For a start, they don’t do what they say on the tin. It was named after the thing the inventors thought it would do, i.e. bind to and modulate GABA receptors, the ones which cause sedation and anxiolysis. But they have no activity at these receptors. Intuitively then they wouldn’t have an effect on sleep or anxiety.
They also don’t bind to dopamine receptors — you would think then that they wouldn’t be helpful for psychosis (most antipsychotics antagonise dopamine receptors).
And they don’t bind to opioid receptors, so they’re surely not useful for treating pain.
But they do! Gabapentinoids are prescribed for sleep, anxiety, bipolar disorder, and epilepsy, as well as neuropathic pain and restless legs syndrome.
Ok so what do they bind to then
Gabapentinoids bind to the α2δ protein, a subunit of voltage-gated calcium channels (hence their alternative name of α2δ ligands). Usually the concentration of calcium ions outside the cell is thousands of times higher than inside; these channels respond to a voltage by opening and allowing calcium to flood in. Depending on the cell they’re attached to, this can cause muscle contraction, neuronal signalling, and protein synthesis.
Specifically they bind α2δ-1 and α2δ-2, but only exert their effect through the former (as proven by trials on α2δ-2-knockout mice). There seems to be an as-yet undiscovered natural ligand for α2δ-1 and -2 which binds to the same site as gabapentinoids.
Importantly they don’t block calcium channels — instead they inhibit the release of monoamines (serotonin, norepinephrine, dopamine) and substance P1 triggered by calcium influx. They also inhibit calcium channel-dependent release of glutamate and glycine in various brain tissues.
Sensitized calcium channels
There are states in which calcium channels become ‘sensitized’, such as in the case of neuronal injury, and gabapentinoids might selectively work in these conditions.
Activation of protein kinase C is required for gabapentinoids to reduce substance P released caused by capsaicin
Gabapentinoids reduce the size of postsynaptic currents in certain tissues in hyperalgesic rats (which have been bred to feel more pain), but not in normal rats
Glutamate release triggered by substance P is blocked by gabapentinoids
As they don’t simply block calcium channels, they have big advantages over drugs that do — they only minimally change synaptic function, unlike calcium channel blockers. They can essentially restore ‘normal’ functioning in overexcited calcium channels while leaving healthy ones alone.
Natural gabapentinoids in the body
Anticlockwise from top: gabapentin, leucine, isoleucine
Gabapentinoids have a suspicious structural similarity to leucine and isoleucine, two amino acids. Radiolabelling these amino acids shows they also bind the α2δ protein, and L-isoleucine blocks certain effects of gabapentinoids, suggesting they compete for binding at the same site.
Some people have reported relief of their restless legs syndrome from acetylleucine, a leucine analog, which suggests it’s acting in a similar way to gabapentinoids (Fields 2021). Curiously this drug is very hard to find except in France, where it’s sold over-the-counter.
Gabapentin vs pregabalin
Unlike lots of drugs, gabapentinoids seem to be actively transported into the body by LAT1, the large neutral amino acid transporter.
This is a disadvantage over other drugs, because it limits how much and how quickly gabapentin can be absorbed. Gabapentin often has to be taken multiple times per day to avoid saturating these transporters. It also competes with other amino acids (the ones above) for these transporters.
Pregabalin, another gabapentinoid, is superior here because it is transported by other carriers, not just LAT1, so its uptake doesn’t saturate in the same way. It binds α2δ much more strongly than gabapentin, and in animals is more potent as an analgesic and anticonvulsant.
Can they block synapse formation?
Even weirder: Eroglu 2009 found that α2δ-1 is a neuronal receptor for thrombospondin, a molecule which promotes synaptogenesis in astrocytes. Specifically, it forms part of a larger signalling system. It acts as the extracellular receptor for a “synaptogenic signalling complex”; when thrombospondin binds, it causes a cascade of events which switches on this complex and leads to the start of synapse development.
As gabapentinoids also bind to this protein… does that mean they reduce synaptogenesis? In vitro, yes: gabapentin powerfully blocks synapse formation. Though this sounds slightly terrifying it’s also probably an important mechanism for gabapentinoids’ effects in epilepsy and neuropathic pain — synapse formation can be triggered by neuronal injury in these conditions and might well contribute to the pathology of these conditions (although this is uncertain).
It’s worth noting that gabapentin and thrombospondin, while both binding to the same protein, don’t bind to the same part of that protein.
