FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3175383339> ?p ?o ?g. }

• W3175383339 abstract "Article Figures and data Abstract eLife digest Introduction Results Discussion Materials and methods Appendix 1 Data availability References Decision letter Author response Article and author information Metrics Abstract Alzheimer’s disease (AD) in elderly adds substantially to socioeconomic burden necessitating early diagnosis. While recent studies in rodent models of AD have suggested diagnostic and therapeutic value for gamma rhythms in brain, the same has not been rigorously tested in humans. In this case-control study, we recruited a large population (N = 244; 106 females) of elderly (>49 years) subjects from the community, who viewed large gratings that induced strong gamma oscillations in their electroencephalogram (EEG). These subjects were classified as healthy (N = 227), mild cognitively impaired (MCI; N = 12), or AD (N = 5) based on clinical history and Clinical Dementia Rating scores. Surprisingly, stimulus-induced gamma rhythms, but not alpha or steady-state visually evoked responses, were significantly lower in MCI/AD subjects compared to their age- and gender-matched controls. This reduction was not due to differences in eye movements or baseline power. Our results suggest that gamma could be used as a potential screening tool for MCI/AD in humans. eLife digest Alzheimer’s disease is one of the most common forms of dementia, characterised by declining memory and thinking skills, and behavioural changes that worsen over time. It affects millions of people worldwide, mostly in older age, and yet early indicators of the disease are lacking. Most cases are only diagnosed once a person’s brain function becomes noticeably impaired, even though known biological changes underpin the disease. Detecting Alzheimer’s disease early could aid diagnosis and enable early intervention, while also improving the chances of finding treatments to halt or reverse the disease. Currently, brain function is measured by performing cognitive tests, such as remembering a set of words, imaging the brain with MRIs or CT scans, and blood or spinal fluid tests. Many of these tests can be invasive and expensive, so researchers are exploring whether measuring oscillations in the brain’s electrical activity can be a non-invasive and chepaer way of testing brain function. Gamma oscillations are rhythmic signals, thought to be involved in attention and working memory. Animals used to study Alzheimer’s disease have shown some abnormalities in gamma oscillations, and studies of healthy humans have also observed a decline in the strength and frequency of these oscillations with age. These findings have spurred an interest in understanding the link between gamma oscillations and AD in humans. To investigate this link, Murty et al. measured patterns of brain activity in elderly people chosen from the community using electrodes placed on their scalps (a technique called electroencephalography). These participants watched certain images previously shown to elicit gamma oscillations. Participants who were later diagnosed with early Alzheimer’s disease had weaker gamma oscillations than their cognitively healthy peers in the part of the brain that processes visual images. These results build upon previous findings from animal research suggesting that gamma oscillations may be disrupted in early Alzheimer’s disease. The work by Murty et al. could lead the way to new ways of diagnosing Alzheimer’s disease, where early indicators are urgently needed. Introduction Alzheimer’s disease (AD) is a predominant cause of dementia (decline in cognitive abilities) of old age and substantially contributes to the socioeconomic burden in the geriatric population, necessitating early diagnosis. Advances in our understanding of cellular pathology of AD in rodent models and its link to gamma rhythms in brain have spurred interest to investigate diagnostic and therapeutic potential of gamma rhythms in AD and other forms of dementia (Mably and Colgin, 2018; Palop and Mucke, 2016). Gamma rhythms are narrow-band oscillations in brain’s electrical activity with center frequency occupying ~30–80 Hz frequency range (Gray et al., 1989). These are suggested to be generated from excitatory-inhibitory interactions of