Skip to main content
Download PDF

Joint factor analysis and approximate equipercentile linking of common trait health anxiety measures: a cross-sectional study of the 14-, 18- and 64-item health anxiety inventory, the illness attitude scale, and the 14-item Whiteley Index

BMC Psychiatry volume 23, Article number: 658 (2023) Cite this article

Abstract

Background

Research on health anxiety has bloomed in recent years, but summaries of the literature are complicated by the use of dissimilar self-report questionnaires. Furthermore, these instruments have rarely been administered in parallel, and especially not in clinical samples. In this study, we aimed to investigate the relationship between five widespread health anxiety measures, and to draft guidelines for the conversion of different sum scores.

Methods

Clinical trial participants with principal pathological health anxiety (n = 335) and a sample of healthy volunteers (n = 88) completed the 14-item Whiteley Index (WI-14), the Illness Attitude Scale (IAS), and the 14-, 18-, and 64-item Health Anxiety Inventory (the HAI-64, HAI-18, and HAI-14). Cross-sectional data from all participants were pooled (N = 423) and we conducted a joint factor analysis and approximate equipercentile linking of the WI-14, IAS, HAI-64, HAI-18, and HAI-14.

Results

Inter-scale correlations were high (rs ≥ 0.90 and ≥ 0.88 in adjusted analyses), and the scree plot of the joint factor analysis spoke for a unifactorial solution where 89/105 items (85%) had loadings ≥ 0.40. Most items at the core of this broad trait health anxiety factor pertained to the worry about health, the fear of having or developing a serious disease, and to some extent bodily preoccupation. We present a cross-walk table of observed equipercentile linked sum scores.

Conclusions

This study speaks clearly in favor of the WI-14, IAS, HAI-64, HAI-18, and HAI-14 all tapping into the same trait health anxiety construct, the core of which appears to concern the worry about health, the fear of having or developing a serious disease, and to some extent bodily preoccupation. Based on recently reported cut-offs for the HAI-14, a reasonable cutoff for pathological health anxiety in a psychiatric setting probably lies around 7–8 on the WI-14, 52–53 on the IAS, 82–83 on the HAI-64, and 26–27 on the HAI-18.

Trial registration

ClinicalTrials.gov NCT01966705, NCT02314065.

Peer Review reports

Introduction

Health anxiety is a multifaceted psychological trait that is primarily characterized by a fear of, or preoccupation with, having or developing a serious health condition such as terminal cancer, a severe cardiovascular disease, or a progressive neurological disorder [1]. Higher levels of trait health anxiety commonly imply increased anxiety sensitivity and intolerance of uncertainty [2], increased bodily preoccupation [3], increased frequency and intensity of behaviors aimed at reducing health anxiety in the short term [4], and a more pronounced fear of death [5]. In the general population, trait health anxiety varies on a continuum, from benign levels of concern about health, to pathological health anxiety that is both recurrent and excessive [6,7,8,9]. Based on its clinical characteristics, pathological health anxiety can be considered a de facto anxiety or perhaps obsessive-compulsive spectrum disorder [2, 10,11,12,13,14].

Over the past decades, research focusing on health anxiety has been growing rapidly for several reasons; one being the relative success of the cognitive-behavioral framework in guiding treatment efforts and promoting interest in experimental work [11, 15]. Other factors sparking an interest in health anxiety include a rise in work focusing on the role of information technology in exacerbating health concerns [16], emergent discrepancies in widespread diagnostic taxonomies [14, 17], and an increased interest in the role of psychological factors in somatic disease including covid-19 [18, 19]. However, an obstacle to the interpretation of this increasingly diverse research field is that a large number of different trait health anxiety measures are in circulation, and that these are not easily compared. Thus, it is often difficult to determine whether participants of different studies suffer from similar levels of trait health anxiety, and if so, whether these levels are indicative of pathological health anxiety.

Self-report measures of trait health anxiety have existed since the mid-1960s. The 14-item Whiteley Index with dichotomous (“yes”/”no”) items (WI-14) was developed based on hospital staff definitions of “hypochondriasis” [20]. Its psychometric properties are usually found to be acceptable but not ideal, and revised versions with Likert-type items are now more common [21]. The Illness Attitude Scale (IAS; and sometimes referred to in the plural) was probably the most widely used measure of trait health anxiety from about the late 1980s to the early 2000s [22], and was developed on the basis of statements by patients who exhibited abnormal illness behavior or believed they had an undiagnosed disease [23]. Its psychometric properties are usually found to be good, though the factor structure is disputed [24]. The Health Anxiety Inventory [1] is perhaps the most widely used self-report measure of trait health anxiety today, and exists in many forms, the three most common probably being the 64-, 18-, and 14-item versions (HAI-64, HAI-18, and HAI-14). This questionnaire was developed to capture the cognitive and emotional components of DSM-IV hypochondriasis which was the prototypical pathological health anxiety diagnosis up until 2013. Common versions of the Health Anxiety Inventory are all widely believed to possess good to excellent psychometric properties [1, 25]. In a recent study, a cutoff of 22 on the HAI-14 was found to be appropriate for identifying patients with pathological health anxiety in the psychiatric setting [26]. When the respondent is known to suffer from pathological health anxiety, a score of 28 or higher is indicative of moderate symptoms, and 33 of substantial symptoms [26]. In summary, the WI-14, IAS, HAI-64, HAI-18 and HAI-14 are all examples of widespread measures of trait health anxiety, but knowledge about how scores can be converted from one measure to another is lacking.