(It’s kind of nuts that it took decades for one of the key mechanisms of action for this class of drug to be discovered. Makes you wonder what else we don’t know, about gabapentinoids and other drugs.)
Memory, executive function, and dementia
Worryingly this suggests that gabapentinoids might affect the normal formation of synapses. Could this cause other deficits, such as in memory formation?
Behroozi 2023 attempted to test this and did not find an effect, although they were looking specifically at improvements in memory formation.
Gabapentinoids can certainly cause brain fog and slower processing. Eghrari 2025 also found an increased risk of cognitive impairment and dementia in patients with chronic low back pain prescribed gabapentin; when stratified by age, patients taking gabapentin had twice the risk of dementia and mild cognitive impairment. This risk was further increased in patients who had taken gabapentin more throughout their lives. Presumably this effect would also extend to pregabalin.
A billon-dollar scandal
Gabapentinoids are frequently prescribed off label (when a doctor prescribes a drug outside of the conditions for which it’s approved). Not necessarily a bad thing: doctors use their discretion to decide when to do this, and for a drug with as broad a therapeutic profile as gabapentinoids it doesn’t seem wholly surprising.
But there are strict rules around advertising a drug for this sort of thing, or pushing doctors to prescribe it off label. The drug is approved for specific conditions and drug companies (in countries where they’re allowed to advertise) can only push for it to be prescribed for these conditions.
Their subsidiary Parke-Davis promoted Neurontin (gabapentin) for at least eleven unapproved conditions, flying doctors to lavish retreats, paying kickbacks, and commissioning ghostwritten journal articles. Off-label prescribing accounted for 78% of Neurontin sales.
Pfizer pleaded guilty to criminal charges and paid $945 million in settlements. In a separate 2009 case, they paid a further $2.3 billion for off-label marketing of several drugs including Lyrica (pregabalin).
Separately, top pain researcher Scott Reuben admitted to fabricating data in at least 21 studies – including Pfizer-funded trials of Lyrica – without ever enrolling a single patient. He was jailed in 2010.
Can they make you suicidal?
More controversial. One epidemiological survey looked at a cohort of individuals before and after they were prescribed gabapentin, and found no increase in suicidality, as well as a reduction in suicide attempts in psychiatric patients (Gibbons 2011). A large Swedish cohort study found a significant increase in suicide – but only for pregabalin, and not gabapentin (Molero 2019).
It’s not clear why this would be the case, as the drugs work in exactly the same way (as far as we know). In fact, pregabalin was found to increase suicidal behaviour/deaths from suicide, unintentional overdoses, head and body injuries, road traffic accidents and offences, and arrests for violent crime, where gabapentin had no or almost no effect (and actually reduced road traffic incidents and arrests).
The obvious explanation is that pregabalin is simply more powerful, both due to the pharmacokinetic gap described above and because it binds α2δ much more strongly. The highest doses of gabapentin simply can’t compete with the highest doses of pregabalin.
Are they fun?
Certainly for some people they are. Gabapentinoids are notorious for diversion, where people score prescriptions and then sell the drugs on. The prescription rate for these drugs in prisons is double that of the general population. In some ways pregabalin is the drug of choice in UK prisons. In France, 81% of recreational teenage pregabalin users reported to poison control centres were homeless or living in migrant shelters (Dufayet 2021).
This should surprise us; they don’t have any dopaminergic or opioidergic activity, so they don’t tick the obvious addictive drug boxes. Nonetheless, some people clearly find them enjoyable, with effects somewhat similar to alcohol/benzodiazepines, and develop dependence on them.
This might explain why there are more than 50 million gabapentinoid prescriptions issued every year in the US alone.
In the hilarious Drug User’s Bible, in which the author takes basically every drug imaginable, there’s this snippet from taking 300mg pregabalin (which is a hefty dose, users are started on 75mg typically):
I totally underestimated this drug. I am basically zombified and largely mistuned to what is going on around me, which appears to be distant. My hands are numb and I am, essentially, stupefied, with head spinning.
This one was a shock. I clearly took far too much and paid a price in terms of a strong intoxication which at times was extremely uncomfortable.
Conclusion
Gabapentinoids are weirder than I had realised.
Of course, the conditions they are prescribed for are horrific – anxiety, chronic pain etc can be a living hell, and a drug which effectively treats them is miraculous. But it’s wild that it took us decades to actually understand the first thing about how these drugs work.