pyramidal cell-interneuron networks (Buzsáki and Wang, 2012) involving parvalbumin (PV) and somatostatin (SOM) interneurons (Cardin et al., 2009; Sohal et al., 2009; Veit et al., 2017). These have been proposed to be involved in certain cognitive functions like feature binding (Gray et al., 1989), attention (Chalk et al., 2010; Fries et al., 2001; Gregoriou et al., 2009), and working memory (Pesaran et al., 2002). Some studies have reported abnormalities in gamma linked to interneuron dysfunction in AD. For example, Verret et al., 2012 reported PV interneuron dysfunction in parietal cortex of AD patients and hAPP mice (models of AD). They found aberrant gamma activity in parietal cortex in such mice. Further, some recent studies have suggested therapeutic benefit of entraining brain oscillations in gamma range in rodent models of AD. For example, Iaccarino et al., 2016 suggested that visual stimulation using light flickering at 40 Hz entrained neural activity at 40 Hz and correlated with decrease in Aβ amyloid load in visual cortices of 5XFAD, APP/PS1 mice models of AD. Based on such reports in rodents in both visual and auditory modalities, some investigators have suggested a paradigm termed GENUS (gamma-entrainment of neural activity using sensory stimuli) and have claimed to show neuro-protective effects in rodent models of AD (Adaikkan et al., 2019; Martorell et al., 2019). Recent studies in human EEG (Murty et al., 2020; Murty et al., 2018) and MEG (Pantazis et al., 2018) have reported existence of two gamma rhythms (slow: ~20–34 Hz and fast: ~36–66 Hz) in visual cortex, elicited by Cartesian gratings. Age-related decline in power and/or frequency of these stimulus-induced gamma rhythms has been shown in cognitively healthy subjects (Gaetz et al., 2012; Murty et al., 2020). However, abnormalities in such stimulus-induced visual narrow-band gamma rhythms in human patients of mild cognitive impairment (MCI, a preclinical stage of dementia [Albert et al., 2011; Petersen et al., 1999; Sosa et al., 2012]) or AD have not been demonstrated till date. We addressed this question in the present double-blind case-control EEG study involving a large cohort of elderly subjects (N = 244; 106 females, all aged >49 years) recruited from urban communities in Bangalore, India. These were classified as healthy (N = 227; see Murty et al., 2020), or suffering from MCI (N = 12) or AD (N = 5) based on clinical history and Clinical Dementia Rating (CDR; Hughes et al., 1982; Morris, 1993) scores. We studied narrow-band gamma rhythms induced by full-screen Cartesian gratings in all these subjects. We also examined steady-state visually evoked potentials (SSVEPs) at 32 Hz in a subset of these subjects (seven MCI and two AD subjects). We monitored eyes using an infrared eye tracker to rule out differences in gamma power due to potential differences in eye position or microsaccade rate (Yuval-Greenberg et al., 2008). Results We presented achromatic full-screen static sinusoidal grating stimuli that varied in spatial frequency (1, 2, and 4 cpd) and orientation (0°, 45°, 90°, and 135°; Figure 1A) to a large cohort of elderly subjects (227 healthy, 12 MCI, and 5 AD subjects; see 'Materials and methods' for details of subject selection and classification). We first examined whether gamma power depended on orientation and spatial frequency. We have previously shown that gamma oscillations in EEG recorded from healthy young subjects have low orientation selectivity (Murty et al., 2018). Consistent with this, we found that stimulus-induced change in fast gamma power did not vary significantly across different orientations (Figure 1—figure supplement 1; two-way ANOVA with spatial frequency and orientation as factors; F(3,2723) = 0.87, p=0.46) in healthy subjects, with very low orientation selectivity (mean ± SEM: 0.02 ± 0.002; see 'Materials and methods' for details). Slow gamma varied with orientation (F(3,2723) = 6.4, p=0.0003); however, orientation selectivity calculated across the four orientations was small (mean ± SEM: 0.03 ± 0.002; see Figure 1—figure supplement 1 for details). We therefore averaged all data across the four orientations. Figure 1 with 1 supplement see all Download asset Open asset Fixation task. Every trial started with the onset of a fixation spot (0.1°) at the center