Health anxiety and medical utilization: The moderating effect of age among patients in primary care

Accepted Manuscript Title: Health Anxiety and Medical Utilization: The Moderating Effect of Age among Patients in Primary Care Authors: Thomas A. Fergus, Jackson O. Griggs, Scott C. Cunningham, Lance P. Kelley PII: DOI: Reference:

S0887-6185(17)30034-8 http://dx.doi.org/doi:10.1016/j.janxdis.2017.06.002 ANXDIS 1953

To appear in:

Journal of Anxiety Disorders

Received date: Revised date: Accepted date:

1-2-2017 5-6-2017 11-6-2017

Please cite this article as: Fergus, Thomas A., Griggs, Jackson O., Cunningham, Scott C., & Kelley, Lance P., Health Anxiety and Medical Utilization: The Moderating Effect of Age among Patients in Primary Care.Journal of Anxiety Disorders http://dx.doi.org/10.1016/j.janxdis.2017.06.002 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

HEALTH ANXIETY AND MEDICAL UTILIZATION

Health Anxiety and Medical Utilization: The Moderating Effect of Age among Patients in Primary Care Thomas A. Fergus,a Jackson O. Griggs,b Scott C. Cunningham,c and Lance P. Kelleyb a

Department of Psychology and Neuroscience, Baylor University, Waco, Texas, USA

b

Waco Family Medicine Residency Program, Heart of Texas Community Health Center, Waco,

Texas, USA c

Hankamer School of Business, Baylor University, Waco, Texas, USA

Corresponding author: Thomas A. Fergus, Ph.D. Department of Psychology and Neuroscience Baylor University Waco, TX 76798 Voice: 1-254-710-2651 Fax: 1-254-710-3033 E-mail: [email protected]

1

HEALTH ANXIETY AND MEDICAL UTILIZATION

2

Highlights -

Examined if age moderates health anxiety-medical utilization association

-

Health anxiety-medical utilization association strongest among older adults

-

Effect specific to the somatic/body preoccupation health anxiety dimension

-

Ethnoracial differences in health anxiety either not supported or small in magnitude

Abstract Health anxiety is commonly seen in medical clinics and is related to the overutilization of primary care services, but existing studies have not yet considered the possible moderating effect of age. We examined if age moderated the association between health anxiety and medical utilization. A secondary aim was to examine potential racial/ethnic differences in health anxiety. An ethnoracially diverse group of patients (N = 533) seeking treatment from a primary care clinic completed a self-report measure of health anxiety. Three indices of medical utilization were assessed using medical records, including the number of: (a) clinic visits over the past two years, (b) current medications, and (c) lab tests over the past two years. Age moderated the effect of health anxiety on multiple indices of medical utilization. Supplemental analyses found that the moderating effect of age was specific to a somatic/body preoccupation, rather than health worry, dimension of health anxiety. Mean-level differences in health anxiety were either not supported (health anxiety composite, somatic/body preoccupation) or were small in magnitude (health worry) among self-identifying Black, Latino, and White participants. Results indicate that assessing for health anxiety could be particularly important for older adult patients who frequently seek out medical services.

Keywords: age; ethnicity; health anxiety; medical utilization; race

HEALTH ANXIETY AND MEDICAL UTILIZATION

3

1. Introduction Health anxiety broadly represents the wide range of worry individuals can have about their health (Asmundson & Taylor, 2005), with cognitive-behavioral models conceptualizing health anxiety as originating because of cognitive factors (Rachman, 2012). For example, Salkovskis and Warwick (2001) proposed that dysfunctional beliefs related to overestimating the likelihood and cost of health problems, as well as beliefs related to difficulties coping with and inadequacy of medical resources treating health problems, contribute to health anxiety. To date, extant research supports the role of those dysfunctional beliefs, as well as other cognitive factors (e.g., attentional bias to health threat, dysfunctional symptom attributions), in relation to health anxiety (Fergus, 2014; Neng & Weck, 2015; Norris & Marcus, 2014; Witthöft et al., 2016). Existing data generally converge on viewing health anxiety as a dimensional construct, such that differences in the severity of health anxiety are best conceptualized quantitatively rather than qualitatively (Ferguson, 2009; Longley et al., 2010). Clinically severe presentations of health anxiety were once conceptualized as hypochondriasis (Warwick & Salkovskis, 1990), but hypochondriasis was eliminated as a disorder in the fifth edition of the Diagnostic and Statistical Manual for Mental Disorders (i.e., DSM-5; American Psychiatric Association, APA, 2013a). More precisely, hypochondriasis was split into two DSM-5 diagnoses: somatic symptom disorder and illness anxiety disorder (Rief & Martin, 2014). The symptom presentation of somatic symptom disorder primarily consists of health anxiety and severe somatic symptoms, whereas the symptom presentation of illness anxiety disorder primarily consists of health anxiety and either no or mild somatic symptoms (APA 2013a/b). The majority of individuals diagnosed with hypochondriasis appear to qualify for a diagnosis of somatic symptom disorder rather than illness anxiety disorder (Bailer et al., 2016). Nevertheless, the traditional definition of

HEALTH ANXIETY AND MEDICAL UTILIZATION hypochondriasis conceptually parallels illness anxiety disorder (Rief & Martin, 2014; Scarella, Laferton, Ahern, Fallon, & Barsky, 2016). One reason being that somatic symptom disorder includes somatic symptoms that are medically explained (Rief & Martin, 2014), which contrasts with conceptualizations of hypochondriasis as being a presentation of a misinterpretation of symptoms as a medical condition (Warwick & Salkovskis, 1990). Moreover, illness anxiety disorder captures fundamental aspects of hypochondriasis (i.e., preoccupation with having or acquiring a serious illness; Rief & Martin, 2014). Whereas additional research examining the distinctiveness of somatic symptom disorder and illness anxiety disorder is needed, health anxiety remains prominently featured in the DSM-5. Consistent with the symptom presentations of somatic symptom disorder and illness anxiety disorder, two prominent dimensions of health anxiety are the tendency to worry about health (i.e., health worry) and a preoccupation with somatic/body symptoms (Asmundson, Carleton, Bovell, & Taylor, 2008; Welch, Carleton, & Asmundson, 2009). Health worry relates to the tendency to perseverate about one’s health and somatic/body preoccupation relates to the tendency to focus on the possibility that somatic/body sensations may be indicators of health problems. Although research examining these two specific dimensions of health anxiety is in its relative infancy, preliminary research findings suggest the possibility of distinct correlates. For example, Bardeen and Fergus (2014) found that somatic/body preoccupation, rather than health worry, relates particularly robustly to the tendency to act impulsively in the face of negative emotions. Based upon these findings, Bardeen and Fergus (2014) noted that individuals experiencing heightened health anxiety-related preoccupation with their body may be particularly likely to engage in reassurance seeking behavior to mitigate associated distress.