And the effects on synaptogenesis, unknown effects on memory, increased risk of various kinds of death and dangerous behaviour (in the case of pregabalin), huge abuse in prisons and migrant shelters, and increased risk of cognitive deficits and dementia should probably worry us given how widely they’re prescribed.
(with apologies to Sasha Shulgin)
Gabapentinoids are weird.
For a start, they don’t do what they say on the tin. It was named after the thing the inventors thought it would do, i.e. bind to and modulate GABA receptors, the ones which cause sedation and anxiolysis. But they have no activity at these receptors. Intuitively then they wouldn’t have an effect on sleep or anxiety.
They also don’t bind to dopamine receptors — you would think then that they wouldn’t be helpful for psychosis (most antipsychotics antagonise dopamine receptors).
And they don’t bind to opioid receptors, so they’re surely not useful for treating pain.
But they do! Gabapentinoids are prescribed for sleep, anxiety, bipolar disorder, and epilepsy, as well as neuropathic pain and restless legs syndrome.
Ok so what do they bind to then
Gabapentinoids bind to the α2δ protein, a subunit of voltage-gated calcium channels (hence their alternative name of α2δ ligands). Usually the concentration of calcium ions outside the cell is thousands of times higher than inside; these channels respond to a voltage by opening and allowing calcium to flood in. Depending on the cell they’re attached to, this can cause muscle contraction, neuronal signalling, and protein synthesis.
Specifically they bind α2δ-1 and α2δ-2, but only exert their effect through the former (as proven by trials on α2δ-2-knockout mice). There seems to be an as-yet undiscovered natural ligand for α2δ-1 and -2 which binds to the same site as gabapentinoids.
Importantly they don’t block calcium channels — instead they inhibit the release of monoamines (serotonin, norepinephrine, dopamine) and substance P1 triggered by calcium influx. They also inhibit calcium channel-dependent release of glutamate and glycine in various brain tissues.
Sensitized calcium channels
There are states in which calcium channels become ‘sensitized’, such as in the case of neuronal injury, and gabapentinoids might selectively work in these conditions.
As they don’t simply block calcium channels, they have big advantages over drugs that do — they only minimally change synaptic function, unlike calcium channel blockers. They can essentially restore ‘normal’ functioning in overexcited calcium channels while leaving healthy ones alone.
Natural gabapentinoids in the body
Anticlockwise from top: gabapentin, leucine, isoleucine
Gabapentinoids have a suspicious structural similarity to leucine and isoleucine, two amino acids. Radiolabelling these amino acids shows they also bind the α2δ protein, and L-isoleucine blocks certain effects of gabapentinoids, suggesting they compete for binding at the same site.
Some people have reported relief of their restless legs syndrome from acetylleucine, a leucine analog, which suggests it’s acting in a similar way to gabapentinoids (Fields 2021). Curiously this drug is very hard to find except in France, where it’s sold over-the-counter.
Gabapentin vs pregabalin
Unlike lots of drugs, gabapentinoids seem to be actively transported into the body by LAT1, the large neutral amino acid transporter.
This is a disadvantage over other drugs, because it limits how much and how quickly gabapentin can be absorbed. Gabapentin often has to be taken multiple times per day to avoid saturating these transporters. It also competes with other amino acids (the ones above) for these transporters.
Pregabalin, another gabapentinoid, is superior here because it is transported by other carriers, not just LAT1, so its uptake doesn’t saturate in the same way. It binds α2δ much more strongly than gabapentin, and in animals is more potent as an analgesic and anticonvulsant.
Can they block synapse formation?
Even weirder: Eroglu 2009 found that α2δ-1 is a neuronal receptor for thrombospondin, a molecule which promotes synaptogenesis in astrocytes. Specifically, it forms part of a larger signalling system. It acts as the extracellular receptor for a “synaptogenic signalling complex”; when thrombospondin binds, it causes a cascade of events which switches on this complex and leads to the start of synapse development.
As gabapentinoids also bind to this protein… does that mean they reduce synaptogenesis? In vitro, yes: gabapentin powerfully blocks synapse formation. Though this sounds slightly terrifying it’s also probably an important mechanism for gabapentinoids’ effects in epilepsy and neuropathic pain — synapse formation can be triggered by neuronal injury in these conditions and might well contribute to the pathology of these conditions (although this is uncertain).
It’s worth noting that gabapentin and thrombospondin, while both binding to the same protein, don’t bind to the same part of that protein.