of the screen on which the subjects had to maintain fixation. After an initial blank period of 1000 ms (gray screen), 2–3 stimuli were randomly shown for 800 ms. These consisted of sinusoidal luminance gratings presented full screen at full contrast. Inter-stimulus interval (ISI) was 700 ms. Each stimulus (of a particular combination of spatial frequency, temporal frequency, and orientation) was presented for a total of ~30–40 times according to the subjects’ comfort and willingness, and is referred to as a ‘stimulus repeat’ in this paper unless otherwise stated. (A) Gamma experiment: static gratings (temporal frequency = 0 Hz) were presented at three spatial frequencies (SFs): 1, 2, and 4 cycles per degree (cpd) and four orientations: 0°, 45°, 90°, and 135°. This experiment lasted for ~25 min, with 1–2 short breaks (for 3–5 min) between blocks. (B) SSVEP experiment: gratings were randomly presented at a temporal frequency of 0 (static) or 16 cycles per second (cps). SF and orientation combination of gratings was fixed across the experiment. This was the combination that showed high change in slow and fast gamma power for each subject during preliminary analysis performed during the session. This experiment followed Gamma experiment during the same session and lasted for ~5 min completed in one block. As reported previously, gamma power varied across spatial frequencies as well (same two-way ANOVA as above; slow/fast gamma: F(2,2723) = 31.1/50.7, p=4.6×10−14/2.4×10−22). In particular, we found that the difference in gamma between cases and controls was more prominent at spatial frequencies of 2 and 4 cpd, so we used the data for these two spatial frequencies for main analysis. The 1 cpd condition, as well as the power averaged across all three spatial frequencies, gave similar, albeit slightly weaker, results (as shown in Figure 2—figure supplement 3). Because the sample sizes were severely unbalanced between cases and controls, for each case subject, we selected controls who were age- (±1 year) and gender-matched and averaged their spectral data. All results shown here are based on pairwise comparison between cases and their averaged controls. Non-pairwise comparison (e.g., between 12 MCI and their 74 age- and gender-matched controls) yielded similar results. Change in gamma power, but not alpha suppression, was reduced in case group compared to control group First, we examined how the two gamma rhythms differed in MCI subjects as compared to their healthy age- and gender-matched controls. We averaged spectral data for all analyzable bipolar electrodes (as described in Murty et al., 2020) from nine occipital and parieto-occipital pairs (marked in black enclosures in Figure 2D; see EEG data analysis subsection in 'Materials and methods') for each subject. We compared the change in power spectral densities (PSDs) for each MCI with the mean change in power of their corresponding age- and gender-matched controls. Figure 2A shows the median stimulus-induced change in PSDs for 12 MCIs (yellow) and their controls (dark orange; light shaded regions show ± SD of median after bootstrapping for 10,000 iterations). While both slow and fast gamma ‘bumps’ were conspicuously visible for MCI as well as control groups, power in both slow gamma (20–34 Hz) and fast gamma (36–66 Hz) ranges (but not alpha, 8–12 Hz range) was significantly lower in the MCI group compared to the control group (Kruskal-Wallis [K-W] test, significance as shown in Figure 2A). This could also be seen in the median time-frequency change in power spectrograms (baseline: −500–0 ms of stimulus onset) for cases and controls in Figure 2B. Change in band-limited power was significantly less for both gamma bands in the MCI group compared to the control group (Figure 2C; slow gamma: χ2(23) = 4.09, p=0.043; fast gamma: K-W test, χ2(23) = 5.61, p=0.018). However, alpha power was not significantly different (χ2(23) = 0.33, p=0.56). Results were similar when we combined both gamma bands to a single band (20–66 Hz; χ2(23) = 5.08, p=0.024) or used the ‘traditional’ gamma band (30–80 Hz; χ2(23) = 5.34, p=0.021). Figure 2—figure supplement 1 shows the time-frequency change in power spectra of individual MCI subjects and the mean of their controls, sorted by increasing gamma power in the MCI subjects. Although there