Somatic symptom burden and health anxiety in the population and their correlates

Highlights

• Concomitant high health anxiety and somatic distress had the poorest outcomes.

• High somatic distress was most closely associated with impairment and service use.

• High health anxiety was most closely associated with dissatisfaction with doctors.

• Findings contribute to the evidence base relevant to DSM-5 Somatic Symptom disorder.

Abstract

Objective

Somatic symptom burden and health anxiety demonstrate overlapping clinical characteristics but their relationship in the general population is unclear. This study examined the association between these dimensions after adjustment for confounders and their respective correlation with outcome measures.

Methods

A randomly selected population-based sample of 3014 respondents aged 15–65 was interviewed by telephone using a structured questionnaire that included the 15-item Patient Health Questionnaire (PHQ-15), Whiteley-7, Kessler-6, Sheehan Disability Scale, socio-demographic variables and items regarding health care utilization. Respondents who scored 10 or above on PHQ-15 and 4 or above on Whiteley-5 were regarded as having high somatic symptom burden and high health anxiety respectively.

Results

Somatic symptom burden and health anxiety are moderately correlated after adjustment for confounders (p < .001). Both have important effects on psychological distress, functional impairment and health care utilization independent of each other (ps < .001). A considerable number of respondents (5.7%) reported both high somatic symptom burden and high health anxiety and this group showed the greatest general psychological distress, functional impairment and health care utilization.

Conclusion

This study demonstrates the close association of somatic symptom burden and health anxiety but also their independent association with psychological distress, functional impairment and health care utilization. The findings support the concept of the DSM-5 category of somatic symptom disorder, but also demonstrate that individuals with high somatic symptom burden or high health anxiety alone may merit separate diagnoses. More sophisticated studies of the relationship between somatic symptom burden and health anxiety are needed.

Introduction

Somatic symptom burden and health anxiety are common in both general population and primary care settings [1], [2], [3], [4]. Clinical studies have indicated that somatic symptom burden is associated with health anxiety. Somatisation disorder is much more common in patients with hypochondriasis than those without [5], [6] and in primary care 20% of somatisation disorder patients were found to also have hypochondriasis [7]. Other primary care studies have shown somatic symptom burden to be positively associated with health anxiety [8], including in Hong Kong [9].

Apart from diagnostic overlap, studies have also demonstrated overlaps in the correlates for somatic symptom burden and health anxiety. People who suffered from somatic symptom burden and/or health anxiety exhibited somatic concern, increased disability, elevated healthcare utilization and dissatisfaction with doctors' explanation of their symptoms and management of their illnesses [4], [9], [10]. A recent clinical study confirmed that patients with both high somatic symptom burden and high health anxiety showed greatest dissatisfaction with doctors' explanation of their symptoms and doctor–patient communication, and the poorest functional outcomes [11].

Nonetheless, the overlap between somatic symptom burden and health anxiety is not complete. In primary care one-tenth of those with somatic symptom burden did not have high health anxiety, and a similar proportion of the respondents with high health anxiety did not show high somatic symptom burden [8]. There are differences between the two conditions in terms of correlates; only somatic symptom burden is associated with female sex [4]. Cognitive behavior therapy may alleviate both conditions [12], [13] but some studies indicated that the treatment response of health anxiety was much better than that of somatic symptom burden [14], [15]. Predictive validity may differ as one study showed that health anxiety predicted mental (but not physical) functioning more clearly than somatic symptom burden [16].

Partly because of the above reasons somatic symptom burden and health anxiety have tended to be studied separately as somatoform pain disorder/undifferentiated somatoform/somatisation disorders and hypochondriasis [17], [18] despite the fact that somatic symptom burden and health anxiety often cluster in patients [19]. The DSM-5 diagnosis of somatic symptom disorder (SSD) combines somatic symptom burden and certain features of health anxiety as well as allows a separate diagnosis of illness anxiety disorder [19], [20]. The diagnosis of SSD has been commended for abandoning the distinction between medically explained and unexplained symptoms and being based on positive psychological criteria, including both somatic symptom burden and health anxiety. However, it has posed diagnostic controversies, such as whether it is an over-inclusive condition [14], [21], [22]. Clinical studies of somatic symptom burden and hypochondriasis or high health anxiety were likely to be biased by help-seeking behavior and the selection of patients with more severe conditions. Their findings need to be replicated in population-based studies in which a broad spectrum of individuals with the two conditions is examined. A recent validation study of the Chinese Whiteley-7 in the Hong Kong general population examined both somatic symptom burden and health anxiety, but provided only preliminary evidence for a positive association between the two conditions [9]. The study did not address several issues including examination of somatic symptom burden and health anxiety as dimensionally distributed variables [9], [10] or the issue of the clinical relevance of these dimensions in people with high scores. The study also did not address the correlates of somatic symptom burden and health anxiety including functional impairment, health care utilization and patients' satisfaction health care utilization [9]. We included satisfaction with doctors in this study because it is closely associated with somatic symptom burden, health anxiety, and health care utilization in the literature, but their relationship has not been examined in population-based studies.

Therefore, the present study aimed to examine the relationship of somatic symptom burden and health anxiety and their correlates in a population-based sample. We examined: 1) the association of somatic symptom burden and health anxiety; 2) the independent effects of the high levels of each on several clinically relevant psychosocial correlates; and 3) the prevalence and correlates of high somatic symptom burden alone, high health anxiety alone and their comorbidity. Using cut-off scores on two commonly used dimensional scales (namely, the PHQ-15 and Whiteley-7), we captured four groups of respondents who exhibited different degrees of somatic symptom burden and health anxiety, namely, “high scores on both”, “high somatic symptom burden alone”, “high health anxiety alone” and “low scores on both” groups respectively.