In this study, we aimed to investigate the relationship between the WI-14, IAS, HAI-64, HAI-18, and HAI-14 using a composite dataset derived from two clinical trials for pathological health anxiety and a sample of healthy volunteers. In a joint factor analysis, we aimed to test our hypothesis that the scales would be highly correlated and tap into the same latent trait health anxiety construct. Should such a broad latent trait health anxiety factor be present, we aimed to determine what cognitive, emotional, and behavioral characteristics that lie at the core of this factor. Furthermore, we intended to relate the sum scores of the WI-14, IAS, HAI-64, HAI-18, and HAI-14 by means of equipercentile linking. Using the resulting linking table, we intended to make use of recently developed guidelines for interpreting the HAI-14 [26] so as to draw conclusions about approximate cut-offs and guidelines for interpreting severity in terms of the WI-14, IAS, HAI-64, and HAI-18.

Methods

Design

This was a psychometric study based on cross-sectional data from a composite aduld sample (pooled N = 423) of 335 adult participants of two clinical trials of cognitive behavior therapy for pathological health anxiety [27, 28] and 88 healthy volunteers recruited via newspaper advertisements [29]. Notably, the two clinical trials included 336 participants but 1 was dropped from the present study due to missing WI-14 data. This study was a collaboration between Gustavsberg Primary Care Clinic and Karolinska Institutet, Stockholm, Sweden. All procedures were approved by the regional ethics review board of Stockholm (2013/375 − 31/5, 2014/1530-31/2), all participants gave informed consent to participate in research, and both clinical trials were preregistered at ClinicalTrials.gov (NCT01966705, NCT02314065).

Procedure

All participants in the clinical trials exhibited a fear of, or preoccupation with, severe illness and met full criteria for a principal diagnosis of DSM-5 somatic symptom disorder or illness anxiety disorder as determined by a clinical psychologist aided by the Health Preoccupation Diagnostic Interview [HPDI; 30] and the Mini-International Neuropsychiatric Interview [MINI; 31]. The main exclusion criteria were a serious somatic condition, a substance use disorder, a psychotic disorder, a bipolar disorder, severe depression, and recurrent suicidal ideation. The healthy volunteers were assessed using the MINI and included only if found to be healthy. Prior to the eligibility interview, all 423 participants (both the clinical trial participants and the healthy volunteers), completed the self-report trait health anxiety measures as listed below.

Outcomes

We administered the HAI-64, IAS, and WI-14 online in Swedish, using previously evaluated translations [1, 20, 23, 29]. Participants completed the questionnaires via their web browser, with black text on white background and radio buttons to mark responses. On the HAI-64, each of the 64 items renders a score of 0–3 and the respondent is encouraged to select one of four statements that best corresponds to their level of trait health anxiety (e.g., from “I do not worry about my health” to “I spend most of my time worrying about my health”). In this study, the HAI-64 had a theoretical range of 0-192 and was also rescored as the HAI-18 with a range of 0–54, and the HAI-14 with a range of 0–42, so as to enable approximate linking. The main difference, besides the number of items, between the HAI-14 on the one hand and the HAI-64 and HAI-18 on the other, is that the latter versions include a “negative consequences” subscale, which measures the perceived negative consequences of developing a serious disease. In this study, internal consistency was excellent for the sum scales of all versions of the Health Anxiety Inventory, i.e., the HAI-64 (α = 0.99), HAI-18 (α = 0.97), and HAI-14 (α = 0.97). On the IAS, 27 items are each scored 0–4, and responses indicate the frequency of various experiences pertaining to “worry about illness”, “concerns about pain”, “health habits”, “hypochondriacal beliefs”, “tanatophobia”, “disease phobia”, “bodily preoccupations”, “treatment experience”, and “effects of symptoms”, giving the sum score a theoretical range of 0-108. The internal consistency of the IAS in this study was excellent (α = 0.97). On the WI-14, each item is scored 0 (“no”) or 1 (“yes”), resulting in a theoretical range of 0–14. The internal consistency of the WI-14 in this study was excellent (α = 0.94).

