In 202 patients (61.4% female) with vertigo, dizziness or unsteadiness we investigated gender differences regarding self-perceived disability, anxiety, depression, and its associations. Both genders did not differ significantly in the mean level of self-perceived disability, anxiety, depression, and symptom severity. In men with vertigo, dizziness or unsteadiness the prevalence rates of abnormal anxiety and especially abnormal depression were higher than in women. With respect to the corrected significance value the association between gender and depression was not significant anymore. Our results showed that the associations of abnormal depression and anxiety with self-rated severe disability were significantly stronger in male than in female patients. In comparison to reference values of a general population of Germany  especially in men the prevalence rates of abnormal anxiety and depression were higher (6-fold respectively 2-fold).
As mentioned in the introduction in study-populations of subjects with dizziness, vertigo or unsteadiness the prevalence rates of anxiety and depression vary [2, 4, 6, 12–20]. Diverse patient characteristics for example age, diagnoses, extent of disability, cultural and social aspects and different assessments of anxiety and depression might explain this. The values estimated by us are with 18% at the lower end of previously reported prevalence rates for anxiety and with 20.5% at the upper end of reported prevalence rates of depression. To our knowledge, none of the prior studies explicitly analysed gender differences in the prevalence of anxiety and depression in subjects with vertigo, dizziness or unsteadiness. However, several surveys that estimated the prevalence rates of anxiety and depression in German speaking general populations [21, 40, 43–49] reported higher prevalence rates in women compared to men [21, 44, 46–49]. This raises the questions why we found a different result. One explanation for our conflicting result might be that the higher prevalence rates in women before the onset of dizziness and unsteadiness were picked up by men with the onset of disabling dizziness. Possibly, vertigo or dizziness and associated problems may be greater risk factors for the development of anxiety and depression in men than in women. Because anxiety is described as one cause of depression respectively the development of co-morbid anxiety and depression [45, 50] this may explain why in our sample the prevalence rates of depression (28.9% in men; 15.3% in women) and anxiety (23.7% in men; 14.5% in women) have been of almost the same size when compared within the same gender. About 70% of the patients included in our study suffered ≥ 6 month from vertigo, dizziness or unsteadiness. This might have been the time in which co-morbid anxiety and depression developed.
Different estimations of the level of emotional distress in study-populations composed of individuals with vertigo, dizziness or unsteadiness have been reported. Again the differences may be explained by the diverse size and composition of study-populations, the chosen assessment tools and statistical analyses. We analysed the surveys which applied the HADS [4, 12, 15, 17, 18]. Reported median respectively mean values of the HADS-A (HADS-D) range from 5 (4) in study-participants with peripheral vestibular disorders in Sweden  to 12.6 (6.7) in participants with mixed vestibular disorders in Italy . We estimated in both genders median values of 4 in HADS-D which equals the reference values of Hinz and Schwarz . The estimated median values of 6 points in HADS-A lie two points above the reference values. This may support the hypothesis that the occurrence of vertigo, dizziness or unsteadiness is primarily associated with anxious feelings.
In our survey in the whole study population, age, the different groups of diagnoses, illness duration, employment status, and living condition were not related to anxiety, depression and self-perceived disability. The strongest associations were found between self-perceived disability and anxiety and depression. This association was described in numerous previous studies as mentioned in the introduction [1–8]. We estimated in both genders not only the Spearman correlation coefficients between the DHI-G and HADS but also the associations of the dichotomous variables assessing anxiety respectively depression and self-rated disability. Of the men who rated their disability caused by vertigo, dizziness or unsteadiness as severe 60% had abnormal anxiety and 66.7% suffered from abnormal depression. In women about 26% of the ones with severe disability had anxiety and 19% depression. Especially in male patients the simple self-rating of the severity of disability as mild, moderate or severe might indicate that a patient should be screened for anxiety and depression if he rates the perceived disability as severe.
