Background
Dizziness is one of the most frequent symptoms encountered in medical practice [
1,
2], with a prevalence of approximately 20-30% [
2‐
4]. Dizziness is more common in women [
2,
5]. It is associated with other psychological and physical comorbidities [
6,
7]. Dizziness can refer to vertigo, presyncope, disequilibrium, or to non-specific feelings such as giddiness or foolishness [
8]. Its origin can be vestibular, neurological, cardiovascular or psychological [
9]. More than half of the patients suffering from dizziness have non-vestibular diagnoses [
2]. Furthermore, in about half of the patients of specialized units for oto-neurological disorders, symptoms of dizziness are not fully explained by identifiable medical illnesses but are related to mental disorders such as anxiety [
10‐
12]. Patients who experience dizziness report a variety of symptoms, including nausea, instability, disruptions of normal activity patterns and emotional distress. These symptoms are suggested to handicap the patients significantly [
9,
13]. Dizziness often takes a chronic course [
2,
5,
14].
The health related quality of life (HRQoL) is a multidimensional concept, which reflects core components of functioning (e.g. physical, psychological/emotional and social functioning) in the context of medical conditions [
15]. HRQoL measurements evaluate the impact of medical conditions on subjective well-being [
16]. These measurements are suggested as outcome measures in clinical trials and for assessing the burden of clinical conditions by comparing clinical patients with the general population [
16]. One measure assessing HRQoL is the Medical Outcomes Studies 36-Item Short-Form Health Survey (SF-36) [
17].
Findings on differences between HRQoL of patients suffering from dizziness and the general population are inconsistent. While some studies of patients suffering from dizziness found impaired mental and physical HRQoL as well as impairment of all eight subscales of the SF-36 [
18‐
20], other studies showed no impairment of some of these scales (e.g. physical function, general health and social function) [
21,
22]. Reasons for this inconsistency could be differences in the studied samples. While some studies included patients with acute objective measurable medical problems [
18,
22] others included etiologically heterogeneous patients [
20,
21] and yet others included patients with Meniere’s disease in which dizziness is only one of three major problems (next to hearing loss and tinnitus) [
19].
In addition to these inconsistent results, the associations of clinical symptoms (e.g. duration of illness), psychosocial factors (e.g. emotional distress) and HRQoL have rarely been investigated. As there is evidence that not only clinical symptoms but especially psychosocial factors might be significantly associated to the HRQoL and might play a significant role in treatment outcome [
23‐
25], it seems important to disentangle these associations more precisely. In elderly patients frequency of dizziness, but not the duration of the symptoms correlated with the HRQoL; emotional distress and HRQoL correlated negatively [
20]. However, the contribution of these factors to the variance of the HRQoL has not been investigated. In Meniere’s disease, dizziness (not hearing loss and tinnitus) was most strongly associated with a low HRQoL [
19]. The duration of Meniere’s disease showed no association with HRQoL [
19]. As Meniere’s disease is only one condition of many in which dizziness occurs these results can indicate an association between these factors and the HRQoL but generalization of the findings to the etiologically heterogeneous sample of dizziness patients is difficult.
Altogether the associations between clinical symptom variables (e.g. duration and severity of symptoms), psychosocial factors (e.g. emotional conditions) and HRQoL in patients suffering from dizziness remain unclear. However there is growing evidence that not only clinical symptoms but additional psychosocial factors might influence the perceived HRQoL and the treatment outcome. Therefore, the aim of our cross-sectional study was to identify significant associations between possible predictive clinical symptoms and psychosocial factors and the HRQoL in patients suffering from dizziness. Our sample, which was recruited in the Interdisciplinary Centre for Vertigo and Balance Disorders at the University Hospital of Zurich, was not limited to specific diagnoses. Therefore, the sample was etiologically heterogeneous and included patients with vertigo, non-vertigo and mixed (vertigo and non-vertigo) symptoms of dizziness. We hypothesized that the HRQoL (all subscales) of patients suffering from dizziness is significantly lower than in the general population. We expected that HRQoL in dizziness patients is associated with clinical symptoms (e.g. severity and frequency of symptoms). Furthermore, we expected that psychosocial factors (e.g. emotional distress) influence the variances of the mental as well as physical HRQoL significantly.