4

HEALTH ANXIETY AND MEDICAL UTILIZATION

5

Cognitive-behavioral models suggest that medical utilization is a safety behavior used by individuals experiencing health anxiety. More precisely, medical utilization may temporarily reduce health anxiety and, yet, utilization of medical services typically serves to sustain health anxiety because individuals do not get corrective experiences that dysfunctional beliefs contributing to health anxiety are unfounded (Abramowitz, Schwartz, & Whiteside, 2002; Salkovskis & Warwick, 2001). A repetitive cycle can develop in which individuals experience health anxiety, seek out reassurance from health professionals to mitigate health anxiety, health anxiety subsequently returns, and individuals engage in repeated medical utilization (Abramowitz et al., 2002; Salkovskis & Warwick, 2001). Consistent with these conceptual models, studies have found small-to-moderate positive associations between health anxiety and medical utilization (rs ranging from .20-.45; Fergus, Bardeen, Gratz, Fulton, & Tull, 2015; Fergus & Valentiner, 2009; Fergus & Valentiner, 2010; Longley, Watson, & Noyes, 2005). Given noted conceptual and empirical links between health anxiety and increased medical utilization, it may not be surprising that severe health anxiety (defined using a cutoff score of 20 on the Short Health Anxiety Inventory [SHAI, Salkovskis, Rimes, Warwick, & Clark, 2002]) is reported by nearly 20% of patients attending medical clinics (Tyrer et al., 2011). The prevalence and impact of health anxiety among patients in primary care clinics has received particular attention (Looper & Dickinson, 2014). For example, Fink et al. (2010) found a 41-78% increase in primary care service utilization among patients with severe health anxiety (defined using diagnostic criteria for hypochondriasis in the DSM-IV [APA, 2000] or an alternative diagnostic criteria for health anxiety proposed by Fink et al., 2004). Based upon these findings, Fink et al. (2010) opined that health anxiety should be more consistently assessed and treated in primary care clinics. To help explicate when health professionals in primary care clinics may

HEALTH ANXIETY AND MEDICAL UTILIZATION

6

want to strongly consider assessing for health anxiety, it is important to examine factors that impact the relation between health anxiety and medical utilization. Age is a potential moderating variable to consider for better understanding the relation between health anxiety and medical utilization. Existing research suggests that age is positively associated with medical utilization (Boston & Merrick, 2010; Thomsen et al., 2004), although medical morbidity may help explain that association (Barsky, Orav, & Bates, 2005). Moreover, there have been calls for increased attention to the relation between age and health anxiety. For example, El-Gabalawy, Mackenzie, Thibodeau, Asmundson, and Sareen (2013) noted that older adults are at an increased risk for physical health ailments and, thus, may be particularly susceptible to health anxiety. Boston and Merrick (2010) examined health anxiety among a sample of older adults and found that over 7% of the participants reported severe health anxiety (defined using a cutoff score of 15 on the SHAI). Boston and Merrick (2010) further found that health anxiety (assessed using the SHAI) was related to greater medical utilization independent of the effects of physical health functioning among their sample of older adults. Whereas El-Gabalawy et al. (2013) note that mild health anxiety may lead to adaptive outcomes among older adults (e.g., serving as a motivator for seeking appropriate care), these researchers further note that early detection of elevated health anxiety among older adults is especially important in helping to reduce the likelihood that health anxiety progresses to more severe presentations. Conceptually, age may impact the association between health anxiety and medical utilization because health anxiety can negatively influence older adults’ perceptions of physical health, thereby further worsening their health anxiety (El-Gabalawy, Mackenzie, Shooshtari, & Sareen, 2011; El-Gabalawy et al., 2013). Older adults often face difficulties with emotion regulation in situations in which they experience heightened distress (Charles & Luong,

HEALTH ANXIETY AND MEDICAL UTILIZATION

7

2013). Consequently, older adults with elevated health anxiety may be likely to seek out medical services as a form of reassurance seeking in an attempt to mitigate health-related distress. The present study further examined interrelations among age, health anxiety, and medical utilization, while providing the first known examination of the moderating effect of age on the association between health anxiety and medical utilization. We examined three indices of medical utilization, including the number of: (a) clinic visits over the past two years, (b) current medications, and (c) lab tests over the past two years. We predicted that health anxiety would positively relate to each index of medical utilization. We further predicted that health anxiety would relate to greater medical utilization as age increased. Supplemental analyses examined which of two health anxiety dimensions – health worry and somatic/body preoccupation (Asmundson et al., 2008) – may be most impacted by the moderating effect of age. When examining the potential moderating effect of age on the association between health anxiety and medical utilization, it is important to account for shared variance with covariates. Potentially relevant covariates include other sociodemographic variables, such as gender and race/ethnicity. For example, women report greater health anxiety in respect to certain dimensions of health anxiety, such as health worry, than men (MacSwain et al., 2009). Further, prior research has found women seek out greater primary care services than do men (Bertakis, Azari, Callahan, & Robbins, 2000). There are inconsistent findings regarding differences in the frequency of primary care utilization across racial/ethnic groups, with some research finding no differences in certain primary care settings (Koo, Madden, & Maguen, 2015). Other research indicates patients who are racial/ethnic minorities may be less likely to use primary care services relative to other sources of care relative to patients who are White (Arnett, Thorpe, Gaskin, Bowie, & LaVesit, 2016). Potential racial/ethnic differences in the severity of health anxiety