Statistical analysis

We conducted all statistical analyses in Stata 15.1. First, we validated that the trait health anxiety scales measured the latent construct, and were suitable for equipercentile linking. We calculated Pearson correlations and conducted a joint factor analysis of all (64 + 27 + 14 = 105) items, based on principal axis factoring with promax rotation. Considering that we expected factor loadings to be strong, factors to be few, and there to be many items per factor, we regarded the sample size of 423 as sufficient for this purpose [32, 33]. Because the HAI-64 is considerably longer than the IAS and WI-14, as a sensitivity analysis, we also conducted a secondary factor analysis that only included the items of the HAI-14, IAS, and WI-14. When we had established that all questionnaires tapped into the same latent trait health anxiety construct, we proceeded to equipercentile linking of the WI-14, IAS, HAI-64, HAI-18, and HAI-14. Equipercentile linking is a procedure whereby scores on scales measuring the same thing are linked by means of percentiles, so that scores on two scales are assumed to be equivalent if they correspond to the same percentile of each respective scale distribution. As is commonly done for discrete scales [34], we defined the percentile of each score as the percentage of participants scoring below that score, plus the percentage of participants having exactly that score divided by two. We linked sum scores in the ranges that were observed (represented in the sample), and based on presmoothed frequency distributions that allowed us to model all sum scores (including those not directly observed) within each such range. As a sensitivity analysis, we also report linked sum scores based on observed (non-smoothed) frequency distributions as supplementary material.

Results

Sample characteristics

The pathological health anxiety participants completed their questionnaires between December 4th 2013 and March 22nd 2014 (n = 131), and December 10th 2014 and March 15th 2017 (n = 204), in each respective clinical trial, and the healthy volunteers completed their questionnaires between March 19th 2014 and June 3rd 2014 (n = 88). A typical pathological health anxiety participant was a 38 years old (SD = 12) female (240/335, 72%) with a tertiary education (259/335, 77%) who scored in the moderate pathological range of trait health anxiety (Table 1). In this subsample, 64/335 (19%) had comorbid depression and 207/335 (62%) met full criteria for at least one comorbid anxiety disorder, post-traumatic stress disorder, or obsessive-compulsive disorder. A typical healthy volunteer was a 49 years old (SD = 18) female (59/88, 67%) with a tertiary education (80/88, 91%) who scored low on trait health anxiety (Table 1). None of the healthy volunteers was found to meet criteria for a psychiatric disorder. In the pooled sample (N = 423), the distribution of trait health anxiety scores covered most but not all theoretically possible sum scores (the WI-14 being the exception where all scores, 0–14, were observed; see the supplementary material for details).

Table 1 Sample trait health anxiety sum scores
Full size table

Joint factor analysis and feasibility of equipercentile linking

In the pooled sample, Pearson correlations between the trait health anxiety measures were all high (≥ 0.90 and ≥ 0.88 in adjusted analyses; Table 2) and the Kaiser-Meyer-Olkin (KMO) statistic was indicative of adequate sampling (0.98). Joint factor analysis of the HAI-64, IAS, and WI-14 pointed to one factor being clearly dominant, as is illustrated in the scree plot in Figure S1. All three scales contributed with items that received factor loadings in at least the high 0.80s. The common theme of these core items appeared to be worry about health and the fear of having or developing a serious disease, notable examples being HAI-64 #8 (i.e., HAI-14 and HAI-18 #5: “I am […] afraid that I have a serious illness”), IAS #1 (“Do you worry about your health?“), and WI-14 #1 and #4 (“Do you often worry about the possibility that you have got a serious illness?“, “Do you worry a lot about your health?“). Weak items were primarily found in the HAI-64 Negative Consequences (NC) subset (i.e., #48 onwards), and were primarily items that concerned social interaction in the case of serious illness (Tables S1 and S2). When two factors were retained in the analysis, most of the 17 NC items mapped onto a relatively weak second factor (Tables S3 and S4). Overall, however, this factor analysis indicated that the WI-14, IAS, HAI-64, HAI-18, and HAI-14 tap into one and the same strong trait health anxiety factor as their primary source of variance, and results were similar when the factor analysis included the items of the WI-14, IAS, and HAI-14 only (Figure S2, Tables S5 and S6).

Table 2 Pearson correlation matrix of trait health anxiety measures
Full size table

Equipercentile linking of trait health anxiety sum scores

Because all trait health anxiety scales appeared to be closely associated, we proceeded to equipercentile linking of their sum scores. Table 3 can be used for approximate linking of sum scores on the WI-14, IAS, HAI-64, HAI-18, and HAI-14. With the help of Table 3, each trait health anxiety score that was observed in the present study can be linked to its percentile, which in turn can be linked to a score on another trait health anxiety scale. Thus, for example, a score of 22 on the HAI-14 corresponds to approximately 26–27 on the HAI-18. Linking based on non-presmoothed frequency distributions resulted in relatively similar outcomes.

