Aim of the study

In the present study we endeavour to gain more insight into pre-test psychological functioning of subjects who applied for predictive testing for HD. Psychological functioning was conceptualised as ego strength, general anxiety, and depression level. Hypotheses were tested regarding the role of the following variables: (1) the test participant's age at the (perceived) parental onset of HD, (2) the sex of the affected parent, (3) the perception of the risk for HD, and (4) the subjective proximity of the disease. Moreover, we investigated whether these four variables were associated with post-test psychological functioning (one year after the test).

Hypotheses

The possible role of the test participant's age at the (perceived) parental onset of HD and of the affected parent's sex is based on publications on child psychiatry. Several studies have shown that traumatic events and negative, enduring circumstances during childhood, such as parental loss or parental mental illness, are associated with adult affective symptoms like depression or anxiety.20-24 A crucial aspect of the childhood condition that makes it liable to provoke adult symptoms is the child's age; distress is more marked the younger the child. Age below 5 years is critical25 because attachments are first becoming established at this age and because cognitive skills are lacking at this time. The role of the sex of the parent proved to be less clear. Given the mothers' greater concern and feelings of responsibility for child rearing in our society, one might expect that more adult problems would occur if the negative event concerns the mother of the child rather than the father, but the evidence is not straightforward.21-26 In HD, patients often have young children when the symptoms become manifest. Folsteinet al 27 stated that patients quickly become handicapped in their role as parents because of the illness. The increasing irritability, emotional lability, and lack of judgement of the HD patient interfere with adequate parental behaviour. Moreover, psychotic symptoms, such as major affective disorders, may be present. The partner of the affected person plays a major role in managing the care of the patient, often at an enormous personal price. Many spouses suffer from anxiety disorders, depression, and psychosomatic complaints.28-31 They have the responsibility for taking care of the children, while struggling with their own feelings of sadness, guilt, hostility, or social isolation. Children of an affected parent are not only at genetic risk, but also at psychological risk. Folstein et al 27 have described conduct disorders in the offspring of patients with HD that were related to the instability of the parent, the lack of discipline and structure in the family, and marital conflict.

The first hypothesis states that the younger the test participants were at the onset of HD in their parent, the more vulnerable they are to adult adverse reactions, like depression or anxiety, and the less ego strength they have. The adult distress will be especially marked in those who were in early childhood when the disease started in their parent.

The second hypothesis states that adverse adult reactions are more pronounced when the mother is affected than when the father is affected. Kessler32 also drew attention to the possible role of gender as a major factor in disorganising the family: “when the mother is affected it appears to have a greater impact on effective family functioning than when the father is affected”. Moreover, we would expect an interaction of the child's age at parental onset of HD and the sex of the affected parent. We predict that the effect of the parental disease on adult psychological functioning is larger if the mother becomes affected in the offspring's early childhood than if the father becomes affected, while we would expect no such difference if the parent becomes ill beyond the offspring's childhood.

Hypothesis 3 concerns the relationship between the subjective proximity of the disease and the level of anxiety and depression. The impact of temporally distant events on present behaviour and affect depends on their distance; proximity increases and distance diminishes the influence of the event.33 We predict that the nearer the perceived onset of HD, the higher the pre-test anxiety and depression will be in the test participants. Moreover, this association will still be significant after controlling for the subject's age at the parental onset of the disease. Regarding the post-test period we would expect a negative relationship between the subjective proximity of the disease and the level of anxiety and depression in carriers, while in non-carriers we would expect no relationship.

The fourth hypothesis concerns the relationship between the subjectively perceived risk for HD and the level of anxiety and depression. Major cognitive theories of depression (Beck's cognitive model,34 the learned helplessness theory35) emphasise the role of hopelessness about the future in the aetiology and maintenance of depression. Hopelessness is defined as the belief that desirable outcomes are highly improbable, that adverse outcomes are highly probable, and that one is helpless to change these outcomes. Empirical research36 37 has moreover shown that depressed or anxious subjects make more negative predictions of future adverse events than normal subjects. We predict that a high perceived risk in the pre-test period (predictive pessimism) is associated with higher pre-test anxiety and depression than a low perceived risk (predictive optimism) or an “accurate” risk perception. Moreover, we expect an interaction effect of the subjective proximity of HD and the perceived risk on psychological well being; persons with predictive pessimism will be more vulnerable to anxiety and depression as the disease comes closer than persons with a more accurate risk perception or with predictive optimism.

The second part of hypothesis 4 concerns the post-test period. The Vancouver group1 3 38 39 found that post-test symptoms of depression and anxiety were more likely in those who received a test result contradictory to the expected outcome. Subjects with pre-test predictive optimism (low perceived risk for HD) who received a bad result and those with pre-test predictive pessimism (high perceived risk for HD) who received a good result are expected to be more depressed and anxious after the test than participants who received a test result that corresponded to their pre-test expectations.