HEALTH ANXIETY AND MEDICAL UTILIZATION

8

remain unexamined in the literature. Pursuant to potential differences, Hunter and Schmidt (2010) proposed that individuals who self-identify as Black may report heightened catastrophic interpretations of specific somatic symptoms rather than cognitive symptoms of anxiety (e.g., worry). Such findings provide indirect support for possible racial/ethnic differences in the health anxiety dimensions of health worry and bodily/somatic symptom preoccupation. Two additional covariates to consider are medical morbidity and depression symptoms. As noted, Barsky et al. (2005) noted that age accounts for little unique variance in medical utilization scores once medical morbidity has been taken into account, while finding that primary care patients with severe somatic symptoms had heightened medical morbidity relative to patients without severe somatic symptoms. Additionally, depression positively correlates with both medical utilization (Kimerling, Ouimette, Cronkite, & Moos, 1999) and health anxiety (Wheaton, Berman, Franklin, & Abramowitz, 2010). Statistically accounting for the aforementioned covariates allows for an examination of the robustness of the predicted moderating effect of age on the association between health anxiety and medical utilization. 2. Method 2.1 Participants The initial sample consisted of 538 adults presenting consecutively for treatment at a community health center in a moderately sized (≈ 125,000 residents) southern U.S. city. The average age was 45.5 (SD = 17.4, range 18-90) years and the sample was predominantly female (76.4%). Among the sample, 183 (34.0%) self-identified as Black, 177 (32.9%) as non-Hispanic White, 173 (32.2%) as Latino, three (0.5%) as Asian, and two (0.4%) as “other.” About half of the sample was insured through a state/federal insurance program (50.4%), 23.3% were uninsured, 12.8% were insured through a private insurance program, and 13.5 were insured

HEALTH ANXIETY AND MEDICAL UTILIZATION

9

through another insurance program. To examine possible racial/ethnic differences in health anxiety, as well as to use that variable as a covariate in multivariate analyses, we restricted subsequent analyses to the 533 participants self-identifying as Black, Latino, or White. 2.2 Measures 2.2.1 Whiteley Index (WI; Pilowsky, 1967). The original WI was a 14-item measure that assessed health anxiety using a true/false rating of items. Welch et al.’s (2009) recommended using a 5-point rating scale (ranging from 1 to 5) instead of the dichotomous response option. Welch et al. (2009) further recommended using a revised 6-item version of the WI identified by Asmundson et al. (2008) that is more factorially stable than the original 14-item version of the measure. Asmundson et al.’s (2008) 6-item version of the WI correlates strongly (rs of .63 and .80) with other measures of health anxiety (Fergus, 2013) and was used in this study. Asmundson et al.’s (2008) WI-6 assesses for two dimensions of health anxiety: health worry (e.g., “Do you worry a lot about your health?”) and somatic/body preoccupation (e.g., “Do you find that you are bothered by many different symptoms?”). The WI-6 subscales each consist of three items. The WI-6 scales showed adequate internal consistency in the present study (Cronbach’s αs ranging from .72-.84). 2.2.2 Medical Record Review. A medical record review assessed clinic visits over the past two years, current medications, and lab tests over the past two years. We chose two-years as the time period for two of the indices, which was twice as long as the time period used in prior research examining medical utilization in relation to health anxiety (Abramowitz, Deacon, & Valentiner, 2007; Longley, Watson, & Noyes, 2005), to increase variability in those scores. The medical record review also assessed medical morbidity using the Charlson Comorbidity Index (CCI; Charlson, Pompei, Ales, & MacKenzie, 1987), a widely used index of medical morbidity

HEALTH ANXIETY AND MEDICAL UTILIZATION

10

(e.g., Barsky et al., 2005). Using diagnostic health codes, the CCI weights 17 serious diagnoses in a medical record from 1 to 6 and the ratings are then summed into a total index of medical morbidity. The medical record review finally assessed the severity of depression using responses from the PHQ-2 (Kroenke, Spitzer, & Williams, 2003), a 2-item screening measure for depression. The two items ask respondents the degree to which they were bothered by “little interest or pleasure in doing things” and “feeling down, depressed, or hopeless” over the past two weeks. Items are rated using a 4-point scale (ranging from 0 to 3). Kroenke et al. (2003) identified a cutoff score of 3 or greater on the PHQ-2 as indicating possibly clinically severe depressive symptoms. As part of data collection, we had access to whether participants screened positive (3 or greater) on the PHQ-2. The PHQ-2 was administered to patients once per year as part of standard clinic procedures. As such, we had access to whether participants screened positive on the PHQ-2 within the past 12 months. We coded that screening value as ‘0’ (did not screen positive) or ‘1’ (screened positive) as a marker of elevated depression symptoms. 2.3 Procedure The research was approved by an institutional review board serving the local medical community. Participants were consecutively enrolled. Prospective participants were approached by a trained research assistant in waiting rooms, where the study purpose was described. After obtaining written informed consent, participants completed study measures individually in the waiting room. A more private location was made available. All participants stated that English was their preferred language for communication. Participants consented to have a research team member complete a medical record review. Participants were entered into a raffle to have a chance (≈ 10%) of winning a $20 gift card. 2.4 Data Analytic Strategy

HEALTH ANXIETY AND MEDICAL UTILIZATION

11

Between-groups analysis of variance (ANOVA) analyses examined potential racial/ethnic differences in health anxiety and medical utilization. Zero-order correlations were used to examine bivariate associations between health anxiety and the indices of medical utilization in the total sample. Tests of dependent correlations (Meng, Rosenthal, & Rubin, 1992) were used to examine if health worry versus somatic/body preoccupation correlated significantly stronger with the indices of medical utilization. Hierarchical multiple linear regression analyses examined the main study predictions that age moderates the association between health anxiety and medical utilization in the total sample. We followed Aiken and West’s (1991) guidelines for examining interactive effects using continuous variables, including mean-centering the continuous variables (health anxiety, age) and calculating an interactive effect as a product of the mean-centered variables. The interactive effect was entered into Block 2 of the regression model, with the main effects and covariates entered into Block 1 of the model. Three separate regression models were run, with the only difference in the models being the criterion variable (i.e., clinic visits over the past two years, current medications, and lab tests over the past two years). Supplemental regression analyses were then run using the separate health anxiety dimensions (i.e., health worry, somatic/body preoccupation). The increased possibility of Type I error associated with completing the regression analyses was addressed using the False Discovery Rate (FDR; Benjamini & Hochberg, 1995). Results from the FDR indicated that a familywise alpha level of p < .022 (two-tailed) should be used when examining the statistical significance of the interactive effect in the regression models. 3. Results 3.1 Preliminary Analyses

