
- Noyes R, Kathol RG, Fisher M, Phillips B, Suelzer M, Woodman C: Psychiatric comorbidity among patients with hypochondriasis. Gen Hosp Psychiatry 1994; 15:78–87
- World Health Organization: Composite International Diagnostic Interview (CIDI), version 1.0. Geneva, WHO, 1990
- Robbins JM, Kirmayer LJ, Kapusta MA: Illness worry and disability in fibromyalgia syndrome. Int J Psychiatry Med 1990; 20:49–63[Medline]
- Robbins JM, Kirmayer LJ: Cognitive and social factors in somatization, in Current Concepts of Somatization: Research and Clinical Perspectives, Edited by Kirmayer LJ, Robbins JM. Washington, DC, American Psychiatric Press, 1991, pp 107–141
- Kessler RC, Andrews G, Mroczek D, Ustun B, Wittchen H-U: The World Health Organization Composite International Diagnostic Interview Short-Form (CIDI-SF). Int J Methods Psychiatr Res 1997; 7:171–185
- Pennebaker JN, Susman JR: Disclosure of traumas and psychosomatic processes. Soc Sci Med 1988; 26:327–332
- Noyes R, Stuart S, Langbehn DR, Happel RL, Longley SL, Yagla SJ: Childhood antecedents of hypochondriasis. Psychosomatics 2002; 43:282–289[Abstract/Free Full Text]
- Ware JE, Snow KK, Kosinski M, Gandek C: SF-12 Health Survey: Manual and Interpretation Guide. Boston, Health Institute, New England Medical Center, 1995
- Stewart AL, Hays RD, Ware JE Jr: The MOS Short-Form General Health Survey: reliability and validity in a patient population. Med Care 1988; 26:724–735[Medline]
- Eisen SA, Goldberg J, True WR, Henderson WG: A co-twin control study of the effects of the Vietnam War on the self-reported physical health of veterans. Am J Epidemiol 1991; 143:49–58
- Wells KB, Rogers W, Burnam A, Greenfield S, Ware JE: How the medical comorbidity of depressed patients differs across health care settings: results from the medical outcomes study. Am J Psychiatry 1991; 48:1688–1696
- Patrick DL, Erickson P: Health Status and Health Policy: Allocating Resources to Health Care. New York, Oxford University Press, 1993
- Harrell FE, Lee KL: Regression modeling strategies for improved prognostic prediction. Stat Med 1984; 15:361–387
- Harman HH: Modern Factor Analysis, 3rd ed. Chicago, Chicago University Press, 1976
- SAS/STAT User Guide, Version 8. Cary, NC, SAS Institute, 1999
- Roht LH, Vernon SW, Weir FN, Pier SM, Sullivan D, Reed LJ: Community exposure to hazardous waste disposal sites: assessing reporting bias. Am J Epidemiol 1985; 122:418–433[Abstract/Free Full Text]
- Craig TKJ, Broadman AP, Mills K, Daly-Jones O, Drake H: The South London Somatization Study: I. longitudinal course and the influence of early life experiences. Br J Psychiatry 1993; 163:579–588[Abstract/Free Full Text]
- Barsky AJ, Wool C, Barnett MC, Cleary PD: Histories of childhood trauma in adult hypochondriacal patients. Am J Psychiatry 1994; 151:397–401[Abstract]
- Benjamin S, Eminson DM: Abnormal illness behavior: childhood experiences and long-term consequences. Int Rev Psychiatry 1992; 4:55–70
- Hotopf M, Mayou R, Wadsworth M, Wessely S: Childhood risk factors for adults with medically unexplained symptoms: results from a national birth cohort study. Am J Psychiatry 1999; 156:1796–1800[Medline]
- Leventhal H, Nerenz DR, Steele DJ: Illness representations and coping with health threats, in Handbook of Psychology and Health, Vol 4. Edited by Baum A, Taylor SE, Singer JE. Hillsdale, NJ, Erlbaum, 1984, pp 219–252
- Sensky T: Cognitive therapy in physical illness, in Psychological Treatment in Disease and Illness. Edited by Hodes M, Moorley S. London, Gaskell, 1993, pp 30–45
- Neeleman J, Ormel J, Bijl RV: The distribution of psychiatric and somatic ill health: associations with personality and socioeconomic status. Psychosom Med 2001; 63:239–247[Abstract/Free Full Text]
- Pennebaker JW, Watson D: The psychology of somatic symptoms, in Current Concepts of Somatization: Research and Clinical Perspectives. Edited by Kirmayer JL, Robbins JM. Washington, DC, American Psychiatric Press, 1991, pp 21–35
This article has been cited by other articles:
|
|
|
|
 
A. Martin and F. Jacobi
Features of hypochondriasis and illness worry in the general population in Germany.
Psychosom Med,
September 1, 2006;
68(5):
770 - 777.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
How Sick Are the Worried Well?
Journal Watch Psychiatry,
March 8, 2006;
2006(308):
8 - 8.
[Full Text]
|
|
|
|
|
|
|
How Sick Are the Worried Well?
Journal Watch Women's Health,
January 10, 2006;
2006(110):
8 - 8.
[Full Text]
|
|
|
Get information about faster international access.
Privacy Policy
Copyright © 2005
Academy of Psychosomatic Medicine.
All rights reserved.
Home
| Search
| Current Issue
| Past Issues
| Subscribe
| All APPI Journals
| Help
| Contact Us
|
TOP
ABSTRACT
INTRODUCTION
