Abstract
Background:
Previous research using the Symptom-Checklist-90-Revised (SCL-90-R) has shown that compulsive symptoms in Huntington’s disease (HD) occur in 10% to 52% of all cases. The “Hamburger-Zwangsinventar” (HZI), a self-rating questionnaire comprising 188 items, taps into the domain of compulsiveness in greater detail, but has not been used in HD so far. In addition, little is known about the association of obsessive-compulsive symptoms with pre-clinical stages of HD.
Objective:
Comparison of OC symptoms measured by the SCL-90-R and the HZI in pre-HD and HD.
Methods:
29 premanifest mutation carriers (pre-HD) and 40 manifest HD patients completed both questionnaires. Clinical characteristics of HD were rated by using the UHDRS.
Results:
Compared to data from general population prevalence of OC symptoms were not higher as expected in preHD and only slightly elevated in manifest HD if using HZI. Both instruments detected more OC symptoms in HD patients compared to pre-HD. The SCL-90-R more often detects compulsivity than the HZI. Results of both questionnaires showed correlations to cognition, depression, and disease duration. Compared to findings from OCD patients, there was a subordinate role in the HZI subscale for “washing and cleaning” in HD.
Conclusions:
OC symptoms in pre-HD occur not more often than in general population. The HZI appears to be useful for examining OC symptoms in detail in pre-HD and HD. HZI-subscale ratings for washing and cleaning compulsions were less pronounced in HD compared to OCD patients, possibly due to loss of disgust. The SCL-90-R might overestimate OC symptoms in both groups.
Keywords
INTRODUCTION
Huntington’s disease (HD) is an autosomal-dominant degenerative brain disorder characterized by motor, cognitive and psychiatric symptoms [1]. Aside from depression, psychosis, irritability, aggression, apathy and cognitive impairment, the broad spectrum of psychiatric problems includes obsessive–compulsive symptoms (OC symptoms) [2, 3]. OC symptoms are defined in classifications like the ICD-10 [4] as recurrent obsessional thoughts or compulsive acts. Obsessional thoughts are ideas, images, or impulses that enter the patient’s mind again and again in a stereotyped form. They are almost invariably distressing and the patient often tries, unsuccessfully, to resist them. They are, however, recognized as his or her own thoughts, even though they are involuntary and often repugnant. Compulsive acts or rituals are stereotyped behaviours that are repeated again and again. They are not inherently enjoyable, nor do they result in the completion of inherently useful tasks. Their function is to prevent some objectively unlikely event, often involving harm to or caused by the patient, which he or she fears might otherwise occur. Usually, this behaviour is recognized by the patient as pointless or ineffectual and repeated attempts are made to resist. Anxiety is almost invariably present. If compulsive acts are resisted the anxiety gets worse [5]. It is necessary but sometimes not easy to differentiate between OC symptoms and perseverations. Moreover, we have made the experience in our Huntington Centre that treatment of disturbing OC symptoms in HD is often disappointing, not at least sometimes the relatives are more suffered by the compulsivity of the patients, than the patients themselves. For this reason, we wanted to investigate more closely what the OC symptoms in HD patients are and how they could be studied more closely. OC symptoms may occur up to years before motor symptoms [6, 7]. The most common psychiatric disorder is depression, whereas data regarding the prevalence of OC symptoms are divergent. Different instruments were used to score OC symptoms [8]. Previous research using the SCL-90-R self-rating questionnaire or Yale-Brown Obsessive Compulsive Scale (Y-BOCS) [9], found OCD in HD with a prevalence of 10% to 52% [10].
An overview about the results of several studies using different instruments is given in a recent published article by Oosterloo et al. [11]. Some single case reports about OC symptoms in HD describe a good response to selective serotonin reuptake inhibitors [12–15], and discuss neuropathological changes in the ventromedial caudate nucleus and related neural circuits may be the underlying cause for the association of OC symptoms in HD. Neuropsychological investigation of 17 unmedicated patients with OC symptoms, however, revealed no evidence of cognitive impairment as is typically described for HD patients [16].