HEALTH ANXIETY AND MEDICAL UTILIZATION

12

Participants self-identifying as Latino (M = 39.1, SD = 16.8) were significantly younger (F(2, 530) = 19.6, p < .001) than participants self-identifying as Black (M = 47.1, SD = 16.7) or White (M = 49.9, SD = 17.0). There were no gender differences among the three groups of participants (χ2(2) = 2.7, p = .260). Descriptive statistics for health anxiety and medical utilization across the racial/ethnic groups are presented in Table 1. ANOVA results indicated a significant group difference in health worry, which held (F(2, 529) = 3.27, p = .039) when controlling for age in an analysis of covariance (ANCOVA). Follow-up Fisher’s least significant difference (LSD) tests from the ANOVA indicated that participants who self-identified as Black had significantly greater health worry than participants who self-identified as White. The difference was small in magnitude (Cohen’s d = 0.27). ANOVA results also indicated a significant group difference in current medications, with participants self-identifying as Latino having fewer medications than either participants self-identifying as Black or White. However, this group difference did not hold when controlling for age in an ANCOVA (F(2, 529) = 2.23, p = .109). ANOVA results further indicated a significant group difference in past two year lab tests, which held when controlling for age in an ANCOVA (F(2, 529) = 3.10, p = .046). Follow-up LSD tests from the ANOVA indicated that participants who self-identified as Black had a significantly greater number of lab tests than either participants who self-identified as either Latino or White. Those differences were small in magnitude (ds = 0.23). Because racial/ethnic differences in health anxiety and medical utilization were absent, small in magnitude, or attributable to age differences, the groups were combined into a total sample (N = 533) for the remaining analyses. Race/ethnicity was retained as a covariate for multivariate analyses, with that variable dummy-coded (code 1: 0 = White, 1 = Black; code 2: 0 = White, 1 = Latino). Descriptive statistics and raw correlations are presented in Table 2. As

HEALTH ANXIETY AND MEDICAL UTILIZATION

13

predicted, health anxiety (WI-6 total scale) correlated with each index of medical utilization. The correlations were small-to-moderate in magnitude. Among the two health anxiety dimensions, somatic/body preoccupation shared significantly larger correlations with past two year clinic visits and current medications than did health worry (z-statistics = 4.37 and 6.40, ps < .001). 3.2 Moderating Effect of Age Regression results using the WI-6 total scale are presented in Table 3. With the exception of past two year lab tests, health anxiety continued to share an association with the indices of medical utilization after accounting for the covariates in Block 1 of the regression analyses. Among the covariates gender and medical morbidity shared a relation with each index of medical utilization in Block 1 of the regression analyses. As shown in Block 2 of those analyses, study predictions were largely supported. More precisely, there was an interaction between health anxiety and age in relation to current medications and lab tests over the past two years. Before moving onto simple effects, we completed supplemental analyses to examine the moderating effect of age in relation to the two health anxiety dimensions (i.e., health worry, somatic/body preoccupation). The interaction between health worry and age was non-significant across all three criterion variables (clinic visits over the past two years: β = .02, p = .563; current medications: β = .06, p = .088; and lab tests over the past two years: β = .05, p = .291). Alternatively, the interaction between somatic/body preoccupation and age was generally supported across all three criterion variables (clinic visits over the past two years: β = .09, p = .032; current medications: β = .09, p = .003; and lab tests over the past two years: β = .14, p = .001). Of note, the Type I error correction rendered the interaction between somatic/body preoccupation and age in relation to clinic visits over the past two years non-significant. 3.3 Simple Effects

HEALTH ANXIETY AND MEDICAL UTILIZATION

14

Because the interaction between health anxiety and age was specific to the somatic/body preoccupation dimension of health anxiety, we completed and plotted simple effects using only that dimension of health anxiety. Simple effects examined the relation between the somatic/body preoccupation dimension of health anxiety and the three indices of medical utilization at + 1 SD from the mean age (following Aiken & West, 1991). That meant, for the simple effects, younger age was defined as approximately 28 years of age and older age was defined as approximately 63 years of age. Because the interaction between somatic/body preoccupation and age in relation to clinic visits over the past two years was significant following conventional significance testing, we examined the simple effects in relation to that criterion variable as well to see if the pattern of associations was similar to the simple effects with the other indices of medical utilization. Simple effects are depicted in Figure 1 (plotted with + 1 SD from the mean somatic/body preoccupation score). Somatic/body preoccupation shared an association with clinic visits at older (β = .24, p < .001), but not younger (β = .05, p = .388), age. Somatic/body preoccupation shared an association with medication at older (β = .34, p < .001) and younger (β = .15, p = .001) age, although the effect was significantly stronger at older relative to younger age (indicated by the significant interaction; Aiken & West, 1991). Somatic/body preoccupation shared an association with lab tests at older (β = .24, p < .001), but not younger (β = -.02, p = .621), age. 4. Discussion Health anxiety is prevalent in medical clinics and is related to the overutilization of primary care services (Looper & Dickinson, 2014; Fink et al., 2010; Tyrer et al., 2011). Consistent with prior research, we found health anxiety to be associated with indices of medical utilization (i.e., past two-year clinic visits, current medications, past two-year lab tests) in our primary care sample. The magnitude of that association was consistent with observed

