Many long-COVID cognitive deficits improve spontaneously; without a control group, you cannot separate drug effect from natural history.. explain

The short version: you are right. Without a control group, you cannot tell whether a drug helped or time healed.

That is the central problem with the guanfacine-plus-NAC case series you asked about last time. The evidence you supplied for this follow-up makes the point concrete. A 2025 meta-analysis of 50 controlled studies found that infected people had roughly 2.5 to 4 times the risk of cognitive symptoms like poor concentration and impaired memory compared to uninfected controls [12]. But the same meta-analysis also shows that many of those symptoms resolve over time in a substantial fraction of people.

A 2024 case series of severe COVID ICU survivors found persistent cognitive deficits at 12 months, but the pattern was variable across patients [1]. A 2023 case series of older adults with pre-existing cognitive impairment showed three different trajectories after COVID: one continued to decline, one began to improve, and one stayed stable [2]. A 2022 case report of a young woman with post-COVID POTS described gradual improvement over a year with only lifestyle changes and exercise [6]. Natural recovery happens, and it happens on different schedules for different people.

The 2026 neuroanatomical study is the most telling. It scanned long-COVID patients at baseline and six months later. Gray matter density in the medial frontal cortex increased significantly over those six months [20]. That is the brain healing on its own. The same study found that cognitive differences between patient subgroups and healthy controls had medium-to-large effect sizes but did not survive statistical correction for multiple comparisons [20]. The signal was there, but the sample was too small to be sure.

Now apply this to the guanfacine case series. The patients had been sick for 1.9 to 2.7 years [18]. That is long enough that spontaneous recovery is less likely, but it is not zero. The study had no control group, so any improvement could be natural history, placebo effect, or drug effect. You cannot separate them. The only way to know is a randomized placebo-controlled trial with objective cognitive testing. That trial has not been done.

My call: the guanfacine protocol is biologically plausible but unproven. The evidence does not rule out that natural recovery explains the improvements. Confidence: low.