The CNSM test was computer-administered, and the resultant scores were recorded. The MMSE and WMS Immediate Memory (IM) Index (combining immediate verbal with immediate non-verbal memory scores) were recorded as well as the Wechsler Working Memory (WM) Index Score. The TOVA was computer-administered, and the omission, commission, response time, and variability scores were recorded. Clinical and cognitive interviews also were conducted, and based upon combined assessments the patients were categorized into 4 MCI domains: amnestic single-domain, amnestic multi-domain, non-amnestic single-domain, or non-amnestic multi-domain.
Statistical Analyses
Student’s t-test was calculated between the HM group and the NM group, as well as between subjects with and without memory complaints, for P300 latency and amplitude, TOVA reaction time (RT), MMSE score, WMS IM, and CNSM.
It is noteworthy that TOVA RT measures continuous performance, which requires action, whereas P300 is an internal cerebral reaction performance test that does not require action. Consequently, the difference between P300 latency and TOVA RT correlates to the time difference between thought and action. Thus, we calculated 2 different scores using simple transformation mathematics. In one case, the difference between the P300 latency and TOVA RT (P300-TOVA Diff) was calculated, as well as the absolute value of this difference (Abs Val Diff). Moreover, to account for age differences in P300, a transformation of raw P300 latency was calculated as a difference score between the obtained latency and the predicted latency from the age adjustment [obtained latency–(300+age)], such that positive numbers reflect slower latencies and negative numbers reflect faster latencies (P300AgeDiff). Logistic regression was then utilized to classify patients into groups based on the neurophysiology and memory measures as predictors adjusted to the percent of the sample being predicted.
RESULTS
Correlation of PET with Evoked Potentials and Neuropsychological Test Results
Characteristics of the study sample, consisting of 43 patients with HM and 187 patients with NM, are listed in Table 2. The HM group was significantly older (t=3.52, p=.007) than the NM group. Table 2 also displays statistics for the memory scores and neuropsychological (P300) measures between the HM and NM groups. As expected, the HM group had lower scores than the NM group for the MMSE (t=-2.73, p=.01) and CNSM (t=-2.82, p=.0001). The WMS IM Index (e.g., Verbal + Visual) and WMS WM Index were not significantly different between the 2 groups. Using a cutoff of -1.5, SD subjects (as recommended by the test details are considered to be borderline at the -1.5 SD) with impaired WMS IM Index scores did not differ significantly from those with intact scores when compared with hypometabolic and normal metabolic groups.
Table 2: Means, SDs, and t-tests of HM vs. NM subjects
| Hypometabolic n = 43 | Normal n = 187 | Significance | |
| Age | 59.8 + 19.3 | 54.1 + 12.8 | 0.007 |
| P300 Latency | 346.7 + 29.7 | 327.5 + 27.3 | .00006 |
| P300 Amplitude | 3.5 + 2.4 | 4.7 + 2.3 | .003 |
| P300AgeDiff | -15.7 + 26.1 | -26.6 + 26.5 | .015 |
| TOVA* Reaction Time (RT) | 346.8 + 29.7 | 327.5 + 27.3 | .00006 |
| P300–TOVA RT | 83.8 + 95.0 | 30.5 + 67.0 | .002 |
| Absolute Value of Difference of P300–TOVA RT | 93.7 + 85.0 | 54.1 + 49.8 | .009 |
| MMSE* | 26.3 + 5.1 | 28.7 + 1.9 | .010 |
| CNSM* | 84.7 + 31.5 | 100.5 + 18.6 | .000099 |
*TOVA = Test of Variables of Attention; MMSE = Mini-Mental State Exam; CNSM = Memory score from the CNS Vital Signs Test.
However, the comparison of NM/HM to non-impaired/impaired CNSM scores (-1.5SD) was significant (χ2=10.7, p=.003), with more HM patients presenting poor memory (n=11/23) than those with NM (n=19/156). Poor scores on the MMSE (-1.5SD) also differed