was substantial variability across subjects (also observed in Figure 2C), only 3/12 MCIs showed higher slow gamma power and only 2/12 MCIs showed higher fast gamma than controls. Figure 2 with 4 supplements see all Download asset Open asset Alpha, slow, and fast gamma in mild cognitive impairments (MCIs) and controls. (A) Change in power spectral densities (PSD) for 12 MCI subjects and their respective controls. Solid traces indicate median PSD across 12 MCIs (yellow) and median of mean PSDs for 12 sets of healthy controls (dark orange). Shaded regions indicate ± SD from medians after bootstrapping over 10,000 iterations. Vertical lines represent alpha (8–12 Hz, violet), slow (20–34 Hz, pink), and fast gamma (36–66 Hz, orange). Colored bars at the bottom represent significance of differences in medians (K-W test, black: p=0.01–0.05, green: p<0.01; not corrected for multiple comparisons). (B) Median change in power spectrograms for MCIs (right) and controls (left). White horizontal lines represent alpha (dotted), slow gamma (solid), and fast gamma (dashed) bands. (C) Median change in power in alpha, slow gamma, and fast gamma bands for MCIs (yellow) and controls (dark orange). Data of individual MCI and mean for their respective controls are represented as gray circles. Error bars indicate ± SD from medians after bootstrapping over 10,000 iterations. Black asterisks represent significance of differences in medians (K-W test, p<0.05, not Bonferroni-corrected). (D) Average scalp maps of 112 bipolar electrodes (disks) for cases (bottom row) and controls (top row) for alpha (left), slow gamma (middle), and fast gamma (right). Color of disks represents change in power in respective frequency bands. Electrode groups used for calculation of band-limited power are enclosed in black. We note that we have used very stringent conditions for computation of gamma power, similar to our previous work (Murty et al., 2020; Murty et al., 2018). For example, for all subjects, we used the same set of electrodes over which gamma was computed, as well as same time and frequency ranges. Further, we computed the total power within a band by simply summing the absolute power values within the band separately in baseline and stimulus periods and then taking a ratio. This estimate has larger contribution from lower frequencies in the band because of the power-law distribution of PSDs in baseline/stimulus periods. Consequently, if the traces are overlapping at lower frequencies within the band and diverge at higher frequencies, which was the case in the slow gamma range, the total change in power in the band may not be significantly different. Therefore, these results could be further improved by customizing the low frequency limit of the gamma band for each subject, as well as choosing only electrodes that show stronger gamma. For example, taking slow and fast gamma ranges as 26–34 Hz and 44–56 Hz improved the p-values (slow gamma: χ2(23) = 7.69, p=0.005; fast gamma: χ2(23) = 8.34, p=0.004). Although we have refrained from such customization here because we wanted to study the efficacy of a simple and subject-independent computational procedure, such data-driven subject specific optimization holds promise for improving the efficacy of a gamma-based biomarker. Figure 2D shows the median scalp maps (see EEG data analysis subsection in 'Materials and methods') of change in band-limited power across 112 bipolar electrode pairs (shown as discs) for alpha, slow, and fast gamma bands. Stimulus-induced change in power across all three bands was most prominent in the nine electrode pairs as above. However, this change in power was less in the MCI group compared to the control group in slow and fast gamma bands (but not alpha band) as noted in Figure 2C. For two MCI subjects (M1 and M4 as shown in Figure 2—figure supplement 1), slow gamma power was less than 0 dB. This could be because visual stimuli typically suppress power at low frequencies (<30 Hz), and if the visual stimulus does not induce sufficiently strong slow-gamma rhythm, there is an overall reduction in power between 20 and 35 Hz. However, our results remained consistent even when these two MCI subjects were removed from analysis (χ2(19) = 4.81, p=0.028, for slow gamma between 26 and 34 Hz). Similarly, our results did not change when