Discussion
The current study aimed at differentiating and determining the association between clinical characteristics, psychosocial factors, and the HRQoL in patients with dizziness. In this etiologically heterogeneous group of patients suffering from dizziness, HRQoL was markedly impaired compared to the general population. This impairment was associated with symptom severity (DHI), vertigo characteristics of dizziness, days of sick leave, education (PCS-36) and with emotional distress (HADS), living with someone and education (MCS-36).
The HRQoL of patients suffering from dizziness was markedly impaired in both component scores and all subscales of the SF-36 compared to the general population [
27]. This general population is an image of the population aged 18 and above and includes healthy subjects and people with acute and/or chronic diseases. The mean age of our sample (44.7 years) and the norm population (47.7 years) was comparable and the sex-ratio differed only slightly (our sample: 52.2% women, norm population: 55.6% women). While three other dizziness studies [
18‐
20], found similar results, two studies [
21,
22] found less impairment. One of these two studies [
22] included a selected group of elderly patients who had an acute medical condition (Benign Paroxysmal Positional Vertigo). This acute developed medical condition might have caused lower degrees of impairment than the mainly long lasting dizziness (mean 174 weeks) in our sample. Overall, dizziness was related to the physical and mental HRQoL. This impairment seems to be substantial and others as well as our results suggest that the impact of dizziness on HRQoL may be significantly underestimated [
9,
41].
In the hierarchical regression, 69 percent of the variance of MCS-36 was explained. The most significant associated factor was emotional distress, explaining 64 percent of the variance. Patients suffering from dizziness are often fearful of subsequent dizziness-attacks and the consequences of these attacks [
42]. In an attempt to avoid situations that could possibly provoke dizziness, sufferers may restrict their daily activities, and in this way might increase their suffering (lower mental HRQoL) and emotional distress even more. This circle between dizziness, HRQoL, and emotional distress seems to be supported by the current results, which showed that higher emotional distress and lower MCS-36 were significantly associated and that emotional distress in patients with dizziness contributed about 90 percent to the explained variance of MCS-36. A second important contribution variable of MCS-36 was living with a significant other. Living with a significant other was associated with a better MCS-36 and added two percent to its explained variance. The modulating effect of living with a significant other implies that assistance in every-day life by significant others might play an important role in the self-management of chronic diseases. In our study neither gender nor age was associated with MCS-36. Even though there is evidence that dizziness is more common in women and occurs more often in older patients [
43] there is limited evidence that this is reflected in the HRQoL [
9]. While studies examining heterogeneous groups of patients do not tend to find gender differences regarding generic HRQoL [
21] one other study examining patients with Meniere’s disease reported differences between women and men in MCS-36 [
19]. However another study found no gender differences in generic HRQoL (SF-12) in Meniere’s patients [
44]. Overall, results suggest no significant relationship between gender and MCS-36 in patients suffering from dizziness. The results of the current study showed no association between either the duration nor the frequency of dizziness or the severity of dizziness symptoms and the MCS-36. Given that, our results support the previous observations that clinical symptoms of dizziness seem not to be crucial for the impaired mental HRQoL [
44,
45] but that psychosocial factors such as emotional distress mainly contribute to the variance of the mental HRQoL [
13].