The aim of this study was the comparison of compulsive symptoms measured by the SCL-90-R [17] and the HZI in pre-HD and HD. We selected the HZI for comparison with the commonly used SCL-90-R questionnaire, since the HZI to our knowledge asks the most precise and detailed questions, with regard of the motivation and underlying psychopathology of action. Some other instruments are rather dependent scales and / or very general like the Y-BOCS, the SCOPI was not available in the German language at the time we conducted this study.
MATERIALS AND METHODS
The study was approved by the Ethics Committee of the Medical Faculty of the Ruhr-University Bochum, Germany. Patients were recruited from the in- and outpatient clinic of the Huntington Centre, NRW. Participants who had severe dementia symptoms (defined by a cognitive sum score of the Cognitive Inventory of the Unified Huntington’s Disease Rating Scale (UHDRS)<100), with other severe diseases like cancers or other neurological diseases, severe internal ailment or suicidality were excluded. The premanifest mutation carriers and the manifest HD patients completed both the SCL-90-R and the HZI questionnaire in randomized order. In addition, the participants were rated by using the Unified-Huntington-Disease-Rating-Scale (UHDRS) consisting of motor rating, cognitive testing (verbal fluency test, symbol digit test, stroop test), Total Functional Capacity and Independence Scale. Depression was assessed by using the Beck Depression Inventory (BDI) [18, 19].
The symptom-checklist-90 revised (SCL-90-R)
The SCL-90-R is a widely-used self-rating questionnaire of neuropsychological symptoms with in total 90 items, 10 of them tapping into OC symptoms [20–22]. There is no differentiation between categories of compulsive symptoms like washing and cleaning or controlling. The SCL-90-R gathers with ninety questions information about quality and intensity of several dimensions of psychopathology [22]. The items are rated on a Likert-scale from 0 (=not at all) up to 4 (=extremely). The analysis indicates the psychological stress on nine scales and three global indices of distress. The ten items of the subscale for obsessive-compulsive symptoms range from slight trouble of concentration up to severe obsessive-compulsive symptoms (see Table 1). Symptoms considered are connected to the clinical syndrome of OCD like thoughts, impulses, and compulsive behaviour, perceived as constant, irresistible, non-alterable and ego-dystonic.
SCL-90-R items of the subscale for obsessive-compulsive symptoms
The Hamburger Zwangsinventar (HZI)
Another validated German self-rating questionnaire called “Hamburger-Zwangsinventar” (HZI) asks for specific OC symptoms in more detail with 188 items, rated with “yes” or “no”. For example, the first three questions are: “1: Are you always thinking exactly what you want to say before you make a call? 2: Do you wash your hands after getting too close to an animal? 3: Do you just tablecloths or runners back when you think they were moved?” Moreover the HZI divides the symptoms into six different categories of compulsive behaviour like controlling behaviours; washing and cleaning; ordering and symmetry; obsessions and compulsions associated with counting, touching and talking; obsessions associated with words, pictures or chains of thought; thoughts and fantasies of harming oneself or others. Severity of symptoms is assessed by using a Stanine Scale (Standard score with nine categories). Beside others the HZI was used for discrimination of patients with high from patients with low pronounced symptomatology, for description of OC symptoms in OCD, Parkinson’s disease, Tourette’s syndrome or eating disorders [23–26]. The HZI has not been used in HD research so far. The collected values on the six scales of the HZI are converted in a Stanine scale by using standard value tables [27]. Here, the mean is at Stanine 5 and standard deviation is 2. For any scale, three standard reference points are stated on the scoring sheet: The mean (Stanine 5) was calculated from the test values of all OCD patients (n = 121). Only values of severe disturbed OCD patients have been used for the calculation of the second mean (extreme range). Criterion for the use for calibration of the extreme range was a subjective rating of the patient of at least 50 mm on a 99 mm scale. The third mean value (normal range) was obtained from the values of tested general population (n = 120).
Statistical analysis was performed using SPSS 19.0.0. For statistical significance a level of p≤0.05 was accepted.