(It’s kind of nuts that it took decades for one of the key mechanisms of action for this class of drug to be discovered. Makes you wonder what else we don’t know, about gabapentinoids and other drugs.)
Memory, executive function, and dementia
Worryingly this suggests that gabapentinoids might affect the normal formation of synapses. Could this cause other deficits, such as in memory formation?
Behroozi 2023 attempted to test this and did not find an effect, although they were looking specifically at improvements in memory formation.
Gabapentinoids can certainly cause brain fog and slower processing. Eghrari 2025 also found an increased risk of cognitive impairment and dementia in patients with chronic low back pain prescribed gabapentin; when stratified by age, patients taking gabapentin had twice the risk of dementia and mild cognitive impairment. This risk was further increased in patients who had taken gabapentin more throughout their lives. Presumably this effect would also extend to pregabalin.
A billon-dollar scandal
Gabapentinoids are frequently prescribed off label (when a doctor prescribes a drug outside of the conditions for which it’s approved). Not necessarily a bad thing: doctors use their discretion to decide when to do this, and for a drug with as broad a therapeutic profile as gabapentinoids it doesn’t seem wholly surprising.
But there are strict rules around advertising a drug for this sort of thing, or pushing doctors to prescribe it off label. The drug is approved for specific conditions and drug companies (in countries where they’re allowed to advertise) can only push for it to be prescribed for these conditions.
Their subsidiary Parke-Davis promoted Neurontin (gabapentin) for at least eleven unapproved conditions, flying doctors to lavish retreats, paying kickbacks, and commissioning ghostwritten journal articles. Off-label prescribing accounted for 78% of Neurontin sales.
Pfizer pleaded guilty to criminal charges and paid $945 million in settlements. In a separate 2009 case, they paid a further $2.3 billion for off-label marketing of several drugs including Lyrica (pregabalin).
Separately, top pain researcher Scott Reuben admitted to fabricating data in at least 21 studies – including Pfizer-funded trials of Lyrica – without ever enrolling a single patient. He was jailed in 2010.
Can they make you suicidal?
More controversial. One epidemiological survey looked at a cohort of individuals before and after they were prescribed gabapentin, and found no increase in suicidality, as well as a reduction in suicide attempts in psychiatric patients (Gibbons 2011). A large Swedish cohort study found a significant increase in suicide – but only for pregabalin, and not gabapentin (Molero 2019).
It’s not clear why this would be the case, as the drugs work in exactly the same way (as far as we know). In fact, pregabalin was found to increase suicidal behaviour/deaths from suicide, unintentional overdoses, head and body injuries, road traffic accidents and offences, and arrests for violent crime, where gabapentin had no or almost no effect (and actually reduced road traffic incidents and arrests).
The obvious explanation is that pregabalin is simply more powerful, both due to the pharmacokinetic gap described above and because it binds α2δ much more strongly. The highest doses of gabapentin simply can’t compete with the highest doses of pregabalin.
Are they fun?
Certainly for some people they are. Gabapentinoids are notorious for diversion, where people score prescriptions and then sell the drugs on. The prescription rate for these drugs in prisons is double that of the general population. In some ways pregabalin is the drug of choice in UK prisons. In France, 81% of recreational teenage pregabalin users reported to poison control centres were homeless or living in migrant shelters (Dufayet 2021).
This should surprise us; they don’t have any dopaminergic or opioidergic activity, so they don’t tick the obvious addictive drug boxes. Nonetheless, some people clearly find them enjoyable, with effects somewhat similar to alcohol/benzodiazepines, and develop dependence on them.
This might explain why there are more than 50 million gabapentinoid prescriptions issued every year in the US alone.
In the hilarious Drug User’s Bible, in which the author takes basically every drug imaginable, there’s this snippet from taking 300mg pregabalin (which is a hefty dose, users are started on 75mg typically):
Conclusion
Gabapentinoids are weirder than I had realised.
Of course, the conditions they are prescribed for are horrific – anxiety, chronic pain etc can be a living hell, and a drug which effectively treats them is miraculous. But it’s wild that it took us decades to actually understand the first thing about how these drugs work.
And the effects on synaptogenesis, unknown effects on memory, increased risk of various kinds of death and dangerous behaviour (in the case of pregabalin), huge abuse in prisons and migrant shelters, and increased risk of cognitive deficits and dementia should probably worry us given how widely they’re prescribed.
References