we removed one MCI subject (subject M5) who had negative fast gamma power (χ2(21) = 4.56, p=0.032). Our study had many control subjects for each case subject (range 4–19). To rule out the possibility that our results were influenced by this imbalance, we first ordered the control subjects based on the difference in experiment date from the case subject and then chose only the first N control subjects. We tried different values of N and found that our results remained consistent in all cases, with less slow and fast gamma power in cases vs controls. For example, Figure 2—figure supplement 2 shows the results for N = 1 (a single control for each case). Although the error bars shown in Figure 2C appear smaller for controls than cases, that is only because each data point for the control group is already an average across many subjects. When a single control subject was used per case (Figure 2—figure supplement 2), the error bars were comparable (standard error of the medians: 0.33, 0.20, and 0.34 for controls vs 0.28, 0.08, and 0.36 for cases, for slow gamma, fast gamma, and alpha, respectively). Similarly, if we pooled all the control subjects in one group without averaging (N = 74 controls vs 12 MCI subjects), the standard deviations of slow gamma, fast gamma, and alpha power were 0.94, 0.72, and 0.67 for controls and 0.79, 0.36, and 1.00 for cases. Therefore, the variability in power was comparable in the two groups. Figure 2—figure supplement 3 shows the results for individual spatial frequencies as well as after pooling all three spatial frequencies. MCI subjects had less slow and fast gamma than controls at all spatial frequencies, although the effect was stronger at spatial frequencies of 2 and 4 cpd. We also tested whether the event-related potentials (ERPs) varied across MCIs and healthy controls. Consistent with literature, we noticed three prominent peaks in the ERPs for these subjects: P1, N1, and P2 (Figure 2—figure supplement 4). The ERPs were not different between MCIs and controls (Figure 2—figure supplement 4, panel A). Specifically, we did not find any significant difference between P1/N1/P2 peak amplitude (Figure 2—figure supplement 4, panel B). These analyses indicate that the differences that we observed for MCIs and controls were limited only to slow and fast gamma power. Figure 3 shows the same results for the five AD subjects in our cohort. We interpret these results with caution, since the number of subjects is small (although the study would have benefitted from a larger sample size for both MCI and AD categories, it was not possible to increase the sample size due to the COVID-19 pandemic). Nonetheless, we observed a strong reduction in the slow and fast gamma bands (Figure 3A and B), which was significant for both slow gamma (χ2(9) = 4.81 p=0.028) and fast gamma (χ2(9) = 3.94, p=0.047), but not alpha (χ2(9) = 0.01, p=0.92). Data from individual AD subjects and their controls are shown in Figure 3—figure supplement 1. All five AD subjects had less slow gamma power, and only one AD subject (A5) had more fast gamma power relative to the controls. Further, similar to MCI subjects, AD subjects had less slow and fast gamma than controls at all spatial frequencies, although the effect was stronger at spatial frequencies of 2 and 4 cpd (Figure 3—figure supplement 2). Figure 3 with 3 supplements see all Download asset Open asset Alpha, slow, and fast gamma in AD subjects and controls. Same format as in Figure 2, but for five AD subjects and their respective controls. Because four out of five AD subjects had only mild AD (CDR = 1), we also tested the results after combining the MCI and AD data sets (Figure 3—figure supplement 3). While slow gamma (χ2(33) = 9.51, p=0.002) and fast gamma (χ2(33) = 8.88, p=0.003) power were strongly reduced in cases vs controls, there was no difference in alpha power (χ2(33) = 0.25, p=0.61) or frequencies higher than ~65 Hz. We further tested the dependence of power in gamma/alpha bands on CDR score using a linear regression model (see Regression analysis section in 'Materials and methods') while accounting for age and gender. In the matched condition in which data from all the age- and gender-matched control subjects for each case were averaged (yielding 17 cases and 17 controls), the coefficient for CDR was