Table 3 Cross-walk table of equipercentile linked scores on trait health anxiety measures
Full size table

Discussion

This study was an unusual attempt at a joint analysis of five common trait health anxiety self-report questionnaires. We found that the WI-14, IAS, HAI-64, HAI-18, and HAI-14 all loaded heavily on the same broad latent trait health anxiety factor, as illustrated by the fact that out of 105 items in total, 35 had loadings ≥ 0.80 and 89 had loadings ≥ 0.40. A strength of the present study is that all non-healthy participants had pathological health anxiety as opposed to other primary psychopathologies, so that a clear gradient in trait health anxiety, without substantial interference of partially overlapping constructs such as somatic disease and panic disorder symptoms, could be modelled for the purpose of factor analysis and the linking of sum scores. The strong unifactorial solution seen in this study speaks in favor of the linking of sum scores derived from the WI-14, IAS, HAI-64, HAI-18, and HAI-14.

Based on estimates derived from the cross-walk table (Table 3), an HAI-14 cutoff of 22 to screen for pathological health anxiety in the psychiatric setting [26] corresponds to a score of ca. 26–27 on the HAI-18, 82–83 on the HAI-64, 52–53 on the IAS, and 7–8 on the WI-14. These tentative cutoffs are slightly higher than those previously reported [35], most probably because the present estimates are derived from an analysis where consecutive psychiatric patients constituted the reference group [26] whereas previous estimates were derived from a study that employed a more pragmatic reference group [35]. In respondents with confirmed pathological health anxiety, based on the recent suggestion that scores below 28 on the HAI-14 are indicative of mild symptoms [26], the same could be said of scores below ca. 33–34 on the HAI-18, 105–106 on the HAI-64, 67–68 on the IAS, and 10–11 on the WI-14. Similar to scores of at least 33 on the HAI-14 [26], symptoms are probably to be regarded substantial even within the clinical range if the respondent scores at least ca. 40–41 on the HAI-18, 128–129 on the HAI-64, 80–81 on the IAS, or 12–13 on the WI-14. These approximate conversions and rule of thumb guidelines for the interpretation of scores derived from common trait health anxiety scales could be of use both in research and the clinic.

In this study, items that pertained to worrying about health and fearing the prospect of having or developing a serious disease were at the heart of the trait health anxiety construct. One implication of this finding is that, should a minimal set of questions be used for the purpose of identifying individuals with pathological health anxiety, for example by the general practitioner or as part of a screening procedure, these questions should ideally focus on worrying about health and the fear of having or developing a serious disease. Thus, for example, it would probably be more fruitful to ask “Would you say that you worry a lot about your health, and the possibility of having a serious disease?” than to ask about repeated doctor shopping or other aspects of the trait health anxiety construct. Importantly, this view of health worries and the fear of disease as the core of the trait health anxiety construct contrasts with certain widespread conceptualizations of pathological health anxiety, such as the ICD-11 hypochondriasis diagnosis, which focuses more on the presence of excessive health-related behaviors. Results for bodily preoccupation were mixed, and our impression is that items tapping into worry such as IAS #20 (“When you notice a sensation in your body, do you find it difficult to think of something else?”, 0.90–0.94) had substantial factor loadings whereas items that conceived of bodily preoccupation in terms of perceptual changes or “being aware” of the body only had slightly weaker or even modest loadings (e.g., HAI-64 #3: “I am constantly aware of bodily sensations or changes”, 0.74–0.75; WI-14 #3: “Do you find that you are often aware of various things happening in your body?”, 0.53–0.58). Items pertaining to the fear of death (IAS #13–14) showed mixed results in the 0.59–0.78 range, thus only partly corroborating the common view that the fear of death is an integral component of pathological health anxiety [5, 36]. Interestingly, not responding to medical reassurance (IAS #11, WI-14 #10), which used to be a criterion for hypochondriasis during the DSM-IV era, had relatively weak loadings (0.50–0.59). One possible explanation for this is that these medical reassurance items were phrased in a manner that led participants to reply based on whether they usually feel reassured in the very short term, as opposed to whether this reduction tends to persist for a longer period of time (“When your doctor tells you that you have no physical disease to account for your symptoms, do you refuse to believe him?”, “Is it hard for you to believe the doctor when he tells you there is nothing for you to worry about?”). The focus on this short time frame is unfortunate, considering that the cognitive-behavioral view of pathological health anxiety would contend that it is not so much the lack of a short-term reduction in anxiety that is characteristic of this clinical problem, but rather the reduced likelihood of a reduction in health anxiety being maintained over time [e.g., 37, 38]. Generally speaking, many items pertaining to overt behaviors such as symptom checking, reassurance seeking, and various avoidance behaviors were also in the lower range, highlighting that whereas high levels of trait health anxiety always imply an increased fear of or preoccupation with having or developing a serious disease, different individuals engage in different behaviors in the hope of reducing or managing this problem in the short term [4]. For example, an individual worrying about skin cancer may be more inclined to “examine [his or her] body for disease” (IAS #9), than an individual worrying about a severe congenital heart defect or pancreatic cancer. Interestingly, both being afraid of seeking health care (HAI-64 #22) and the inclination to seek healthcare (IAS #23 and #24) loaded on the same latent trait health anxiety trait which corroborates the previous finding that both patterns are common, and may even be found in the same individual and fluctuate over time [39]. Abnormal health behaviors are probably important for trait health anxiety, but measuring these using specific examples results in clear psychometric challenges.