HEALTH ANXIETY AND MEDICAL UTILIZATION

15

associations between health anxiety and medical utilization in prior studies (Fergus et al., 2015; Fergus & Valentiner, 2009; Fergus & Valentiner, 2010; Longley et al., 2005). Extending extant research linking health anxiety to greater medical utilization, we next examined if the association between health anxiety and medical utilization may be better understood by considering age. Consistent with predictions, the association between health anxiety (as assessed unidimensionally) and medical utilization was generally impacted by age. The exception to this pattern of findings was past two-year clinic visits, where health anxiety showed a main effect but only showed a trending interactive effect in relation to this index of medical utilization. Supplemental analyses indicated that using a health anxiety total score likely attenuated the magnitude of the interactive effect, as the interaction between the health worry dimension of health anxiety and age was not supported across any of the indices of medical utilization. The interaction between the somatic/body preoccupation dimension of health anxiety and age was generally supported across each of the three indices of medical utilization. The exception was that the Type I error adjustment rendered the interaction between somatic/body preoccupation and age non-significant in relation to past two-year clinic visits, although that interactive effect was supported using conventional standards for significance testing. The moderating effect of age was not attributable to shared variance with a number of covariates, including gender, race/ethnicity, medical morbidity, and elevated depression symptoms. The present results indicate that the interaction between health anxiety and age in relation to medical utilization is specific to the somatic/body preoccupation dimension of health anxiety. The pattern of the interactive effect indicated that somatic/body preoccupation only shared an association with indices of medical utilization (past two-year clinic visits and lab tests) at older, relative to younger, age or shared a significantly stronger association with an index of medical

HEALTH ANXIETY AND MEDICAL UTILIZATION

16

utilization (current medications) at older, relative to younger, age. For plotting and examining simple effects, older age was defined as one standard deviation from the mean age score (Aiken & West, 1991). In the present study, that meant older age examined at approximately 63 years of age. Although the exact age demarcating older adulthood has differed across existing studies in the health anxiety literature, 63 years is consistent with cutoffs of years of age used to define older adulthood in prior research examining health anxiety (El-Gabalawy et al., 2013). El-Gabalawy et al. (2013) note that older adults tend not to seek professional mental health services for anxiety. Consequently, older adults with health anxiety may over-utilize physical healthcare services and under-utilize mental health services relative to younger adults. El-Gabalawy et al. (2013) further note that many older adults do not receive adequate screening, referrals, or treatment for health anxiety from medical practitioners, therefore resulting in the continuation and potential worsening of their health anxiety. Overall, the results offer support to El-Gabalawy et al.’s (2013) proposal that it is important for medical professionals to assess for the potential role of health anxiety among older adults who frequently seek out medical services. The present findings suggest a subtle, yet important, elaboration of this statement, as the effect of age in the present study was specific to understanding how the somatic/body preoccupation dimension of health anxiety relates to medical utilization. As discussed, health anxiety is a cardinal symptom of two DSM-5 psychological disorders: somatic symptom disorder and illness anxiety disorder (APA, 2013a/b). Whereas debate about the relative merits of these disorders exists (Rief & Martin, 2014), individuals with somatic symptom disorder evidence a greater preoccupation with somatic symptoms than do individuals with illness anxiety disorder (Bailer et al., 2016). Bailer et al. (2016) further found that somatic symptom disorder was related to a greater number of physicians consulted as a

HEALTH ANXIETY AND MEDICAL UTILIZATION

17

result of somatic concerns. The present results cannot directly speak to those two disorders per se. However, the specificity of the observed interactive effect in relation to preoccupation with bodily/somatic symptoms may indirectly support the possibility that age is particularly important for understanding the frequency in which individuals with the symptom presentation of health anxiety and concerns about somatic symptoms seek out primary care services. Research examining distinct correlates of the two health anxiety dimensions of the WI-6 is limited. One study found that somatic/body preoccupation was more relevant to the emotion regulation dimension of difficulty remaining in control of behavior when experiencing negative emotion than was health worry (Bardeen & Fergus, 2014). Bardeen and Fergus (2014) speculated that distress related to focusing on the possibility that somatic/body sensations may be indicators of health problems could be particularly linked to behavioral dyscontrol, such as excessive reassurance seeking (e.g., medical utilization), when faced with negative emotion. A possible role of emotion regulation in understanding the present results is consistent with the strength and vulnerability integration (SAVI) model developed by Charles (2010). The SAVI model proposed that older adults generally show expertise in applying emotion regulatory strategies and evidence a greater motivation to use such strategies than do younger adults. Importantly, the SAVI model further proposed that chronic stressors mitigate, and may in fact reverse, age-related improvements in emotion regulation (Charles, 2010). Charles (2010) noted that deteriorating health is one potential chronic stressor that can mitigate age-related strengths in using emotion regulation strategies, with El-Gabalawy et al. (2013) raising the possibility that medical morbidity may be a primary factor involved in health anxiety among older adults. Medical morbidity correlated with health anxiety, particularly the somatic/body preoccupation dimension, as well as age and medical utilization in the present

HEALTH ANXIETY AND MEDICAL UTILIZATION

18

study. However, in the present study, the moderating effect of age was not attributable to shared variance with medical morbidity. In addition, El-Gabalawy et al. (2013) noted that many older adults experience medical morbidity and yet only a minority will experience severe health anxiety. Additional factors other than medical morbidity thus likely warrant consideration. Following from the SAVI model (Charles, 2010) and extant findings (Bardeen & Fergus, 2014), emotion regulation may be a particularly important variable for understanding increased medical utilization among older adults with elevated health anxiety. Older adults with physical health conditions may be more likely to experience health anxiety (El-Gabalawy et al., 2013). Prior research suggests anxiety affects older adults’ perceptions of their physical health, as older adults with anxiety disorders tend to appraise their physical health as poorer (El-Gabalawy et al., 2011). Those misappraisals may contribute to an even greater escalation in health anxiety (Abramowitz et al., 2002; Salkovskis & Warwick, 2001). The SAVI model would predict that such a context (i.e., experiencing a chronic stressor and emotional distress) attenuates or even reverses age-related benefits in emotion regulation among older adults (Charles, 2010). Emotion regulation deficits related to difficulty remaining in control of behavior when experiencing negative emotion could be particularly important in understanding the resulting engagement in medical utilization among older adults with elevated health anxiety (Bardeen & Fergus, 2014). A secondary aim of the present study was to provide the first known examination of potential racial/ethnic differences in health anxiety. There were generally no mean differences in health anxiety across participants self-identifying as Black, Latino, or White with the exception of health worry, where participants self-identifying as Black reported greater health worry than participants self-identifying as White. This group difference was unexpected given that individuals who are Black tend to underreport cognitive symptoms of anxiety, such as worry,

