The following study has received a lot of press, generating excitement for neuromodulation as a new potential treatment for dementia, including Alzheimer’s disease. Briefly, adults aged 65 years or older were recruited for a series of three experiments. Participants were randomized to neuromodulation or sham and repeatedly evaluated in terms of memory performance.
Grover, S., Wen, W., Viswanathan, V. et al. Long-lasting, dissociable improvements in working memory and long-term memory in older adults with repetitive neuromodulation. Nat Neurosci (2022). Long-lasting, dissociable improvements in working memory and long-term memory in older adults with repetitive neuromodulation | Nature Neuroscience
The authors reported that “The rate of memory improvements over 4 days predicted the size of memory benefits 1 month later. Individuals with lower baseline cognitive function experienced larger, more enduring memory improvements.” As far as I can tell these observations are based on Figures 4 and 5, which do not include the sham group, and I suspected regression-to-the mean with respect to these specific statements. However, after zooming in on the tiny spaghetti plots in Figure 3, which do include the sham group, it appears there could be a real effect of neuromodulation on memory.
After reading some of Frank Harrell’s work I suspect there are better ways to do the analysis.
The authors posted the data here (OSF | Repetitive theta/gamma neuromodulation in older adults selectively improves dual auditory-verbal memory for one month.), so if anyone has the time and interest, they could do an alternative analysis.