RESULTS
29 pre-HD participants (17 females) defined by a positive gene test and absence of specific HD symptoms (Diagnostic Confidence Level [DCL] <4, item 17 of the UHDRS Motor Assessment) and 40 manifest, genetically confirmed HD patients (23 females) were included in this study. None of the pre-HD mutation carriers received antipsychotic medication. Sixteen of the 40 participants with manifest HD were treated with one or two antipsychotics, i.e., nine received tiapride, three tetrabenazine, four risperidone, two quetiapine and one sulpiride. Two participants from our pre-HD group and 8 from our HD group were treated with a SSRI (i.e., citalopram, escitalopram or sertraline). Other antidepressants used were clomipramine, venlafaxine, mirtazapine and duloxetine in single participants. An overview about clinically data is given in Table 2.
Demographic and clinically data
Analysis of the SCL-90-R questionnaire identified 5 overall case-findings in our pre-HD group and 22 in the HD group. T-values ≥60 should be analysed in more detail, a suspicious t-value in the SCL-90-R is defined as ≥63. Four of the five overall identified cases from the SCL-90-R in the pre-HD-group and 20 of the 22 overall identified HD cases show suspicious t-values ≥63 for the subscale compulsiveness next to others. The age and sex-corrected t-values of the scale compulsiveness of the SCL-90-R showed in the total sample a minimum of 35 and a Maximum of 80, the mean value was 57.9 (standard deviation 12.3). For the pre-HD subgroup, the age and sex-corrected t-values of the scale compulsiveness showed a minimum of 35 and a Maximum of 70 (mean: 51.1; standard deviation 12.3). The manifest HD subgroup showed a minimum of 35 and a Maximum of 80 (mean: 62.9; standard deviation 11.7).
The results of the HZI-subscales are shown in Fig. 1.

“Hamburger Zwangsinventar” (HZI) results of the subscales for premanifest mutation carriers for Huntington’s disease (pre-HD) and manifest Huntington’s disease patients (HD) with cutoff stanine 5 = OCD-symptoms compared to patients with an obsessive-compulsive-disorder.
Both instruments (SCL-90-R p < 0.0001, HZI p = 0.001) detected more OC symptoms in HD patients compared to pre-HD. The SCL-90-R more often detects compulsivity than the HZI (total test value) in both groups (pre-HD 17.2% versus 3.5%, and HD-patients 60.0% versus 22.5%, Pearson Chi-square test p = 0.016). Both questionnaires showed in the total sample (69 participants) a significant correlation to cognition, as well as depression assessed by using the BDI, and disease duration (the worse the cognition and the depressiveness, and the longer the disease duration, the higher are the findings in both questionnaires, p < 0.01 or smaller, see Table 3), but not correspondingly for the subgroups.
Correlations of SCL-90-R and Hamburger-Zwangsinventar (HZI) to cognition-test-results, depression measured by the Beck-Depression-Inventory (BDI), and disease duration
The analysis of the HZI-subscales showed a subordinate role for washing and cleaning compulsions in the manifest HD group compared to patients with an OCD (see Fig. 1).
Only two of the manifest HD-patients were known to be suffered clinically by OC symptoms. The first patient showed conspicuous data in the subscales of the HZI for “counting, touching, talking” (Stanine 5) and “arrange” (Stanine 4). The total value of the HZI was inconspicuous (Stanine 1), the SCL-90-R detected no “case” in this patient, too. Only a slight sign for “compulsivity” was shown in the SCL-90-R by further detailed examination of the subscale. The patient took escitalopram 20 mg daily without any actually signs of a clinically significant depression (BDI: 8), but ongoing clinically relevant OC symptoms. The UHDRS cognitive testing showed a cognitive impairment with a sum-score of 149 in this patient. Also, the second patient showed an inconspicuous total value of the HZI (Stanine 1), with only the subscale “arrange” being conspicuous (Stanine 4). The SCL-90-R showed inconspicuous values for all scales in this patient and likewise this patient had no signs of depression (BDI: 2, cognitive sum-score: 175).
