- Top of page
Huntington's disease (HD) is a neurodegenerative disease caused by a gene mutation located on chromosome 4 (HDCRG, 1993). Its clinical manifestation is heterogeneous, with symptoms and signs occurring in three domains: motor, psychiatry, and cognition (Tabrizi et al., 2011).
Attentional deficits in HD patients have been widely demonstrated (Aron et al., 2003; Muller et al., 2002; Peavy et al., 2010). In pre-manifest HD subjects, i.e., gene carriers without overt clinical symptoms, differences in attentional processing compared with controls have also been found (Lawrence et al., 1998), but with inconsistent results (Campodonico, Codori & Brandt, 1996; Farrow et al., 2007). Reports on longitudinal change in attentional functioning in pre-manifest HD are few (Beglinger et al., 2010; Lemiere, Decruyenaere, Evers-Kiebooms, Vandenbussche & Dom, 2004; Verny et al., 2007), and again contradictory results have been found (Hart, Middelkoop, Jurgens, Witjes-Ane & Roos, 2011; Witjes-Ane et al., 2007).
A practical test to investigate attentional processing is the Sustained Attention to Response Task (SART; Robertson, Manly, Andrade, Baddeley & Yiend, 1997), a simple Go/No-go task with little motor involvement. The P300, an event-related potential (ERP) that can be deduced from the EEG, is proposed to be an electrophysiological substrate of attentional and inhibitory processes (Duncan et al., 2009; Kok, 1997). In combination, these two assessments have demonstrated ability to detect lapses in attention (Datta et al., 2007).
Hart et al. (2012) combined the SART with a simultaneous P300 registration to investigate attentional functioning in both a manifest and a pre-manifest HD (PMHD) group cross-sectionally. They demonstrated that attentional control was deficient in manifest HD subjects, apparent in the inability to directly resume task requirements after having made an error. The manifest subjects showed higher error rates corroborated by abnormalities in the P300 signal. While the attentional control deficit in manifest HD (MHD) was evident, the performance of the PMHD group did not differ from that of controls, and no P300 abnormalities were found.
Recent MRI studies reported early brain changes involving grey and white matter (van den Bogaard, Dumas, van der Grond, van Buchem & Roos, 2012; Bohanna, Georgiou-Karistianis, Hannan & Egan, 2008) in pre-manifest gene carriers even far from expected disease onset. This raises the possibility that these changes can also be measured with functional assessments, such as P300. Indeed, differences in ERPs between PMHD subjects and controls have been found (Nguyen, Bradshaw, Stout, Croft & Georgiou-Karistianis, 2010). A majority of these studies have found reduced neurophysiological measures in the absence of abnormal clinical performance in the pre-manifest groups. This coincides with findings in MRI studies where pathological changes in PMHD gene carriers are seen before onset of clinical symptoms (Bohanna et al., 2008)
To date, one longitudinal study has been performed using electrophysiological assessment in HD. Here, somatosensory-evoked potentials were studied in a group of MHD subjects. The amplitude of these potentials demonstrated progressive decline over the 2-year follow-up period (Ehle, Stewart, Lellelid & Leventhal, 1984). To our knowledge, no longitudinal electrophysiological research has been performed in PMHD subjects yet.
The current pilot study aimed to investigate attentional deficits in PMHD and MHD subjects longitudinally. Observed early brain changes in PMHD literature lead us to expect possible longitudinal change for SART error scores and P300 characteristics over the 3-year interval in this group. For the manifest group, we expected to find abnormalities similar to those found earlier by Hart and colleagues: longer reaction time, more errors, and a lower No-go P300 amplitude than controls and PMHD subjects. Furthermore, due to the progressive nature of HD, we expect a longitudinal worsening of these outcome values.
- Top of page
In this pilot-study, we present preliminary data on longitudinal SART and simultaneous P300 assessment in HD. To our knowledge, we are the first to perform a longitudinal ERP study in pre-manifest HD. We found that over a course of 3 years, MHD subjects showed an increase in RT on the SART, and exhibited a prolongation of the P300 latency on specific trials of the SART. PMHD subjects only showed specific increase in reaction time just before having prevented a possible error on the SART.
In the MHD group, we replicated the findings of the cross-sectional study by Hart et al. (2012) on the SART and P300 in HD where MHD subjects make more errors on the SART than both other groups, and took longer to react to the trials (follow-up visit). The increase in reaction time of the patients compared with the other two groups is not surprising, and is likely to reflect motor slowing. This conclusion is strengthened by only specific, and not overall, RT increase in the PMHD group. In the earlier study, the authors hypothesized that the observed slowing could reflect a speed-accuracy trade-off, a (conscious or unconscious) strategy applied by MHD participants to maintain task requirements (Samavatyan & Leth-Steensen, 2009). The increase in RT seen in the current MHD cohort does not, however, prevent these subjects from making more mistakes than the other two groups. So, there is no task-related benefit from the motor slowing. One explanation for this absence of benefit could be that the employment of the speed-accuracy trade-off has reached a maximum. If this level is reached, the impairment can no longer be compensated and task-related benefit is absent.