We are aware of one previous study where more than one health anxiety scale was included in the same factor analysis [40]. In that study, 503 undergraduate students completed the HAI-18 and the Multidimensional Inventory of Hypochondriacal Traits [MIHT; 41]. When a second-order health anxiety factor was added, the MIHT Affective/Worry subfactor which focuses on worry about health and the fear about serious illness had the strongest factor loading (0.96), followed by a factor representing the first 14 items of the HAI-18 (i.e., the HAI-14; 0.83). Similar to the present study, the MIHT Behavioral/Reassurance subfactor had a modest factor loading of 0.55, and the Negative Consequences (NC) items of the HAI-18 did worse (0.51). Thus, on the whole, the outcome of the previous study was similar to the present one in the sense that the worry about health and the fear of serious illness was at the core of the trait health anxiety construct [40], which speaks for the validity of our findings. The fact that many of the HAI-64 NC items had relatively weak factor loadings on the broad trait health anxiety factor in both studies also has clinical implications in that this speaks for further use of the HAI-14 rather than the HAI-64 or HAI-18 if construct validity in the field as a whole is to be promoted. Simply put, the HAI-14 appears to focus on the core aspects of trait health anxiety.

This study had notable strengths. Several trait health anxiety questionnaires were administered in parallel which is unusual, especially in clinical samples. Furthermore, data could be derived from a combination of healthy volunteers and patients with pathological health anxiety which means that the full range of trait health anxiety scores were available for analysis. This study also had limitations. Participants were primarily self-referred, reported high average educational attainment, and were primarily female. This implies a threat that results may not generalize as well to populations that are not actively seeking treatment, that are less educated, and that are primarily male. Notably, relatively little is known about measurement invariance with regard to psychometric measures of health anxiety (for one of few noteworthy investigations, see MacSwain et al., 2009 [42]). This means that it is not clear to what degree the WI-14, IAS, HAI-64, HAI-18, and HAI-14 behave differently in psychometric terms for example as a function of demographic characteristics. Another limitation is that the HAI-14 and HAI-18 were scored from the corresponding items of the HAI-64 as opposed to administered separately, which may have affected the outcome due to intermediate items giving rise to framing and ordering effects [e.g., 43]. Because the HAI-64, IAS, and WI-14 were not administered in weighted or random order, ordering effects could also have affected the study overall, which highlights the preliminary nature of these findings. A limitation of this study is also that the full theoretical ranges of sum scores were not observed. Specifically, none of the participants had a score of 42 on the HAI-14, 54 on the HAI-18, 0–8 or 175–192 on the HAI-64, or 0–4 or 104–108 on the IAS. These sum score ranges were therefore not modelled. Last, we wish to highlight that several common trait health anxiety questionnaires were not included in the present study. For example, based on a decision taken around 2010 [44], in the research program from which these data were derived, we administered a dichotomous (“yes”/”no”) item version of the WI-14 as opposed to a Likert-version which is more common nowadays [21].

Conclusion

This study indicates that the WI-14, IAS, HAI-64, HAI-18, and HAI-14 are all valid measures of the same trait health anxiety construct, the core of which appears to be the worry about health, the fear of having or developing a serious disease, and to some extent bodily preoccupation. Approximate linking guidelines enable clinicians and researchers to convert sum scores between questionnaires, and to determine how trait health anxiety levels compare over published studies. For the clinician, a take home message from this study could also be that it is probably more fruitful to identify patients with pathological health anxiety by asking them about health worries and the fear of serious disease (which appears to lies at the core of health anxiety) than to ask about specific behavioral patterns such as healthcare consumption or reassurance seeking which differ considerably between patients, and commonly change for the same patient over time.