HEALTH ANXIETY AND MEDICAL UTILIZATION

19

relative to individuals who are White (Hunter & Schmidt, 2010). A potentially important contextual factor for understanding these results is that the assessed cognitive symptoms of anxiety pertained to health in the present study, whereas prior research has focused on domaingeneral levels of worry or domain-specific levels of worry other than health (Hunter & Schmidt, 2010). Hunter and Schmidt (2010) noted a particular salience of physical health concerns among individuals who are Black. Because physical health represents a particularly salient threat for individuals who are Black, it is possible that individuals who are Black evidence increased cognitive anxiety in relation to specific domains (i.e., health). Although a tenable possibility, it is important to acknowledge that the group difference in health worry was only small in magnitude and the present pattern of findings require replication before firmer conclusions are drawn. The above considerations should be considered with the following study limitations in mind. We considered health anxiety as a precursor to medical utilization in accordance with prior studies and existing conceptual models (Abramowitz et al., 2007). However, the study design precludes conclusions if health anxiety is a cause or consequence of medical utilization. As part of our data collection efforts, we had access only to whether participants screened positive for clinically severe depression using the PHQ-2 within the past 12 months. The PHQ-2 evidences good sensitivity and specificity in relation to diagnoses of major depressive disorder obtained using clinical interviews (Kroenke et al., 2003); however, future research would benefit from completing a fuller assessment of depressive and related symptoms, such as anxiety (e.g., Deacon, Lickel, & Abramowitz, 2008). Whereas we accounted for several theoretically relevant covariates, other relevant covariates were unassessed in the present study. Socioeconomic status is an example of one variable found to be relevant to medical utilization that may be important to statistically control for in future research (e.g., Blackwell, Martinez, Gentleman, Sanmartin, &

HEALTH ANXIETY AND MEDICAL UTILIZATION

20

Berthelot, 2009). It also may be important for future research to assess aspects of patient interface with medical settings, such as differential medical treatment practices and physicianpatient interactions, as there can exist differences in how patients interface with medical settings across older and younger adults (e.g., Robb, Chen, & Haley, 2002). Our sample had relative strengths (e.g., diversity of age, racial/ethnic diversity), although possible idiosyncrasies of the sample (e.g., geographic location, presenting problems) highlight the importance of replicating the findings in other primary care settings. Whereas the magnitude of the interactive effect between health anxiety and age is consistent with the typical effect size among interactions in questionnaire-based studies (i.e., 1-3%; Aiken & West, 1991), the interaction was only small in magnitude. The magnitude of the interaction highlights the importance of future research continuing to examine variables that are important to furthering our understanding of the impact of health anxiety on medical utilization in primary care settings. Limitations notwithstanding, the present results offer additional support for a link between health anxiety and primary care service utilization. The results indicate patient age is important for understanding healthcare service utilization among primary care patients reporting elevated health anxiety. Routinely assessing for health anxiety in primary care settings, particularly among older adults who frequently seek out medical services, can aid in the identification and treatment of patients experiencing severe health anxiety.

HEALTH ANXIETY AND MEDICAL UTILIZATION

21

References Abramowitz, J. S., Deacon, B. J., & Valentiner, D. P. (2007). The Short Health Anxiety Inventory: Psychometric properties and construct validity in a non-clinical sample. Cognitive Therapy and Research, 31, 871-883. Abramowitz, J. S., Schwartz, S. A., & Whiteside, S. P. (2002). A contemporary conceptual model of hypochondriasis. Mayo Clinic Proceedings, 77, 1323-1330. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington DC: Author. American Psychiatric Association. (2013a). Diagnostic and statistical manual of mental disorders (5th ed.). Washington DC: Author. American Psychiatric Association. (2013b). Highlights of changes from DSM-IV-TR to DSM-5. Washington DC: Author. www.psychiatry.org/dsm5 (accessed 5/2013). Aiken, L. S., & West, S. G. (1991). Multiple regression: testing and interpreting interactions. Thousand Oaks, CA: Sage. Arnett, M. J., Thorpe, R. J., Gaskin, D. J., Bowie, J. V., & LaVesit, T. A. (2016). Race, medical mistrust, and segregation in primary care as usual source of care: Findings from the Exploring Health Disparities in Integrated Communities Study. Journal of Urban Health: Bulletin of the New York Academy of Medicine, 93, 456-467. Asmundson, G. J. G., Carleton, R. N., Bovell, C. V., & Taylor, S. (2008). Comparison of unitary and multidimensional models of the Whiteley Index in a nonclinical sample: Implications for understanding and assessing health anxiety. Journal of Cognitive Psychotherapy, 22, 87-96.

HEALTH ANXIETY AND MEDICAL UTILIZATION

22

Asmundson, G. J.G., & Taylor, S. (2005). It’s not all in your head: How worrying about your health could be making you sick – and what you can do about it. New York, NY: Guilford. Bailer, J., Kerstner, T., Witthöft, M., Diener, C., Mier, D., & Rist, F. (2016). Health anxiety and hypochondriasis in the light of the DSM-5. Anxiety, Stress, & Coping, 29, 1-34. Bardeen, J. R., & Fergus, T. A. (2014). An examination of the incremental contribution of emotion regulation difficulties to health anxiety beyond specific emotion regulation strategies. Journal of Anxiety Disorders, 28, 394-401. Barsky, A. J., Orav, J., & Bates, D. W. (2005). Somatization increases medical utilization and costs independent of psychiatric and medical comorbidity. Archives of General Psychiatry, 62, 903-910. Benjamini, Y., & Hochberg, Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of Royal Statistical Society: Series B, 57, 289-300. Bertakis, K. D., Azari, R., Callahan, E. J., & Robbins, J. A. (2000). Gender differences in the utilization of health care services. Journal of Family Practice, 49, 147-152. Blackwell, D., Martinez, M., Gentleman, J., Sanmartin, C., & Berthelot, J. M. (2009). Socioeconomic status and utilization of health care services in Canada and the United States: Findings from a binational health survey. Medical Care, 47, 1136-1146. Boston, A. F., & Merrick, P. L. (2010). Health anxiety among older people: An exploratory study of health anxiety and safety behaviors in a cohort of older adults in New Zealand. International Psychogeriatrics, 22, 549-558.