Although some of the four predictor variables may be correlated, we predict that each will have an independent contribution to the participant's pre-test well being. In other words, we expect that it will be possible to identify the proportion of the variance in the participant's pre-test functioning that can be attributed uniquely to each of the four variables.

Results

Table 1 presents some sociodemographic data for carriers and non-carriers. The differences between the groups were not statistically significant (chi-square and Kolmogorov-Smirnov test for independent samples and two tailed ttest47). Table 2 summarises the data for the dependent and the independent variables. The differences between carriers and non-carriers were not significant (chi-square and Kolmogorov-Smirnov test for independent samples and two tailedt test). The direct testing group did not significantly differ from the linkage group. The means of the pre- and post-test ego strength, anxiety, and depression scores were in the normal range. Individual psychometric scores showed that the number of test applicants with at least a mild level of depression (BDI >10, corresponding with approximately the 80th centile in the general population)43 was 11 (16%) in the pre-test period and 6 (9%) in the post-test period (one year after the test). The number of participants with a high score for anxiety (STAI trait >80th centile)41 was 8 (12%) in the pre-test period and 7 (10%) in the post-test period.

Table 3 displays the intercorrelations of the variables in this study (Pearson correlation coefficient). The correlations of the perceived proximity of the disease onset and of perceived risk with post-test psychometric scores were presented separately for carriers and non-carriers because different predictions were made for the two groups.

HYPOTHESIS 1

We found low but significant correlations between the test participant's age at parental onset of the disease (TAPO) and pre-test psychological functioning (table 3); only 5.8% of the variance of ego strength, 6.2% of the variance of anxiety, and 7.8% of the variance of depression level in the pre-test period was explained by the test participant's age at parental onset of HD. We found no significant correlations of TAPO with post-test psychological functioning. We specified in the first hypothesis that adult distress is most marked if the adverse condition starts in the early childhood of the test participant. To test this hypothesis, two groups of test participants were identified, those who were in early childhood when the disease started in their parent (TAPO ⩽5) and those who were older at the parental onset of HD (TAPO >5). One tailedt tests were conducted to compare the means on the psychometric tests for the two groups. The results are in line with the interaction hypothesis for pre-test psychological functioning (table 4), but not for post-test functioning.

HYPOTHESIS 2

The sex of the affected parent was not significantly associated with pre- and post-test psychological functioning in the test participant (table 3). To test the interaction hypothesis, an analysis of variance was performed with pre- and post-test ego strength, general anxiety, and depression level as dependent variables and with the child's age at parental onset of HD (TAPO), sex of the affected parent, and their product term as independent (categorical) variables (not in table). The variable TAPO was reduced to two levels (TAPO ⩽5 and TAPO >5). The main effect of TAPO was significant (p<0.01) for pre-test functioning. The interaction effect of TAPO and the sex of the affected parent on psychological functioning was not statistically significant. Moreover, we investigated whether there was an interaction effect of the sex of the test participant with the sex of the affected parent, but this was not found in our data.

HYPOTHESIS 3

Table 3 shows that the subjective proximity of the disease onset (SP) is significantly correlated with pre-test level of anxiety (r=−0.41) and depression (r=−0.56). It is also significantly correlated with the carriers' post-test depression level (r=−0.32), but not with their anxiety level (r=−0.23, p=0.10). This means that carriers are more depressed when the onset of the disease comes closer. As expected, the correlation was not significant for non-carriers. To test the interaction between the subjective proximity and test result, we performed two hierarchical multiple regression analyses with post-test anxiety and depression as dependent variables and with subjective proximity, test result, and their product as independent variables. However, the interaction term was not significant. These results did not change after elimination of the three participants with a negative SP score. After adjustment for the test participant's age at the parental onset of the disease, the contribution of SP remained significant for pre-test anxiety (p<0.05) and depression (p<0.001) but not for post-test depression level in carriers (p>0.05) (not in table).

HYPOTHESIS 4

Table 3 shows that a higher perceived risk for HD in the pre-test period is positively correlated with higher pre-test depression (r=0.33, p<0.01), but not with a higher anxiety level (r=0.14, p>0.05). To test the interaction hypothesis, hierarchical multiple regression analyses were conducted, with the perceived risk, the subjective proximity of the disease, and their product as predictors. The results (table 5) show that the effects of the perceived risk (PR) and the subjective proximity (SP) were both significant for pre-test depression. For pre-test anxiety, only the subjective proximity had a significant effect. In addition, there was a significant interaction effect of the perceived risk and the subjective proximity on pre-test depression and anxiety. The total variance explained (R²) increased from 0.38 to 0.52 for depression and from 0.17 to 0.27 for anxiety. The interactions remained significant after the elimination of the three participants with a negative proximity score.