Data Availability

The data analyzed during the current study are not publicly available due to the relevant Swedish and European Union data protection and privacy legislation, but are available from the corresponding author on reasonable request. Such requests will be considered on a case-by-case basis, so as to ensure that data and materials are stored, managed, and shared in accordance with the local policies of the sponsor and the relevant Swedish and European Union data protection and privacy legislation.

References

  1. Salkovskis PM, Rimes K, Warwick H, Clark D. The Health anxiety inventory: development and validation of scales for the measurement of health anxiety and hypochondriasis. Psychol Med. 2002;32(5):843–53.

    CAS  PubMed  Google Scholar 

  2. Deacon B, Abramowitz JS. Is hypochondriasis related to obsessive-compulsive disorder, panic disorder, or both? An empirical evaluation. J Cogn Psychother. 2008;22(2):115–27.

    Google Scholar 

  3. Höfling V, Weck F. Assessing bodily Preoccupations is sufficient: clinically effective screening for hypochondriasis. J Psychosom Res. 2013;75(6):526–31.

    PubMed  Google Scholar 

  4. Abramowitz JS, Moore EL. An experimental analysis of hypochondriasis. Behav Res Ther. 2007;45(3):413–24.

    PubMed  Google Scholar 

  5. Aan de Stegge BM, Tak LM, Rosmalen JGM, Oude Voshaar RC. Death anxiety and its association with hypochondriasis and medically unexplained symptoms: a systematic review. J Psychosom Res. 2018;115:58–65.

    PubMed  Google Scholar 

  6. Asmundson GJR, Taylor S, Carleton RN, Weeks JW, Hadjstavropoulos HD. Should health anxiety be carved at the joint? A look at the health anxiety construct using factor mixture modeling in a non-clinical sample. J Anxiety Disord. 2012;26(1):246–51.

    PubMed  Google Scholar 

  7. Longley SL, Broman-Fulks JJ, Calamari JE, Noyes R, Wade M, Orlando CM. A taxometric study of hypochondriasis symptoms. Behav Ther. 2010;41(4):505–14.

    PubMed  Google Scholar 

  8. Ferguson E. A taxometric analysis of health anxiety. Psychol Med. 2009;39(2):277–85.

    CAS  PubMed  Google Scholar 

  9. Norbye AD, Abelsen B, Forde OH, Ringberg U. Distribution of health anxiety in a general adult population and associations with demographic and social network characteristics. Psychol Med. 2022;52(12):2255–62.

    PubMed  Google Scholar 

  10. Axelsson E, Hedman-Lagerlöf E. Validity and clinical utility of distinguishing between DSM-5 somatic symptom disorder and illness anxiety disorder in pathological health anxiety: should we close the chapter? J Psychosom Res. 2023;165:111133.

    PubMed  Google Scholar 

  11. Leonidou C, Panayiotou G. How do illness-anxious individuals process health-threatening information? A systematic review of evidence for the cognitive-behavioral model. J Psychosom Res. 2018;111:100–15.

    PubMed  Google Scholar 

  12. Warwick HM, Salkovskis PM. Hypochondriasis. Behav Res Ther. 1990;28(2):105–17.

    CAS  PubMed  Google Scholar 

  13. Olatunji BO, Deacon BJ, Abramowitz JS. Is hypochondriasis an anxiety disorder? Br J Psychiatry. 2009;194(6):481–2.

    PubMed  Google Scholar 

  14. Stein DJ, Kogan CS, Atmaca M, Fineberg NA, Fontenelle LF, Grant JE, Matsunaga H, Reddy YCJ, Simpson HB, Thomsen PH, et al. The classification of obsessive-compulsive and related Disorders in the ICD-11. J Affect Disord. 2016;190:663–74.

    CAS  PubMed  Google Scholar 

  15. Axelsson E, Hedman-Lagerlöf E. Cognitive behavior therapy for health anxiety: systematic review and meta-analysis of clinical efficacy and health economic outcomes. Expert Rev Pharmacoecon Outcomes Res. 2019;19(6):663–76.

    PubMed  Google Scholar 

  16. Starcevic V, Berle D. Cyberchondria: towards a better understanding of excessive health-related internet use. Expert Rev Neurother. 2013;13(2):205–13.

    CAS  PubMed  Google Scholar 

  17. Dimsdale JE, Creed F, Escobar J, Sharpe M, Wulsin L, Barsky A, Lee S, Irwin MR, Levenson J. Somatic symptom disorder: an important change in DSM. J Psychosom Res. 2013;75(3):223–8.

