The main finding of this study was a significant delay in peak latencies of cortical potentials to sounds in FAD mutation carriers relative to noncarriers from the same families. Delays in ERP latencies were observed in subjects having minimal or no cognitive deficits, and thus may indicate changes in cortical function prior to the clinical expression of AD. FAD carriers also had smaller slow wave amplitudes and larger P200 amplitudes.
Auditory ERPs are generated by synchronous activity in populations of cortical neurons that are time-locked to stimulus onset.15
Prolonged ERP latencies, in particular the P300, are observed in many neurologic and psychiatric disorders such as dementia,16
closed head injury,17
and major depression.19
The N200 has been studied less intensively than the P300, but also has a longer latency in older MCI and AD patients compared to age-matched controls.20
The N200, P300, and slow wave are considered cognitive potentials that are present when selectively attending to a task,21
while the N100 and P200 are typically considered sensory components that can be influenced by cognitive factors.22,23
The N100 and P200 largely reflect auditory cortical responses,24,25
while the P300 is generated by parietal, temporal, and prefrontal association cortex.26
Taken together, ERP components can be affected by cortical dysfunction in subjects either with or without notable cognitive deficits, but changes may not be specific to a particular neurologic or psychiatric disorder.
There were no significant group differences in behavioral measures, indicating ERP differences among groups are not secondary to different task demands. The 2 groups also did not differ in the clinical and cognitive screening tests used for dementia. Clearly neuropsychological tests would be needed for a comprehensive cognitive assessment. However, the absence of frank cognitive deficits in FAD subjects suggests that ERP measures were sensitive to the effects of specific gene mutations prior to the clinical expression of AD neuropathology. Note that the results were unaffected even when 2 subjects with somewhat low MMSE scores were removed. Discriminant analysis using ERP measures in the small subject pool of this study showed promising sensitivity (86%) and specificity (82%). The sensitivity and specificity results were comparable to studies using ERP8
and EEG measures27
in larger groups of MCI and sporadic AD subjects, although more FAD subjects would be needed for direct comparisons.
The pathology underlying delays in ERP latencies in persons with PSEN1
mutations is uncertain. Decreases in whole brain and medial temporal lobe volume28
and hypometabolism in temporoparietal cortex29
have been reported to occur during the presymptomatic period in FAD. The relationship of these changes to ERP latencies, however, is unclear. Compromise of white matter tracts is another candidate mechanism for slowed cortical processing in FAD mutation carriers. Imaging of white matter integrity in FAD carriers using diffusion tensor MRI found decreased fractional anisotropy in FAD carriers vs noncarriers.30
Decreased fractional anisotropy was seen for total white matter and select regions of interest including the corpus callosum and connections of the medial temporal lobe, and has been related to decreased myelination or diffuse white matter damage. Given that the ERPs in this study originate in neocortex and are assumed to reflect coordinated activity among multiple cortical regions,26
we speculate that latency delays in FAD carriers may reflect, in part, white matter damage.
Previous studies of older subjects with MCI show that P50 amplitude and P300 latency distinguish MCI from controls,31,32
while N100 and P200 latencies did not differ. Increases in P300 latency are present in both amnestic MCI and FAD mutation carriers. Both amnestic MCI and FAD carriers have abnormal early auditory ERPs, but the specific components differed between groups (MCI and P50, FAD and P200). Amplitudes of the P50 and N100 are greater in amnestic MCI vs older controls,31,33
with the largest amplitudes in MCI subjects who later convert to dementia within 1–4 years.8
In the present study there was a small amplitude increase of an auditory ERP component in FAD carriers, but it was the P200 rather than P50 or N100. The group effect was attenuated somewhat by one noncarrier subject who had a very large P200 component (2.1 SD), but group differences were nonetheless significant at lateral sites. Unlike the findings in this smaller sample of FAD subjects, overall amplitudes and latencies of the P200 do not differ in amnestic MCI or AD relative to older controls.9,31
Factors such as the expected time until dementia, subject age, and disease processes may account for differences between ERP results in amnestic MCI and sporadic AD vs findings in FAD carriers in the current study. Auditory ERP differences in amnestic MCI were most apparent in those subjects who converted to AD within 4 years.8
In contrast, FAD mutation carriers in the current study would not be expected to have diagnosable dementia for more than a decade. Age may also be an important factor, as the impact of AD pathology and possible compensatory responses may differ in younger and older subjects. Although postmortem studies show similar results for FAD and sporadic AD, there are some neuropathologic differences.6
Thus differences in disease processes may contribute to ERP differences in amnestic MCI and sporadic AD vs FAD subjects.
Most of the subjects in this study were women. One prior study reported sex differences in P300 latency in older healthy controls that depended on APOE
The ε4 allele is a risk factor for AD,35
but all subjects had normal scores on cognitive testing. In women, but not men, APOE
ε4 genotype was associated with longer P300 latencies than subjects having only ε2/ ε3 alleles. Consequently, we caution against generalizing the results from this study to men.