significantly negative for both gamma bands (βCDR = –0.57/–0.25, p=0.0017/0.0063 for slow/fast gamma). Results were similar for the unmatched condition, in which all the controls were considered separately (112 healthy, 12 MCI, and 6 AD subjects), albeit the CDR slope (βCDR) was significantly negative only for slow gamma (βCDR = –0.62/–0.29, p=0.0071/0.0702 for slow/fast gamma). On the other hand, similar to the results in previous analyses, alpha power did not depend on CDR (βCDR = –0.30/–0.30, p=0.14/0.21 for matched/unmatched conditions). We have previously shown that gamma power decreases with age in healthy elderly, and females have more gamma than males (Murty et al., 2020). Consistent with these results, we found that coefficient for age was significantly negative (βAGE = –0.018/–0.0094, p=0.014/0.06 for slow/fast gamma), while coefficient for gender was significantly positive (βGENDER = 0.33/0.35, p=0.0070/3.15 × 10−5 for slow/fast gamma) when the regression analysis was performed on the full set of healthy subjects (N = 227). However, when the regression analysis was performed with cases and controls as described above, the coefficients were not significant (matched: βAGE = 0.0042/4.13 × 10−5, p=0.73/0.99 and βGENDER = –0.036/0.11, p=0.87/0.36 for slow/fast gamma; unmatched: βAGE = –0.0071/–0.0009, p=0.51/0.91 and βGENDER = 0.17/0.19, p=0.30/0.10 for slow/fast gamma). This also suggests that CDR is a stronger predictor of gamma power than age or gender. These results suggest that the alternate hypothesis (gamma power in controls was greater than cases) was more likely than the null hypothesis (controls and cases had comparable gamma power). We quantified this by estimating the Bayes factor (BF), which is the ratio of the marginal likelihood of the alternate hypothesis and the null hypotheses, given the data that we observed (see 'Materials and methods for details). For the MCI group (N = 12), BF computed using single-tailed paired t-test was ~1.89 for both slow and fast gamma. However, as before, choosing a more ‘sensitive’ range for slow gamma (26–34 Hz) and fast gamma (44–56 Hz) improved the BF to 3.34 and 4.60, respectively, suggesting substantial evidence for the alternate hypothesis over the null hypothesis. For the AD group (N = 5), BF was 2.85 for slow gamma and 1.83 for fast gamma, suggesting weak evidence, which did not improve substantially when more sensitive ranges were used. However, when both MCI and AD groups were combined, BF increased to 13.70 for slow gamma (26–34 Hz) and 8.60 for fast gamma (44–56 Hz), further strengthening the evidence in favor of alternate hypothesis. On the other hand, evidence for alpha band power was in favor of the null hypothesis (BF was 0.088 when only MCIs were considered, 0.29 for ADs, and 0.077 when both MCI and AD subjects were considered). Difference in gamma power was not due to differences in eye position or movement Previous studies have correlated increases in gamma power with occurrence of small involuntary eye movements called microsaccades (Yuval-Greenberg et al., 2008). These have been described in previous literature using plots called ‘main sequence,’ which show peak velocity on ordinate and maximum velocity on abscissa on a log-log scale. These plots reveal the ballistic nature of microsaccades, that is, the initial velocity and maximum displacement of the eye in the visual field are correlated during microsaccadic movements (Engbert, 2006). We compared eye data between MCI/AD subjects and their respective controls and found comparable eye positions and microsaccade profiles (Figure 4A), similar main sequence (Figure 4B), and similar pupillary reactivity (Figure 4C) to stimulus presentation (measured as coefficient of variation of pupil diameter across time; see Murty et al., 2020). Further, the trends described in Figures 2 and 3 did not change qualitatively when we reanalyzed the data after removing stimulus repeats containing microsaccades (see 'Materials and methods') from analysis (Figure 4—figure supplement 1), although these did not reach significance due to lesser number of trials (~45% of original analysis) and fewer subjects compared to the original analysis (see figure legend for details). Similarly, the trends held true when we reanalyzed only those repeats that had at least one