In order to understand the nature of the interaction effects, we have illustrated them in figs 1 and 2. Therefore, categorical variables were introduced for PR and SP. For PR we made two categories, high risk perception and accurate or low risk perception. For SP the categories were based on the median SP score (near=smaller than the median SP score (mean SP score=1.9 years, SD=7.3, n=29) and far=larger than the median SP score (mean SP score=19.0, SD=4.1, n=31)).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Interaction between risk perception and perceived proximity on depression.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Interaction of risk perception and perceived proximity on anxiety.

To test the second part of the hypothesis, we performed hierarchical multiple regression analyses with post-test general anxiety and depression levels as dependent variables and with perceived risk, test result, and their product as independent variables (those with an “accurate” or a “vague” risk perception were not included in the analyses). We did not find a significant interaction effect of perceived risk and test result on post-test depression and anxiety.

The contribution of the combination of the four mentioned variables to predicting pre-test adjustment was investigated by means of multiple regression analyses (table 6). This resulted in an explained variance of 0.20 (F(4,59)=3.35, p<0.05) for pre-test ego strength, 0.41 (F(4,59)=9.54, p<0.001) for depression, and 0.18 (F(4,59)=3.08, p<0.05) for anxiety. The subjective proximity of the onset had a significant unique contribution to predicting the three post-test variables. The perceived risk also explained unique variance of pre-test depression level. Adding the interaction effect between perceived risk and perceived proximity of the disease, in a second step, contributed significantly to the explanation of the variance in pre-test ego strength, depression, and anxiety, resulting in R²=0.29 (F(5,59)=4.32, p<0.01), R²=0.52 (F(5,59)=11.82, p<0.001), and R²=0.28 (F(5,59)=4.15, p<0.01).

With regard to post-test psychological functioning (one year after the test), none of the four selected variables had a significant value in the multiple regression equations. In the next step, the pre-test psychometric tests were added first in the regression equation, followed by the four variables mentioned. However, the results of the multiple regression analyses showed that none of the four variables explained unique variance of the post-test psychological measures, beyond that accounted for by the pre-test measures. The pre-test psychometric scores accounted for 54% of the variance of post-test ego strength (F(1,59)=67.26, p<0.001), 18% of the variance of post-test depression (F(1,59)=12.62, p<0.001), and 30% of the variance of post-test anxiety (F(1,59)=25.26, p<0.001). Adding the test result to the previous three models did not significantly improve the R².

Discussion

The present paper aimed to delineate variables associated with pre-test distress of applicants for predictive testing for HD. Based on theoretical considerations, hypotheses were tested regarding the role of (1) the test participant's age at the (perceived) parental onset of HD, (2) the sex of the affected parent, (3) the perception of the risk for HD, and (4) the subjective proximity of the disease. We found low but significant Pearson correlations between the participant's age at the parental onset and pre-test psychometric measures of psychological functioning: the younger the test applicant was at the parental onset of HD, the more likely it was that this person had a lower ego strength, more anxiety, and a higher depression level in the pre-test period. However, this variable did not explain unique variance of pre-test psychological functioning, beyond that accounted for by the other variables in the model. Its effect may be indirect through its influence on subjective proximity of the disease. Moreover, a possible secondary effect of the young age of the proband at the parental onset of the disease may be the length of exposure to the illness in the parent during key developmental stages, particularly in early adolescence, when this would coincide with more severe illness in the parent.

The lack of a unique contribution of the participant's age at parental onset can also be rooted in the self-selection of the sample. As already stated in the introduction, publications on predictive testing for HD have shown that test participants are on average a self-selected group of mentally resourceful people. It cannot be excluded that persons who went through a very traumatic childhood owing to the parental disease do not ask for the test. For these people, having certainty that, one day, the disease in their parent will also happen to them may be unbearable.

Since it was very difficult, if not impossible, to collect reliable information about childhood experiences by means of retrospective reports during the pre-test counselling, we had insufficient data about the early negative experiences of the subject with the affected parent or with a disorganised family. It is plausible that early psychiatric symptoms in the parent are more traumatic for the young child than early cognitive impairment and uncontrollable movements. However, the possible effect of the interaction between the participant's age at the perceived parental onset of HD and the early negative experiences with the disease could not be taken into account in the study. A prospective, longitudinal study of HD families, including (future) test participants and non-participants, might shed more light on these questions.