CONCLUSIONS
According to our study, using the HZI the frequency of OC symptoms in pre-HD and the intensity of isolated symptoms of compulsivity in pre-HD is not relevantly higher than expected, compared to published numbers of prevalence from total population (1–3%) [8, 29]. However, intensity and prevalence is higher in manifest HD patients. The SCL-90-R also used in earlier studies might overestimate compulsive symptoms in pre-HD and HD. In manifest HD this is possibly caused due to participants’ misinterpretation of other HD symptoms, cognitive impairment or perseverative behaviour as OC symptoms. However, these hypotheses cannot explain the elevated results of OC symptoms in pre-HD detected by the SCL-90-R. Ultimately, the reasons for this remain unclear and would need further examination. Thus the use of self-rating scales for assessment of OC-symptoms in HD is discussed critically. The HZI-subscale for washing and cleaning compulsions in manifest HD patients was the lowest compared to other categories and remarkably less relevant compared to OCD patients, which might be caused by the known loss of disgust in HD patients [30].
DISCUSSION
The aim of the present study was to examine OC symptoms in pre-HD and manifest HD patients with two different self-rating scales that are routinely used to evaluate psychopathological signs and symptoms in both clinical and non-clinical samples. Specifically, we were interested in the question how these two scales reflect the emerge of OC symptoms from pre-HD to clinically manifest HD. Accordingly, we utilized two different scales, the SCL-90-R and the HZI, to examine the presence and severity of OC symptoms in the two groups. Remarkably, the frequency of OC symptoms in pre-HD and manifest HD in our study differs significantly, depending on which instrument was used for assessment. By using the HZI questionnaire prevalence of OC symptoms were not higher in pre-HD-group, but higher in manifest HD patients (22.5%) if compared to general population. If SCL-90-R was used, prevalence was clearly increased in both subgroups (pre-HD 17.2%, manifest HD 60.0%). Divergent data regarding the prevalence and intensity of OC symptoms are also described in literature [8]. Following our results, the use of different instruments to examine OC symptoms might be a possible explanation. Both questionnaires only ask for OC symptoms which possibly might be caused by other HD-symptoms like cognitive impairment.
Regarding the single questionnaires, the SCL-90-R was discussed as a very sensitive instrument to also assess subclinical OC symptoms in HD [16]. Even if the sample size of our study is relatively small because it is a pilot study to investigate the first time if the HZI is applicable to investigate OC symptoms in HD, our results seem to indicate, however, the SCL-90-R may overestimate OC symptoms in HD, since the HZI, revealed significantly less OC symptoms in our both HD groups. The assumption that SCL-90-R overestimates OC symptoms is also supported by findings from literature. For pre-HD, Craufurd et al. described them to be present in 10% by using the PBA assessment [31]. Mean scores were subclinical and below those of patients with diagnosed obsessive-compulsive disorder for pre-HD if the SCOPI scale was used and the symptom worrying and checking was related to cognitive errors [32]. We also found correlations between cognitive scales and OC symptoms in our study. This finding might be interpreted as a result of a neurodegenerative process with increasing symptoms. This interpretation is supported by the correlations to the disease burden score. On the other hand, however, the finding might be explained due to participants’ misinterpretation of questions form the OCD questionnaire due to cognitive impairment [33]. For example, increased controlling might be due to cognitive impairment (because somebody really is afraid of forgetting something, but without urge symptomatic to control everything) or it might be due to OC symptoms combined with an urge feeling to do so. This context might be too complex to differentiate for cognitive impaired persons in a self-rating questionnaire. In our study both questionnaires showed in the total sample more OC symptoms in the presence of poorer cognition or depressiveness and a longer disease duration, but not for the subgroups, possibly caused by a too small sample size of the subgroups. Thus, the use of self-rating questionnaires might lead to misinterpretation of results.
This difference seems to be even more relevant if manifest HD patients with higher cognitive impairment were investigated. In our group 60% of HD patients were found to have OC symptoms using the SCL-90-R questionnaire. Most published research found about 10% to 20% OC symptoms in HD patients if clinically criteria were used [9, 34]. This numbers are more or less the same, if semi-structured scales, which require an interview of an investigator, were used like the “Short Problem Behaviours Assessment (PBA-s)” [35] or “Unified Huntington’s Disease Rating Scale OC item” [31, 36–39]. For PBA-s trained interviewers have to judge symptoms and symptom severity after a structured interview. Thus, the interviewers are trained to differentiate between, e.g., OC symptoms (including aspects of urge and ego-dystonic of symptoms) and perseverative behaviour. A possible overlap of symptoms, possibly secondary to the limitations of currently used rating scales in terms of their construct when applied to HD, was also discussed in a recent review by Mestre et al. [19].
