Joe told the First Lady he didn't feel well.

[andym]

37. Older people tend to have peak cognition in the morning

How that impacts someone who is traditionally a night person like yourself is uncertain. However, it suggests that Joe Biden would do better with an afternoon debate. Perhaps he should go on a two-week Hawaiian vacation before the next debate, so early evening will seem like early afternoon to him.

Evidence:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4067093/#:~:text=Older%20adults'%20peak%20performance%20on,to%20be%20in%20the%20afternoon.
Look Out – It’s Your Off-Peak Time of Day! Time of Day Matters More for Alerting than for Orienting or Executive Attention
Marisa Knight and Mara Mather
Exp Aging Res. 2013; 39(3): 305–321.

Abstract
Background
Older adults’ peak performance on memory and cognitive inhibition tasks tends to be in the morning while younger adults’ peak performance tends to be in the afternoon. Although these tasks require efficient attentional processes for optimal performance, previous research examining age differences in the effects of time of day has not measured the distinct aspects of attention quantified by the attentional network test (ANT; Fan, McCandliss, Sommer, & Posner, 2002).

Methods
We examined the relationship between time of testing and the efficiency of alerting, orienting and executive attention networks by randomly assigning younger (18–28 years; N = 27, M = 21.37 years, SD = 2.39) and older adults (65–85 years; N = 32, M = 73.34 years, SD = 5.18) to AM or PM testing of alerting, orienting and executive attention. Mean reaction times for each network was analyzed with a univariate ANOVA with age (younger, older) and time of day (AM, PM) as between-subjects factors.

Results
Consistent with our hypotheses, while time of day had little effect on orienting or executive attention, it affected alerting in opposite ways for younger and older adults, with alerting cues benefitting performance most at participants’ off-peak times of day. A larger benefit from alerting cues was observed when participants’ were tested at their off-peak (M = 30.11 ± 15.66) relative to their peak time (M = 2.18 ± 15.97).

Conclusion
Our findings show that age-related circadian patterns influence the alerting component of attention, with both age groups showing the largest benefit from alerting cues when testing occurs at non-optimal times of day. Overall, our findings underscore the importance of controlling for time of day in investigations of attention and add to our understanding of how age differences in circadian patterns impact attention.