We found no significant relation between the sex of the affected parent and the test participant's psychological adjustment. One explanation is that the effect of maternal disease in early childhood depends on the extent to which the father or a significant other person can replace the mother figure for the child. Overall, several protective factors in the child and in the environment may compensate for maternal mental illness, and parental mental illness in general, in early childhood. Good intelligence and good scholastic attainments exert a protective effect.25 However, the most important protective factor proved to be the subject's social network; a supportive and understanding relationship in childhood and close personal relations in later life proved to reduce the effect of childhood stressors on later adult psychological functioning.23 48 49

The expected association of the subjective proximity of the disease onset and pre-test depression and anxiety was confirmed; a closer perceived proximity of the onset of HD was uniquely associated with more depression and anxiety in the test participants and this relationship remained significant after adjustment for the participant's age at parental onset.

We hypothesised that test participants with predictive pessimism (high risk perception) in the pre-test period would be more anxious and depressed before the test than persons without predictive pessimism. This was confirmed for depression, but not for anxiety. Neither theoretical considerations nor the statistical method allowed the direction of the association between predictive pessimism and pre-test distress to be ascertained. Does a high risk perception cause depression or, conversely, does depression cause a pessimistic risk perception, or is their relationship reciprocal? Another possibility is that both depression and a pessimistic risk perception are the result of the influence of a third set of variables, including the high perceived severity of the disease50 or the lack of perceived controllability51 or both.

The expected interaction effect of subjective proximity of HD and perceived risk on pre-test depression and anxiety was supported; persons with predictive pessimism were more anxious and depressed as the disease came closer than persons with a more accurate risk perception or with predictive optimism. The mean pre-test depression score of test participants at risk for HD with a high perceived risk for HD who were quite close to the perceived onset of the disease reached a value of 10.7 which corresponds to at least a mild depression level; the variability within this group is however large (SD=9.9). The mean scores of the other three groups were significantly lower and remained at the level of the general population. Test participants at risk for HD who are close to the perceived onset of the disease and who have a pessimistic risk perception should be given special attention during pre-test counselling because of their possible negative affective condition at that time.

One year after the test, about 10% of the tested subjects, carriers as well as non-carriers, had at least a mild depression level or a high score for anxiety or both. Pre-test psychological measures were the best predictors of post-test distress, irrespective of the test result, as shown in previous research.5 15-18 In this study we included other variables, such as the participant's age at the parental onset of the disease, the sex of the affected parent, the perceived risk for HD, and the perceived proximity of the disease, but these did not significantly increase the explained variance of post-test functioning, beyond that accounted for by the pre-test psychometric scores. Contrary to Codori et al,5 who found that the perceived years to disease onset was a significant predictor of the post-test hopelessness of carriers (measured with Beck's Hopelessness Scale52), we found no significant unique contribution of the subjective proximity of the disease onset to explain the level of anxiety and depression of carriers one year after the test. It is possible that the effect of the perceived proximity of the disease is reflected in more disease specific distress of carriers, such as intrusion/avoidance processes, rather than in more general anxiety and depression.53Another (associated) explanation is that some carriers use more minimisation and denial as coping mechanisms as the onset of the disease comes closer, which may result in an attenuation of the current level of general anxiety and depression scores.

The expectation that post-test symptoms of depression and anxiety would be higher in those who received a test result contrary to the expected outcome than in those whose result corresponded to their pre-test expectations was not supported by our data. The small number of persons with a low risk perception may have been an obstacle in this context; only four out of 60 participants thought that they were (probably) not a carrier of the HD gene. The hypothesis that, among carriers, those with a low pre-test perceived risk for HD would be more hopeless and depressed than test participants with high perceived risk had also been formulated by Codori et al,5but was not confirmed by their data either.

The present study focused on the contribution of environmental influences (such as negative childhood experiences with an affected parent) and of aspects associated with the genetic nature of the disease (such as perceived genetic risk and perceived proximity of the disease onset) to explain pre- and post-test distress. A large part of the variability in psychological distress remains unexplained. Other variables, such as the condition of affected sibs, the quality of the partner relationship, or the economic resources, may have a considerable impact on the well being of test applicants. The applicant's strategies to cope with a threatening situation can also influence pre- and post-test distress. During counselling, it is important to keep in mind that test participants use psychological defence mechanisms like optimistic thoughts or denial and minimisation to manage strong negative emotions.15-19 Although the role of social support and family systems has been considered very important in coping with threat,3 48 54 it has received little attention until now. Tibben15 reported that satisfaction with the perceived quality (rather than the number) of supportive other people was associated with low avoidance of HD related situations and with hopeful expectation about the future. Future longitudinal research should more directly investigate the role of negative childhood experiences owing to living in a family with a HD parent and the possible protective effect of social relationships, during childhood and also in later life, in explaining psychological well being. It is also important to address the question whether these aspects are associated with the uptake of predictive testing for HD and with the avoidance attitude and the feelings of hopelessness of non-participants.9 55 56 This research should enable clinical geneticists and mental health professionals to refine the pre- and post-test counselling strategies for predictive DNA testing, not only for HD, but also for other serious late onset diseases.