microsaccade (Figure 4—figure supplement 2). These results indicate that the trends described in Figure 2 are independent of the presence or absence of microsaccades. Figure 4 with 2 supplements see all Download asset Open asset Eye position, microsaccades, and pupillary reactivity for healthy/MCI/AD subjects. (A) Left column: Eye position in horizontal (top row) and vertical (middle row) directions; and histogram showing microsaccade rate (bottom row) vs time (−0.5–0.75 s of stimulus onset) for 11 MCI cases (yellow) and their respective healthy controls (dark orange). Solid traces indicate medians, shaded patches represent ± SD of median after bootstrapping over 10,000 samples. Right column: Same plots for four AD cases and their healthy controls. Eye position did not vary significantly across time between MCI/AD and control subjects except in the case of AD vs controls, where it varied slightly (but within ±0.1°). (B) Main sequence plots showing peak velocity and maximum displacement of all microsaccades (number indicated on top) extracted from 11 MCI (top row), 4 AD (bottom row) subjects indicated in yellow, and their corresponding healthy controls (dark orange). Average microsaccade rate (median ± SD of median of 10,000 bootstrapped samples) across all subjects for each group is also indicated at the bottom of the panels. MCI/AD cases had similar microsaccade rates (also seen in panel A) and main sequence plots compared to their healthy controls. (C) Bar plots showing coefficient of variation of pupil diameter (reactivity of pupil to stimulus presentation; see Murty et al., 2020) for 11 MCI (left), 4 AD (right), and their corresponding healthy controls. Data for individual MCIs and average across respective controls is represented by gray circles. Height of bars indicate medians and error bars indicate ± SD of median of 10,000 bootstrapped samples. We did not find any significant difference between the MCI/AD and control groups in pupil reactivity (K-W test, MCI vs controls: χ2(21) = 3.76, p=0.052; AD vs controls: χ2(7) = 0.75, p=0.39). Difference in gamma power was not due to differences in baseline power We also tested if the trends described in Figures 2 and 3 were seen for absolute band-limited power in the baseline condition. The cases (MCIs or ADs) and their respective controls had comparable PSDs (Figure 5) and slopes of PSDs (Figure 5—figure supplement 1) in the baseline condition. Further, baseline power in alpha, slow gamma, and fast gamma frequency ranges did not differ significantly between cases and controls (K-W test; for MCIs: χ2(23) = 0.48/0/0, p=0.49/0.95/1 for alpha/slow/fast gamma, respectively; for AD: χ2(9) = 0.27/0.53/0.27, p=0.60/0.46/0.60 for alpha/slow/fast gamma, respectively). Thus, we concluded that the trends described in Figures 2 and 3 were specific to stimulus-induced change in slow/fast gamma power and did not depend on baseline absolute power or slopes of PSDs. Figure 5 with 1 supplement see all Download asset Open asset Baseline PSDs and alpha/slow/fast gamma power in cases and healthy controls. Left column: Baseline PSDs (top row) and bar plots (bottom row) showing baseline absolute power (calculated in −500–0 ms of stimulus onset) for each of the 12 MCIs and corresponding healthy controls in alpha, slow gamma, and fast gamma bands. Same format as in Figure 2A and C. Data for individual MCIs and averages of corresponding control subjects are shown in gray circles. Corresponding analyses for five AD subjects are shown in right column. None of the differences in MCI and AD groups (compared to controls) were significant (see Results section). SSVEP power at 32 Hz was comparable in case and control groups We next tested whether power of SSVEPs in gamma range also decreased in the MCI group as compared to the control group. We tested for SSVEPs at 32 Hz by presenting full-screen grating" @default.
• W3175383339 created "2021-07-05" @default.
• W3175383339 creator A5004654673 @default.
• W3175383339 creator A5015291490 @default.
• W3175383339 creator A5020697014 @default.
• W3175383339 creator A5022523438 @default.
• W3175383339 creator A5046984070 @default.
• W3175383339 creator A5048452219 @default.
• W3175383339 creator A5050237688 @default.
• W3175383339 creator A5050538810 @default.
• W3175383339 creator A5058708625 @default.