    PubMed  Google Scholar 

  18. Lebel S, Mutsaers B, Tomei C, Leclair CS, Jones G, Petricone-Westwood D, Rutkowski N, Ta V, Trudel G, Laflamme SZ, et al. Health anxiety and illness-related fears across diverse chronic illnesses: a systematic review on conceptualization, measurement, prevalence, course, and correlates. PLoS ONE. 2020;15(7):e0234124.

    CAS  PubMed  PubMed Central  Google Scholar 

  19. Maheu C, Singh M, Tock WL, Eyrenci A, Galica J, Hebert M, Frati F, Estape T. Fear of Cancer recurrence, health anxiety, worry, and uncertainty: a scoping review about their conceptualization and measurement within breast Cancer Survivorship Research. Front Psychol. 2021;12:644932.

    PubMed  PubMed Central  Google Scholar 

  20. Pilowsky I. Dimensions of hypochondriasis. Br J Psychiatry. 1967;113(494):89–93.

    CAS  PubMed  Google Scholar 

  21. Welch PG, Carleton RN, Asmundson GJ. Measuring health anxiety: moving past the dichotomous response option of the original Whiteley Index. J Anxiety Disord. 2009;23(7):1002–7.

    PubMed  Google Scholar 

  22. Kosic A, Lindholm P, Järvholm K, Hedman-Lagerlöf E, Axelsson E. Three decades of increase in health anxiety: systematic review and meta-analysis of birth cohort changes in university student samples from 1985 to 2017. J Anxiety Disord. 2020;71:102208.

    PubMed  Google Scholar 

  23. Kellner R, Appendix A. Illness attitude scales. Somatization and hypochondriasis. edn. New York: Praeger Publishers; 1986: 319–24.

    Google Scholar 

  24. Sirri L, Grandi S, Fava GA. The illness attitude scales: a clinimetric index for assessing hypochondriacal fears and beliefs. Psychother Psychosom. 2008;77(6):337–50.

    PubMed  Google Scholar 

  25. Alberts NM, Hadjistavropoulos HD, Jones SL, Sharpe D. The short health anxiety inventory: a systematic review and meta-analysis. J Anxiety Disord. 2013;27(1):68–78.

    PubMed  Google Scholar 

  26. Österman S, Axelsson E, Lindefors N, Hedman-Lagerlöf E, Hedman-Lagerlöf M, Kern D, Svanborg C, Ivanov VZ. The 14-item short health anxiety inventory (SHAI-14) used as a screening tool: appropriate interpretation and diagnostic accuracy of the swedish version. BMC Psychiatry 2022, 22(1).

  27. Hedman E, Axelsson E, Andersson E, Lekander M, Ljótsson B. Exposure-based cognitive-behavioural therapy via the internet and as bibliotherapy for somatic symptom disorder and illness anxiety disorder: randomised controlled trial. Br J Psychiatry. 2016;209(5):407–13.

    PubMed  Google Scholar 

  28. Axelsson E, Andersson E, Ljótsson B, Björkander D, Hedman-Lagerlof M, Hedman-Lagerlof E. Effect of Internet vs Face-to-face cognitive behavior therapy for health anxiety: a Randomized Noninferiority Clinical Trial. JAMA Psychiatry. 2020;77(9):915–24.

    PubMed  Google Scholar 

  29. Hedman E, Ljótsson B, Andersson E, Andersson G, Lindefors N, Rück C, Axelsson E, Lekander M. Psychometric properties of internet-administered measures of health anxiety: an investigation of the Health anxiety inventory, the illness attitude scales, and the Whiteley Index. J Anxiety Disord. 2015;31:32–7.

    PubMed  Google Scholar 

  30. Axelsson E, Andersson E, Ljótsson B, Wallhed Finn D, Hedman E. The health preoccupation diagnostic interview: inter-rater reliability of a structured interview for diagnostic assessment of DSM-5 somatic symptom disorder and illness anxiety disorder. Cogn Behav Ther. 2016;45(4):259–69.

    PubMed  Google Scholar 

  31. Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, Hergueta T, Baker R, Dunbar GC. The mini-international neuropsychiatric interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry. 1998;59(Suppl 20):22–33.

    PubMed  Google Scholar 

  32. MacCallum RC, Widaman KF, Zhang S, Hong S. Sample size in factor analysis. Psychol Methods. 1999;4(1):84–99.

    Google Scholar 

  33. de Winter JC, Dodou D, Wieringa PA. Exploratory factor analysis with small sample sizes. Multivar Behav Res. 2009;44(2):147–81.

    Google Scholar 

  34. Kolen MJ, Brennan RL. Test equating, scaling, and linking. Springer; 2014.

  35. Hedman E, Lekander M, Ljótsson B, Lindefors N, Rück C, Andersson G, Andersson E. Optimal cut-off points on the health anxiety inventory, illness attitude scales and whiteley index to identify severe health anxiety. PLoS ONE. 2015;10(4):e0123412.