Hierarchical regression of the PCS-36 explained 45 percent of its variance. The most significant associated factor was the severity of dizziness (DHI), explaining 35 percent of the variance (high DHI ~ low PCS-36). Additional four percent of the variance was explained by vertigo characteristics of dizziness, which reduced the PCS-36 further. Vestibular disturbances (e.g. spinning) and provoked autonomic symptoms such as nausea and sweating [
46] potentially reduce physical activities and might influence the physical role of the patient negatively (reduced PCS-36). In addition, due to a higher severity of dizziness symptoms a vicious circle of vestibular symptoms, autonomic symptoms, and increasing physical deconditioning might set in. As a possible consequence a vicious circle of vestibular symptoms, autonomic symptoms, and increasing deconditioning might set in. In this vicious circle, adaptive processes of the central nervous system might be decelerated and dizziness symptoms might be prolonged. In turn these potentially prolonged dizziness symptoms might trigger the autonomic arousal [
47,
48] and possibly maintain a reduced PCS-36. As expected, high scores for sick leave were associated with low scores in PCS-36. Age and gender did not contribute significantly to the variance of PCS-36. Even though dizziness is common in older age a significant relationship between impairment in patients with dizziness and age is not obvious [
49]. While univariate results suggested a significant association between HADS and PCS-36, hierarchical regression showed suppressor effects of HADS. As the univariate correlation was low the suppressor (HADS) rather improved the prediction than contributed to the variance on its own (less than one percent). Furthermore, possible suppressor effects were already indicated in the univariate context which showed a relatively high correlation between DHI and HADS and opposites signs for DHI and HADS and DHI and PCS-36.
Taken the results for PCS-36 and MCS-36 together, patients might assume a physical nature for their symptoms (low PCS-36), which might encourage them to seek medical help. If one assumes that symptoms are of somatic nature, medical visits might be unsatisfactory because no sufficient somatic cause can be found and patients might experience emotional distress by these visits. Thus, the emotional arousal potentially increases, even though the somatic impairment is unchanged or improved due to compensating processes over time [
24]. This however, might result in a low MCS-36 and might explain why both the PCS-36 and the MCS-36 are reduced in patient suffering from dizziness.
Some limitations of this study should be noted. The cross-sectional design of the study prevents conclusions about causality. Longitudinal designs might reveal causal relationships between the various clinical symptoms of dizziness, psychosocial factors and HRQoL. Furthermore, the overall response rate of the study was moderate. However it is known that in non-face-to-face survey studies, it is difficult to obtain a patients response rate of 60% [
50]. The non-face to-face design was used to assess the patients before the clinical examination in the Interdisciplinary Centre for Vertigo and Balance Disorders. Therefore, all our data were pre-consultation data and we do not know whether additional patient characteristics such as comorbid conditions or diagnostic subgroups might be associated with the HRQoL. However, this pre-consultation design was necessary to prevent changes in the answering behaviour due to the interdisciplinary clinical examination and counselling.
Another point is that dizziness seems to be more frequent in women [
2], which is, not reflected in our sample. However, the participating patients did not differ from non-participating patients regarding gender and age. Therefore, our study seems not to suffer from a significant recruiting bias regarding these important socio-demographic variables. However, due to the pre-consultation design we were not able to compare further variables such as symptom duration, or comorbid conditions between included and excluded patients. Another limitation is the recruitment of patients at a tertiary care centre. Patients are not commonly referred to a specialist care centre with early or mild symptoms which might have resulted in patients with more severe dizziness [
2,
14]. This might be reflected by the somewhat higher DHI score (46.0) than in other studies [
9,
51]. Although we think that the current results reflect a vast array of patients suffering from dizziness, they have to be interpreted with caution.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
SW initiated the collaborative project, designed data collection tools, collected and monitored data collection, wrote the statistical analysis plan, cleaned, analysed, and interpreted the data, and drafted and revised the paper. ABB monitored data collection, analysed and interpreted the data, and drafted and revised the paper. DS collected data and monitored data collection, and critically revised the draft paper. SCAH collected data and monitored data collection, and critically revised the draft paper. GK collected data and revised the draft paper. MR initiated the collaborative project, monitored data collection, interpreted the data and drafted and revised the paper. All authors read and approved the final manuscript.