HEALTH ANXIETY AND MEDICAL UTILIZATION

23

Charles, S. T. (2010). Strength and vulnerability integration: A model of emotional well-being across adulthood. Psychological Bulletin, 136, 1068-1091. Charles, S. T., & Luong, G. (2013). Emotional experience across adulthood: The theoretical model of strength and vulnerability integration. Current Directions in Psychological Science, 22, 443-448. Charlson, M. E., Pompei, P., Ales, K. L., & Mackenzie, C. R. (1987). A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. Journal of Chronic Disease, 40, 373-383. Deacon, B., Lickel, J., & Abramowitz, J. S. (2008). Medical utilization across the anxiety disorders. Journal of Anxiety Disorders, 22, 344-350. El-Gabalawy, R., Mackenzie, C. S., Shooshtari, S., & Sareen, J. (2011). Comorbid physical health conditions and anxiety disorders: A population-based exploration of prevalence and health outcomes among older adults. General Hospital Psychiatry, 33, 556-564. El-Gabalawy, R., Mackenzie, C. S., Thibodeau, M. A., Asmundson, G. J. G., & Sareen, J. (2013). Health anxiety disorders in older adults: Conceptualizing complex conditions in late life. Clinical Psychology Review, 33, 1096-1105. Fergus, T. A. (2013). Repetitive thought and health anxiety: Tests of specificity. Journal of Psychopathology and Behavioral Assessment, 35, 366-374. Fergus, T. A. (2014). Health-related dysfunctional beliefs and health anxiety: Further evidence of cognitive specificity. Journal of Clinical Psychology, 70, 248-259. Fergus, T. A., Bardeen, J. R., Gratz, K. L., Fulton, J. J., & Tull, M. T. (2015). The contribution of health anxiety to retrospectively-recalled emergency department visits within a sample

HEALTH ANXIETY AND MEDICAL UTILIZATION

24

of patients in residential substance abuse treatment. Cognitive Behaviour Therapy, 44, 18. Fergus, T. A., & Valentiner, D. P. (2009). Reexamining the domain of hypochondriasis: Comparing the Illness Attitudes Scale to other approaches. Journal of Anxiety Disorders, 23, 760-766. Fergus, T. A., & Valentiner, D. P. (2010). Disease phobia and disease conviction are separate dimensions underlying hypochondriasis. Journal of Behavior Therapy and Experimental Psychiatry, 41, 438-444. Ferguson, E. (2009). A taxometric analysis of health anxiety. Psychological Medicine, 39, 277285. Fink, P., Ørnbøl, E., & Christensen, K. S. (2010). The outcome of health anxiety in primary care. A two-year follow-up study on health care costs and self-rated health. PLoS ONE, 5, e9873. Fink, P., Ørnbøl, E., Toft, T., & Sparle, K. C., Frostholm, L., & Olesen, F. (2004). A new, empirically established hypochondriasis diagnosis. American Journal of Psychiatry, 161, 1680-1691. Hunter, L. R., & Schmidt, N. B. (2010). Anxiety psychopathology in African American adults: Literature review and development of an empirically informed sociocultural model. Psychological Bulletin, 136, 211-235. Kimerling, R., Ouimette, P. C., Cronkite, R. C., & Moose, R. H. (1999). Depression and outpatient medical utilization: A naturalistic 10-year follow-up. Annals of Behavioral Medicine, 21, 317-321.

HEALTH ANXIETY AND MEDICAL UTILIZATION

25

Koo, K. H., Madden, E., & Maguen, S. (2015). Race-ethnicity and gender differences in VA health care service utilization among U.S. veterans of recent conflicts. Psychiatric Services, 66, 507-513. Kroenke, K., Spitzer, R. L., & Williams, J. B. W. (2003). The Patient Health Questionnaire-2: Validity of a two-item depression screener. Medical Care, 41, 1284-1292. Longley, S. L., Broman-Fulks, J. J., Calamari, J. E., Noyes, R., Wade, M., & Orlando, C. M. (2010). A taxometric study of hypochondriasis symptoms. Behavior Therapy, 41, 505514. Longley, S. L., Watson, D., & Noyes, R., Jr. (2005). Assessment of the hypochondriasis domain: The Multidimensional Inventory of Hypochondriacal Traits (MIHT). Psychological Assessment, 17, 3-14. Looper, K., & Dickinson, P. (2014). Epidemiological and economic aspects of hypochondriasis and health anxiety. In V. Starcevic & R. Noyes Jr. (Eds.), Hypochondriasis and health anxiety: A guide for clinicians (pp. 85-112). New York, NY: Oxford University. MacSwain, K. L. H., Sherry, S. B., Stewart, S. H., Watt, M. C., Hadjistavropoulos, H. D., & Graham, A. R. (2009). Gender differences in health anxiety: An investigation of the interpersonal model of health anxiety. Personality and Individual Differences, 47, 938943. Meng, X. L., Rosenthal, R., & Rubin, D. B. (1992). Comparing correlated correlation coefficients. Psychological Bulletin, 111, 172-175. Neng, J. M. B., & Weck, F. (2015). Attribution of somatic symptoms in hypochondriasis. Clinical Psychology and Psychotherapy, 22, 116-124.

HEALTH ANXIETY AND MEDICAL UTILIZATION

26

Norris, A. L., & Marcus, D. K. (2014). Cognition in health anxiety and hypochondriasis: Recent advances. Current Psychiatry Reviews, 10, 44-49. Pilowsky, I. (1967). Dimensions of hypochondriasis. British Journal of Psychiatry, 113, 89-93. Rachman, S. (2012). Health anxiety disorders: A cognitive construal. Behaviour Research and Therapy, 50, 502-512. Rief, W., & Martin, A. (2014). How to use the new DSM-5 somatic symptom disorder diagnosis in research and practice: A critical evaluation and a proposal for modifications. Annual Review of Clinical Psychology, 10, 339-367. Robb, C., Chen, H., & Haley, W. E. (2002). Ageism in mental health and health care: A critical review. Journal of Clinical Geropsychology, 8, 1-12. Salkovskis, P. M., Rimes, K. A., Warwick, H. M., & Clark, D. M. (2002). The Health Anxiety Inventory: Development and validation of scales for the measurement of health anxiety and hypochondriasis. Psychological Medicine, 32, 843-853. Salkovskis, P. M., & Warwick, H. M. (2001). Making sense of hypochondriasis: A cognitive model of health anxiety. In G. J. G. Asmundson, S. Taylor, & B. J. Cox (Eds.), Health anxiety: Clinical and research perspectives on hypochondriasis and related conditions (pp. 47-64). New York, NY: John Wiley & Sons. Scarella, T. M., Laferton, J. A. C., Ahern, D. K., Fallon, B. A., & Barsky, A. (2016). The relationship of hypochondriasis to anxiety, depressive, and somatoform disorders. Psychosomatics, 57, 200-207. Thomsen, D. K., Mehlsen, M. Y., Hokland, M., Viidik, A., Olesen, F., Avlund, K., Munk, K., & Zachariae, R. (2004). Negative thoughts and health: Associations among rumination,