    PubMed  PubMed Central  Google Scholar 

  36. Furer P, Walker JR, Stein MB. Treating health anxiety and fear of death: a practitioner’s guide. New York City: Springer; 2007.

    Google Scholar 

  37. Lucock MP, Morley S, White C, Peake MD. Responses of consecutive patients to reassurance after gastroscopy: results of self administered questionnaire survey. BMJ. 1997;315(7108):572–5.

    CAS  PubMed  PubMed Central  Google Scholar 

  38. Salkovskis PM, Warwick HM. Morbid preoccupations, health anxiety and reassurance: a cognitive-behavioural approach to hypochondriasis. Behav Res Ther. 1986;24(5):597–602.

    CAS  PubMed  Google Scholar 

  39. Newby JM, Hobbs MJ, Mahoney AEJ, Wong SK, Andrews G. DSM-5 illness anxiety disorder and somatic symptom disorder: Comorbidity, correlates, and overlap with DSM-IV hypochondriasis. J Psychosom Res. 2017;101:31–7.

    PubMed  Google Scholar 

  40. Fergus TA, Valentiner DP. The short health anxiety inventory and multidimensional inventory of hypochondriacal traits: a comparison of two self-report measures of health anxiety. Cognit Ther Res. 2011;35(6):566–74.

    Google Scholar 

  41. Longley SL, Watson D, Noyes R Jr. Assessment of the hypochondriasis domain: the multidimensional inventory of hypochondriacal traits (MIHT). Psychol Assess. 2005;17(1):3–14.

    PubMed  Google Scholar 

  42. MacSwain KLH, Sherry SB, Stewart SH, Watt MC, Hadjistavropoulos HD, Graham AR. Gender differences in health anxiety: an investigation of the interpersonal model of health anxiety. Pers Individ Dif. 2009;47(8):938–43.

    Google Scholar 

  43. Knowles ES. Item context effects on personality scales: measuring changes the measure. J Pers Soc Psychol. 1988;55(2):312–20.

    Google Scholar 

  44. Hedman E, Andersson G, Andersson E, Ljótsson B, Rück C, Asmundson GJ, Lindefors N. Internet-based cognitive-behavioural therapy for severe health anxiety: randomised controlled trial. Br J Psychiatry. 2011;198(3):230–6.

    PubMed  Google Scholar 

Download references

Acknowledgements

This work used the BASS platform from the eHealth Core Facility at Karolinska Institutet.

Funding

Open access funding provided by Karolinska Institute. This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Open access funding provided by Karolinska Institute.

Author information

Authors and Affiliations

  1. Division of Family Medicine and Primary Care, Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, Stockholm, Sweden

    Erland Axelsson

  2. Liljeholmen Primary Health Care Center, Region Stockholm, Liljeholmstorget 7, Stockholm, SE-117 94, Sweden

    Erland Axelsson

  3. Academic Primary Health Care Center, Region Stockholm, Stockholm, Sweden

    Erland Axelsson

  4. Centre for Psychiatry Research, Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden

    Susanna Österman

  5. Division of Psychology, Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden

    Erik Hedman-Lagerlöf

  6. Gustavsberg Primary Health Care Center, Region Stockholm, Stockholm, Sweden

    Erik Hedman-Lagerlöf

Authors
  1. Erland Axelsson
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Susanna Österman
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Erik Hedman-Lagerlöf
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

EA had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Conception and design: EA. Acquisition of data: EA and EHL. Statistical analysis: EA. Drafting of the manuscript: EA. Interpretation, critical revision for intellectual content, approved the final manuscript: EA, SÖ, and EHL.

Corresponding author

Correspondence to Erland Axelsson.

Ethics declarations

Competing interests

EA and EHL have authored a self-help book and several book chapters on the topic of health anxiety, available in the public marketplace. EHL is also a shareholder of DahliaQomit AB, a company specializing in online psychiatric symptom assessment. SÖ declares no conflict of interest.

Ethics approval and consent to participate

This research was approved by the regional ethics review board of Stockholm (2013/375 − 31/5, 2014/1530-31/2) and conducted in accordance with the relevant guidelines and regulations. All study participants gave informed consent to participate in research.

Consent for publication

Not applicable.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Axelsson, E., Österman, S. & Hedman-Lagerlöf, E. Joint factor analysis and approximate equipercentile linking of common trait health anxiety measures: a cross-sectional study of the 14-, 18- and 64-item health anxiety inventory, the illness attitude scale, and the 14-item Whiteley Index. BMC Psychiatry 23, 658 (2023). https://doi.org/10.1186/s12888-023-05151-7

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12888-023-05151-7

Keywords

BMC Psychiatry

ISSN: 1471-244X

Contact us