• W3175383339 creator A5076079068 @default.
• W3175383339 creator A5087351819 @default.
• W3175383339 date "2021-02-13" @default.
• W3175383339 modified "2023-09-26" @default.
• W3175383339 title "Author response: Stimulus-induced gamma rhythms are weaker in human elderly with mild cognitive impairment and Alzheimer’s disease" @default.
• W3175383339 doi "https://doi.org/10.7554/elife.61666.sa2" @default.
• W3175383339 hasPublicationYear "2021" @default.
• W3175383339 type Work @default.
• W3175383339 sameAs 3175383339 @default.
• W3175383339 citedByCount "0" @default.
• W3175383339 crossrefType "peer-review" @default.
• W3175383339 hasAuthorship W3175383339A5004654673 @default.
• W3175383339 hasAuthorship W3175383339A5015291490 @default.
• W3175383339 hasAuthorship W3175383339A5020697014 @default.
• W3175383339 hasAuthorship W3175383339A5022523438 @default.
• W3175383339 hasAuthorship W3175383339A5046984070 @default.
• W3175383339 hasAuthorship W3175383339A5048452219 @default.
• W3175383339 hasAuthorship W3175383339A5050237688 @default.
• W3175383339 hasAuthorship W3175383339A5050538810 @default.
• W3175383339 hasAuthorship W3175383339A5058708625 @default.
• W3175383339 hasAuthorship W3175383339A5076079068 @default.
• W3175383339 hasAuthorship W3175383339A5087351819 @default.
• W3175383339 hasBestOaLocation W31753833391 @default.
• W3175383339 hasConcept C126322002 @default.
• W3175383339 hasConcept C135343436 @default.
• W3175383339 hasConcept C15744967 @default.
• W3175383339 hasConcept C169760540 @default.
• W3175383339 hasConcept C169900460 @default.
• W3175383339 hasConcept C180747234 @default.
• W3175383339 hasConcept C2779134260 @default.
• W3175383339 hasConcept C2779918689 @default.
• W3175383339 hasConcept C2984915365 @default.
• W3175383339 hasConcept C548259974 @default.
• W3175383339 hasConcept C71924100 @default.
• W3175383339 hasConceptScore W3175383339C126322002 @default.
• W3175383339 hasConceptScore W3175383339C135343436 @default.
• W3175383339 hasConceptScore W3175383339C15744967 @default.
• W3175383339 hasConceptScore W3175383339C169760540 @default.
• W3175383339 hasConceptScore W3175383339C169900460 @default.
• W3175383339 hasConceptScore W3175383339C180747234 @default.
• W3175383339 hasConceptScore W3175383339C2779134260 @default.
• W3175383339 hasConceptScore W3175383339C2779918689 @default.
• W3175383339 hasConceptScore W3175383339C2984915365 @default.
• W3175383339 hasConceptScore W3175383339C548259974 @default.
• W3175383339 hasConceptScore W3175383339C71924100 @default.
• W3175383339 hasLocation W31753833391 @default.
• W3175383339 hasOpenAccess W3175383339 @default.
• W3175383339 hasPrimaryLocation W31753833391 @default.
• W3175383339 hasRelatedWork W2000817815 @default.
• W3175383339 hasRelatedWork W2051831694 @default.
• W3175383339 hasRelatedWork W2061634617 @default.
• W3175383339 hasRelatedWork W2079468847 @default.
• W3175383339 hasRelatedWork W2180085651 @default.
• W3175383339 hasRelatedWork W2987653986 @default.
• W3175383339 hasRelatedWork W3033723657 @default.
• W3175383339 hasRelatedWork W3105559479 @default.
• W3175383339 hasRelatedWork W4296838050 @default.
• W3175383339 hasRelatedWork W902281470 @default.
• W3175383339 isParatext "false" @default.
• W3175383339 isRetracted "false" @default.
• W3175383339 magId "3175383339" @default.
• W3175383339 workType "peer-review" @default.