HEALTH ANXIETY AND MEDICAL UTILIZATION

27

immunity, and health care utilization in a young and elderly sample. Psychosomatic Medicine, 66, 363-371. Tyrer, P., Cooper, S., Crawford, M., Dupont, S., Green, J., Murphy, D., … Tyrer, H. (2011). Prevalence of health anxiety problems in medical clinics. Journal of Psychosomatic Research, 71, 392-394. Warwick, H. M. C., & Salkovskis, P. M. (1990). Hypochondriasis. Behaviour Research and Therapy, 28, 105-117. Welch, P. G., Carleton, R. N., & Asmundson, G. J. G. (2009). Measuring health anxiety: Moving past the dichotomous response option of the original Whiteley Index. Journal of Anxiety Disorders, 23, 1002-1007. Wheaton, M. G., Berman, N. C., Franklin, J. C., & Abramowitz, J. S. (2010). Health anxiety: Latent structure and associations with anxiety-related psychological processes in a student sample. Journal of Psychopathology and Behavioral Assessment, 32, 565-574. Witthöft, M., Kerstner, T., Ofer, J., Mier, D., Rist, F., Diener, C., & Bailer, J. (2016). Cognitive biases in pathological health anxiety. Clinical Psychological Science, 4, 464-479.

HEALTH ANXIETY AND MEDICAL UTILIZATION

28

Past Two Year Clinic Visits

25 20 15 -1 SD Age 10

+1 SD Age

5 0

Current Number of Medications

-1 SD Health Anxiety

+1 SD Health Anxiety

25 20 15 -1 SD Age 10

+1 SD Age

5 0 -1 SD Health Anxiety

+1 SD Health Anxiety

Past Two Year Lab Tests

25 20 15 -1 SD Age 10

+1 SD Age

5 0 -1 SD Health Anxiety

+1 SD Health Anxiety

Figure 1. The moderating effect of age on the association between health anxiety (somatic/body preoccupation dimension) and indices of medical utilization.

HEALTH ANXIETY AND MEDICAL UTILIZATION

28

Table 1 Racial/Ethnic Differences in Health Anxiety and Medical Utilization. Black Variable

Mean

Latino

SD

Mean

White

SD

Mean

SD

F(2, 530)

WI-6-Total

16.87

6.76

15.51

6.62

15.62

6.72

2.29

WI-6-Worry

9.00

3.72

8.46

3.38

8.02

3.68

3.37*

WI-6-Preoccupation

7.86

3.85

7.05

3.98

7.59

3.97

1.98

12.04

7.56

10.94

8.08

12.69

9.69

1.92

Current Medications

8.15

6.52

5.60

5.31

9.11

6.96

Two Year Lab Tests

19.49

12.69

16.53

13.14

16.48

13.27

Two Year Clinic Visits

Group Comparisons

B>W

14.46** B > L, W > L 3.16*

B > L, B > W

Note. ** p < .001, * p < .05 (two-tailed). Sample sizes: B (Black) n = 183, L (Latino) n = 173, W (White) n = 177.

HEALTH ANXIETY AND MEDICAL UTILIZATION

29

Table 2 Descriptive Statistics and Zero-Order Correlations in Total Sample. Variable

Mean

SD

1

2

3

4

5

6

7

8

1. WI-6-Total

16.01

6.72

2. WI-6-Worry

8.50

3.62

.88**

3. WI-6-Preoccupation

7.51

3.94

.90**

.58**

11.90

8.50

.21**

.10*

.27**

7.64

6.47

.33**

.17**

.41**

.42**

6. Past Two Year Lab Tests

17.53

13.08

.12*

.08

.13*

.72**

.31**

7. Age

45.45

17.39

.09*

-.05

.21**

.16**

.61**

.02

1.51

1.99

.20**

.07

.28**

.25**

.56**

.21**

.50**

.13*

.08

.15**

.14*

.06

.09*

.07

4. Past Two Year Clinic Visits 5. Current Medications

8. Charlson Comorbidity Index 9. Patient Health Questionnaire-2a -

-

-

Note. N = 533. ** p < .001, * p < .05 (two-tailed). a = screened negative (0), screened positive (1).

.10*

HEALTH ANXIETY AND MEDICAL UTILIZATION

30

Table 3 Results from Regression Analyses Predicting Indices of Medical Utilization using Health Anxiety Total Score. Index of Medical Utilization in Regression Model Two Year Clinic Visits Variable Block 1

ΔR2 .14

Gendera

β

p < .001

Current Medications ΔR2 .52

β

p < .001

Two Year Lab Tests ΔR2 .11

β

P < .001

.21

< .001

.02

.002

.20

< .001

White-Blackb

-.06

.215

-.07

.058

.09

.078

White-Latinoc

-.05

.259

-.08

.015

.01

.889

Medical Morbidity

.18

< .001

.30

< .001

.24

< .001

Depression Severityd

.10

.011

-.02

.509

.07

.082

Age

.08

.127

.44

< .001

-.08

.102

Health Anxiety

.16

< .001

.23

< .001

.06

.144

Block 2

.00

Health Anxiety x Age

.083 .07

.103

.01

.004 .09

.004

.01

.014 .11

.014

Note. N = 533. Dummy-coded variables: a 0 = men, 1 = women; b 0 = White, 1 = Black; c 0 = White, 1 = Latino; d 0 = screened negative, 1 = screened positive.