I believe there is a more general point to be made here. Namely, the effects of most (if not all) psychoactive substances don’t decay with their half-life. A more useful framing would be that they don’t track blood concentration all that well, and psychoactive effects can last both significantly longer or shorter than looking at the half-life would predict. I don’t do theoretical analyses such as yours, because I don’t expect getting much insight that is not available through (other people’s) empiricism.[1] Consider a few cases: (half-life numbers found through quick googling) Caffeine - [2] Looking at your applet, it suggests a ~fourth of (effective, after including paraxantine) the peak concentration after 16 hours. But at the same time, I would predict that taking a bolus caffeine dose after waking up would cause less sleep disturbance than taking a paraxantine dose in the evening to produce the same (effective) blood concentration. I presume you would expect this as well? Phenibut—half life 5.3 hours. Psychonaut wiki lists the duration of effects as 10-16 hours which basically matches my experience. By any objective/subjective metric I’d predict we’d find the effects to dimish more slowly than with caffeine. Armodafinil—half life 15 hours. Duration 8-15 hours, which, again, matches my experience. Note, the duration is a bit shorter than with phenibut despite the huge half-life difference. Nicotine Nicotinehalf life 1-2 hours. Duration (from buccal administration [3]which I use) 0.5-1.5 hours. Psychedelics (in general) - I believe psychedelic trips are largely inexplicable through any exponential decay model (of psychoactive effects). I can explain, if you don’t share this intuition.
I could go on, of course. But my overall point is that regardless of whether one uses receptor affinity and occupancy and dowstream metabolites or more/other variables in one’s model, said model will likely still not be accurate enough to predict subjective effects.
I understand the general point. Here I wanted to point out a pretty narrow misconception about caffeine specifically. Of course psychological state isn’t purely a function of blood concentration since there are other variables that change throughout the day, like adenosine levels. It’s also possible that there could be a lag time or complicated mapping from receptor occupancy to psychological effects. But there’s a useful intermediate node in the causal graph, effective concentration of receptor antagonists, that looks pretty different depending on whether you model paraxanthine or not.
I would predict that taking a bolus caffeine dose after waking up would cause less sleep disturbance than taking a paraxantine dose in the evening to produce the same (effective) blood concentration. I presume you would expect this as well?
I don’t think I would expect that, degree of tolerance held constant.
On further thought I would weakly expect more difficulty getting to sleep from paraxanthine if you took it within an hour of going to bed, since there could be some stuff that has to do with the derivative of effective concentration. If you took it several hours before bed to end up with the same effective concentration at bedtime I don’t think I expect a noticeable difference in ability to get to sleep. I also think I’d expect less sleep disturbance overall from paraxanthine, effective concentration at bedtime held constant and barring scenarios where you’re taking it within an hour of bedtime, since the effective concentration resulting from paraxanthine will decay faster than that resulting from caffeine.
I believe there is a more general point to be made here. Namely, the effects of most (if not all) psychoactive substances don’t decay with their half-life. A more useful framing would be that they don’t track blood concentration all that well, and psychoactive effects can last both significantly longer or shorter than looking at the half-life would predict. I don’t do theoretical analyses such as yours, because I don’t expect getting much insight that is not available through (other people’s) empiricism.[1]
Consider a few cases: (half-life numbers found through quick googling)
Caffeine - [2] Looking at your applet, it suggests a ~fourth of (effective, after including paraxantine) the peak concentration after 16 hours. But at the same time, I would predict that taking a bolus caffeine dose after waking up would cause less sleep disturbance than taking a paraxantine dose in the evening to produce the same (effective) blood concentration. I presume you would expect this as well?
Phenibut—half life 5.3 hours. Psychonaut wiki lists the duration of effects as 10-16 hours which basically matches my experience. By any objective/subjective metric I’d predict we’d find the effects to dimish more slowly than with caffeine.
Armodafinil—half life 15 hours. Duration 8-15 hours, which, again, matches my experience. Note, the duration is a bit shorter than with phenibut despite the huge half-life difference.
Nicotine
Nicotine half life 1-2 hours. Duration (from buccal administration [3]which I use) 0.5-1.5 hours.
Psychedelics (in general) - I believe psychedelic trips are largely inexplicable through any exponential decay model (of psychoactive effects). I can explain, if you don’t share this intuition.
I could go on, of course. But my overall point is that regardless of whether one uses receptor affinity and occupancy and dowstream metabolites or more/other variables in one’s model, said model will likely still not be accurate enough to predict subjective effects.
Though I heard about Paraxantine before
Psychonaut wiki lists duration as 2-5 hours which is proably a bit too short, but not by much.
Administration can also alter the duration of effects greatly. Way more than looking at plasma concentrations would predict.
Presumably, there must be research on this?
I understand the general point. Here I wanted to point out a pretty narrow misconception about caffeine specifically. Of course psychological state isn’t purely a function of blood concentration since there are other variables that change throughout the day, like adenosine levels. It’s also possible that there could be a lag time or complicated mapping from receptor occupancy to psychological effects. But there’s a useful intermediate node in the causal graph, effective concentration of receptor antagonists, that looks pretty different depending on whether you model paraxanthine or not.
I don’t think I would expect that, degree of tolerance held constant.
On further thought I would weakly expect more difficulty getting to sleep from paraxanthine if you took it within an hour of going to bed, since there could be some stuff that has to do with the derivative of effective concentration. If you took it several hours before bed to end up with the same effective concentration at bedtime I don’t think I expect a noticeable difference in ability to get to sleep. I also think I’d expect less sleep disturbance overall from paraxanthine, effective concentration at bedtime held constant and barring scenarios where you’re taking it within an hour of bedtime, since the effective concentration resulting from paraxanthine will decay faster than that resulting from caffeine.