[Study of psychometric properties of two new questionnaires assessing beliefs in hypochondriasis in a non-clinical population].

Most instruments focussing on hypochondriasis symptoms do not have for goal to assess beliefs specifically. Instead, these instruments are used to measure specific behaviors. To assess underlying beliefs with these kinds of instruments, you have to extract false beliefs by deduction. In cognitive therapy, it is important to target erroneous beliefs in order to change them. On the other hand, existing instruments are not really suitable to target erroneous health beliefs. Even if some questionnaires are built to assess beliefs directly, it seems that they only measure the conviction of having an illness and do not assess the general health beliefs present in excessive health worriers. However, many researchers argue that this other kind of beliefs are the ones responsible in maintaining hypochondriasis symptoms. Presently, researchers assume that erroneous beliefs can maintain worries about illness among people with hypochondriasis symptoms like false beliefs about worry maintain worries in people with General Anxiety Disorder (GAD). Even if the importance of false beliefs in the maintenance of pathological worries is now recognized, most instruments on hypochondriasis symptoms do not have for goal to assess erroneous beliefs concerning worry about health. For instance, although the questionnaire Why do people worry? (WW) shows good psychometric properties and measures beliefs related to general worries, this questionnaire is not specific enough to correctly evaluate beliefs associated to health worry. A new questionnaire has to emerge in order to assess false beliefs associated to worry about health. This manuscript presents the development and the validation of a new questionnaire: the General Health Beliefs Questionnaire (GHBQ) that assesses general health beliefs, and also presents the development and the validation of a new questionnaire assessing beliefs associated to worry about health: the Why do people Worry about Health? (WW-H) . In this study, the GHBQ's and the WW-H's psychometric qualities and the factorial structure were assessed. More precisely, this study examined the factorial structure, the temporal stability, the convergent, divergent and criteria validities of the GHBQ and the WW-H. Four hundred and twenty nine French-speaking university students (non-clinical participants) completed a battery of questionnaires at the beginning of a class. The questionnaires were: The General Health Beliefs Questionnaire (GHBQ), the Why do people Worry about Health (WW-H), the Illness Worry Scale (IWS), the Beck Depression Inventory-short form (BDI-short form) and the Beck Anxiety Inventory (BAI). A second administration took place three weeks later with the same sample to test the temporal stability of the GHBQ and the WW-H. The principal component analysis with orthogonal rotation (varimax) supports a five components solution for the GHBQ: 1) magical thinking, 2) health, 3) consequences, 4) responsibility, and 5) vulnerability. The principal component analysis with oblique rotation (direct oblimin) (d=0) found a two components solution for the WW-H: 1) utility and 2) magical thinking associated to worries about health. The internal consistency of the GHBQ and the WW-H is excellent (a=.80 and a=.90, respectively). A correlation of 0.49 was found between the GHBQ and the WW-H. The correlation between the GHBQ and the IWS who evaluates the tendency to worry (r=.50) and between the WW-H and the IWS (r=.49) showed that the convergent validity of these questionnaires is adequate. On a three weeks interval, the GHBQ (r=.70) and the WW-H (r=.71) showed a satisfying temporal stability. The means of the high worriers (80 superior percentile at the IWS) (M=38.8, ET=8.93) and the means of the moderate worriers (between the 40 and the 60 percentile at the IWS) (M=32.8, ET=8.00) on the GHBQ have been compared. A significant difference has been found between the two groups [F(1,181)=23,129, p<0,001]. Also, the means of the high worriers (M=27.3, ET=8.59) and the means of the moderate worriers (M=23.8, ET=8.56) on the WW-H haveave been compared. An ANOVA has found a significant difference between these two groups [F(1,180)=7,396, p=0,007]. These results show that general health beliefs and false beliefs associated to worry about health are more often present in high worriers than in moderate worriers. The GHBQ and the WW-H allow psychologists, physicians and psychiatrists to do a quicker and more exhaustive evaluation of general health beliefs and false beliefs associated to worry about health, in less costs. These questionnaires will improve the chances of success of the hypochondriasis' treatment by helping clinicians to detect and correct false beliefs more easily.