Longitudinally, the increase in RT of the MHD group was also significantly larger than the slowing of the PMHD and control subjects, which was expected in view of the neurodegenerative nature of HD. Neurodegeneration is also reflected in the neurophysiological data, as demonstrated by the cross-sectionally longer P300 latencies compared with the other groups, and the latency prolongation over time on Go trials. The P300 latency has been linked to stimulus evaluation time, or generalized, cognitive processing speed (Polich & Criado, 2006), so a lengthened latency in the MHD group could reflect reduced processing speed. Deficits in processing speed have also been demonstrated by other authors using other paradigms (Duff et al., 2010; Maroof, Gross & Brandt, 2011).
The significant post-error slowing for the MHD group that was found in the previous cross-sectional study was again observed in the present study. However, in the current analyses, it was not a specific finding, as slowing over the 3-year follow-up period was also found for trials directly preceding and following a trial where the response was correctly withheld to the appearance of the number 3. These findings most likely reflect the general motor slowing in patients that was already discussed, rather than specific cognitive slowing in response to being confronted with a No-go trial.
Cross-sectionally, pre-manifest subjects perform equally to control subjects, evidenced by the lack of cross-sectional results. Interestingly, in the PMHD group, we did find specific slowing over time. During follow-up, this group became slower in reacting to trials directly preceding correct No-go trials (i.e., correctly withholding the response to the appearance of ‘3’). This slowing was also seen in the MHD group, but only MHD subjects produced more errors than controls. The exact significance of such specific deterioration is difficult to explain, but as the PMHD subjects make the same number of errors as controls and increase in RT where controls do not, this could mean that it takes more attentional demands for PMHD subjects to maintain task performance. As PMHD subjects can keep with the demands of the SART, it could mean that the RT increase is the result of some kind of compensatory mechanism. That it reflects pure motor slowing is less likely as in this case we would have expected increased RT on more variables.
A candidate compensatory mechanism here could be speed-accuracy trade-off, a strategy that was used by MHD subjects in our cross-sectional study (Hart et al., 2012). That growing attentional demands could translate into increased RT is well described in speed-accuracy trade-off literature (Samavatyan & Leth-Steensen, 2009). Additionally, the controls did not slow over time on these specific trials, and produced a similar amount of errors as the PMHD subjects. This strengthens our hypothesis that some kind of compensation is utilized by the latter group. This proof of early compensation by means of speed-accuracy trade-off in PMHD complements our earlier hypothesis that in the current MHD cohort, the maximum level of compensation (by means of speed-accuracy trade-off) is reached and no more task-related benefits are observed.
That compensatory mechanisms are at work in PMHD has been suggested before (Feigin et al., 2006). In another basal ganglia disorder, Parkinson's disease (PD), compensatory mechanisms responsible for delaying overt symptom onset have long been accepted (Zigmond, Abercrombie, Berger, Grace & Stricker, 1990). Here, even several compensatory networks have been identified postponing the final appearance of parkinsonism (Bezard, Gross & Brotchie, 2003). Looking at our overall raw data, we also see that for SART errors and mean RT scores, the pre-manifest subjects perform better not only than the MHD, but also than the control subjects. This could reflect overall compensation to perform at pre-disease level.
As our preliminary data derived from small pilot groups, our conclusions should be considered with caution. Generalization of our results is limited as, due to the sample sizes, statistical power is low. Also, individual fluctuations in such a small sample could have large effects, especially in the MHD group, which consisted of 5 subjects. This is evident in the differences in RT patterns surrounding incorrect responses to a No-go trial, where, at baseline, some post-error slowing was visible (albeit not statistically significant) for MHD and PMHD, while it was absent at follow-up.
The combination of cognitive testing (SART) with simultaneous P300 registration is a strength of this study, in the way that the P300 is independent of motor slowing whereby conclusions about cognitive functioning can be made more easily. Following the specific result in the PMHD group, we recommend replication of this pilot study with larger numbers of participants to be better able to understand the transition from pre-manifest to manifest HD.
In conclusion, this pilot study partly replicated the findings from the cross-sectional study by Hart et al. (2012) that MHD subjects perform worse than both PMHD and control subjects on tests of attentional control. MHD subjects are slower and make more mistakes on the SART, and show longer P300 latencies. Longitudinal change in attentional control was observed for specific trials in the PMHD subjects, which could be suggestive of compensatory mechanisms in this phase of HD.