Our study has some limitations. The HADS is a useful screening instrument which covers relevant emotional distress in subjects with vertigo, dizziness or unsteadiness. But as a screening instrument the HADS gives only limited information on mental and behavioural disorders. The anxiety subscale of the HADS is primarily composed of statements relevant to generalised anxiety [51, 52]. One item assesses the prevalence of panic-attacks. This item will not be sufficient to assess the prevalence of panic-disorders which represented the largest group of comorbid anxiety followed by generalized anxiety disorders and social phobia according to Wiltink et al. . The items of the depression subscale assess simply symptoms of anhedonia (loss of pleasure response). These symptoms belong according to DSM-II-R and ICD-10 to the leading symptoms of the depressive episode, which constitutes only one of several mood disorders [51, 52]. The fact that the HADS only assesses non-somatic symptoms of anxiety and depression may have led to an underestimation of anxiety and depression and possible gender differences, which are described in somatic symptom, may have been undetected. A further weakness of the HADS lies in the different recommended cut-off scores. Beside the suggested cut-offs of the original as well as the German version of the HADS [22, 39] Bjelland and colleagues reported in their review that in most studies caseness was defined by a score of ≥ 8 . A replication of our analyses based on these cut-offs resulted in an increase and equalization of the prevalence rates of anxiety in women and men (34.7%, 32.9%) and a major difference of the prevalence rates of depression in women (17.7%) and men (32.9%). The associations of the categorical variables 'depression' and 'self-perceived severe disability' remained significant only in men. Despite these disadvantages the HADS helps to identify patients with anxiety and depression that should become more specifically assessed and treated by specialists.
In our study-population, female and male patients neither differed significantly in the mean level of self-perceived disability as assessed with the DHI-G, nor in the proportion of both genders in the categories of mild, moderate or severe self-rated disability. Because we found stronger associations between anxiety respectively depression and disability in men compared to women this provokes the question which factors contributed to the level of disability in female patients. In each case we investigated only a small number of possible relevant factors associated with anxiety, depression and disability in patients with vertigo, dizziness or unsteadiness. One of these factors might be the type of vestibular pathology. Some surveys comparing patient groups with different vestibular disorders showed that patients with vestibular migraine and Menière's disease experienced more anxiety, depression and disability than patients with vestibular neuritis or benign paroxysmal positional vertigo [5, 54, 55]. In our survey the subgroup of patients suffering from vestibular migraine (n = 27) did not significantly differ from the other groups in their mean level of anxiety, depression and disability. Further surveys in patients with dizziness and unsteadiness investigated the effect of general co-factors that are known to be associated with emotional distress. Such factors are co-morbidities (e.g. migraine, coronary heart disease, chronic pulmonary disease, cerebrovascular disease, menopause) [1, 2, 10, 11, 23], health behaviour (e.g. smoking, drinking) [1, 2, 10], social status (e.g. income, partnership, employment, family, friends) [1, 2, 10, 15, 56], education [1, 2, 10, 57], personality [10, 18, 58], and the kind of perceptions [3, 7, 8, 10, 13, 15, 19, 56, 57, 59, 60], cognitions [7, 8, 13, 15, 57–60] or coping strategies/illness behaviour [3, 4, 7, 8, 11, 13, 19, 56, 58, 59]. Because of our retrospective analyses we did neither assess specifically these general nor gender specific risk factors of mental disorders. Risk factors for depression that disproportionately affect women include: low socio-economic status, low level of education, housewife, married, mother, single mother, unemployed, poor social support, responsible for family members in need of care, low mastery, violation in childhood, demanding life events, lower self-esteem, self-incrimination, rumination, pregnancy, birth [61–63]. Risk factors which might primarily affect men are: living alone, being divorced, problems at work, reduced bonus payment, retirement, chronic disease, limited perception of need, difficulties in help seeking [61–63].
Our study results can only be generalized with caution. Analyses are based on data primarily collected for a reliability and validity study of the DHI-G. For this study a hypotheses based power calculation was not performed. We recruited patients consecutively with the result of a higher proportion of women participating. This might have affected the investigation of gender differences. Only 26% of the new patients entering our centre answered the letter of enquiry. We do not know whether the patients willing to participate were comparable to the patients who did not answer.
We did not perform a multivariate analysis first of all because of our cross-sectional design. As such the results of our survey are purely descriptive and do not give insight into causal inferences.
This leads to suggestions for future research. Possible risk factors for the development of psychiatric distress and severe disabling dizziness should systematically be investigated in longitudinal studies. Some studies have started to investigate these relationships [5, 8, 11, 13, 56–60]. Because perceptions, emotions, cognitions and behaviour may change during the course of disease  the assessment of these aspects should be done in several time intervals to find out which symptoms and signs and what time points are most relevant. Because in our study-population the median values of the DHI-G and HADS did not differ in both gender, however, gender differences could be shown between women and men with extreme health problems, this should be analysed in more detail in future longitudinal studies. Age dependent gender differences in the prevalence of specific mood disorders, gender differences in the risk factors, symptoms and signs of mood disorders, gender bias in the diagnostic procedures are challenges which also have to be considered in future research in this field.