The HZI questionnaire detects OC symptoms in our manifest HD group only in about 22.5%, which is close to the 10% to 20% described above and as far as assessable similar to results from studies performed by using the SCOPI instrument [2, 40], (a factor analytically derived, self-report instrument) [39]. Thus, despite the fact that the HZI and SCOPI are self-rating questionnaires, both of these questionnaires might reflect OC symptoms in HD more adequately. Accordingly, the SCOPI but not the SCL-90-R was suggested in a recent review for use for assessing OC symptoms in HD. The PBA and Unified Huntington’s Disease Rating Scale, Behavioral Section was suggested as a screening instrument [19]. This is supported by our findings, thus the HZI might represent another appropriate scale. As a limitation, the HZI is only available in the German language so far. Therefore, it will be difficult to get further confirmation of our data by other studies in the future.
Regarding different subcategories of OC symptoms conflicting results for the category “hoarding” were described for use of the SCOPI questionnaire (5 subscales: checking, cleanliness, compulsive rituals, hoarding, and pathologic impulses). Beglinger et al. found all categories changed in HD except “hoarding” [32], whereas Gray et al. described higher findings in their pre-HD group for “hoarding” compared to controls and manifest HD, and more “checking” in their manifest HD group, but no differences for “cleanliness”, “rituals”, “pathological impulses” or SCOPI total score [3]. Remarkably, according to our results the HZI-subscale for washing and cleaning compulsions in the manifest HD group was less pronounced compared to patients with an OCD. This finding needs to be confirmed in an independent study, but might be explained by the loss of disgust described for HD patients [30].
Another effect to results of questionnaires might be caused by be the use of medication like SSRI or SNRI, which potentially modulate OC symptoms. In our study one patient with clinically relevant OC symptoms took 20 mg escitalopram, suffered not from depression, but ongoing from OC symptoms. This may be explained by the fact that 40–60% of patients with an OCD do not response adequately to SRIs [41].
In summary, results for frequency and intensity of OC symptoms in our study differ significantly, depending on which instrument was used for assessment and at which stage of the disease the assessment took place. As discussed by Mestre et al. i) the possibility that behavioural symptoms observed in HD may not fit formal diagnoses suggested by international classification systems, frequently used as “gold standard” for psychiatric diagnoses (e.g., ICD-10), ii) an overlap between the physical manifestations of psychiatric syndromes and the physical manifestations of HD may be present, iii) there might be a co-occurrence of behavioural symptoms in HD, iv) the existence of identifiable disease stages is not reflected in the validation process of behavioural rating scales and v) the presence of cognitive impairment and lack of insight of patients should be considered in scale development in HD [19]. All of the mentioned limitations can be perfectly discussed for our study, too. Even if the HZI might ask for symptoms more in detail, we conclude that there is no clear recommendation for the use of any self-rating scale for assessment of OC symptoms in HD.
CONFLICT OF INTEREST
The authors have no conflict of interest related to this work to report.
Outside the submitted work, Dr. Saft reports grants from Teva Endowed Professorship, grants from ‘Cure Huntington’s Disease Initiative’ (CHDI), grants from Biogen, personal fees from Temmler Pharma GmbH & Co.KG and Desitin Arzneimittel GmbH, outside the submitted work; he received institutional compensation and/or travel or accommodation payments in the context of the Registry Study (Euro-HD-Network), the ENROLL study (CHDI), in context of the MitoNet study, the ACR16 study (Neurosearch), the AFQ-Study (Novartis), the Selisistat-Studies (Siena Biotech), the PRIDE-HD and LEGATO-HD study (TEVA), the Amaryllis study (Pfizer), the BN82451B Study (IPSEN) and the ISIS 443139 Antisense study (IONIS-Pharmaceuticals).
Footnotes
ACKNOWLEDGMENTS
We are grateful to all participants.
