Cannabis has been utilized for millennia across medicinal, spiritual, and recreational contexts. In contemporary discourse, however, it is frequently associated with adverse health outcomes, particularly those involving neural function. Cannabis-related alterations in mood, cognition, and perception can lead to elevated risks for psychiatric conditions, including depression, anxiety, suicidal ideation, and psychosis^52,53. During recent years, legislative reforms and increasing recognition of the therapeutic potential of cannabis have contributed to a marked rise in its use among older adults. Unlike other populations, the effects of cannabis in this population may be complicated by age-related factors, especially for the changes in brain morphology and function found to be affected by cannabis use in younger populations^6,54. This study builds on prior work documenting substantial variability in the relationship between cannabis use and the functioning brain⁵⁵. Despite growing interest, empirical investigations examining the impact of cannabis on the aging brain remain scarce and often yield inconsistent findings. To address this gap, in this study, we capitalized on advances in a fully automated framework to achieve an individual-level estimation of whole brain FC in a purely data-driven manner and comprehensively characterize how they are associated with cannabis use via a widely used LMM approach^36,45,56,57.

To date, the majority of human neuroimaging studies investigating the effects of cannabis on brain function have focused on adolescents and young adults^12,23. In contrast, research examining cannabis use in older adults remains limited, and existing findings are marked by considerable variability and inconsistency. For instance, a study examining adults aged 19 to 55 years reported that cannabis users exhibited reduced FC between the amygdala and cuneus, as well as between the cingulum and occipital and temporal cortices, with no significant group differences observed in hippocampal connectivity⁵⁸. In another investigation, cannabis users demonstrated increased FC within both the salience network and the default mode network (DMN), alongside anticorrelated FC between these two networks⁵⁹. However, these findings diverge markedly from those of a study specifically targeting older adults (ages 60–88), which reported increased FC between the cerebellum and hippocampus in cannabis users—suggesting a potential neuroprotective effect in aging²³. Such discrepancies across studies may stem, in part, from limited sample sizes, which constrain statistical power and increase the risk of inflated or unreliable brain-behavior associations. Given the need for larger, well-powered datasets to establish robust neurobiological correlates of cannabis use, the UKB offers a valuable resource for systematically investigating the effects of cannabis on the aging brain in a more reliable and generalizable manner.

The observed cannabis-related enhancement of FNC within cortical-subcortical-cerebellar circuits aligns with anatomical evidence indicating high concentrations of cannabinoid receptors in these regions^60,61. The endocannabinoid system (ECS), a complex cell-signaling network, plays a critical role in regulating numerous physiological processes⁶² and undergoes age-dependent changes that may modulate the body’s response to cannabinoids such as delta-9-tetrahydrocannabinol (THC) and cannabidiol (CBD)^63,64. The ECS is intricately involved in aging, with cannabinoid receptors regulating cellular mechanisms underlying age-associated inflammation. Notably, older adults exhibit reductions in both cannabinoid receptor density and endogenous endocannabinoid levels, which may contribute to the cognitive decline commonly observed with aging⁶⁵. Preclinical studies have demonstrated age-specific interactions between cannabinoids and the ECS; for instance, administration of low-dose THC in aged animals would improve spatial memory, enhance synaptic density, and reduce neuroinflammation—potentially through reactivation of hypoactive cannabinoid receptor circuits⁶⁶. THC-induced activation of microglia within the receptors-expressed brain regions, along with subsequent neuroinflammatory responses, has been implicated as a potential mechanism underlying impairments in classical conditioning observed in animal models⁶⁷. It is also believed that the activation of cannabinoid receptors restores FC in networks particularly susceptible to age-related declines, such as the cerebellum-subcortical-cortical pathways^68,69. Consequently, our findings of increased FC in brain regions with high cannabinoid receptor expression may reflect the modulatory influence of cannabinoids on ECS activity, potentially facilitating neural plasticity and inter-regional communication, thereby enhancing cognitive function and mitigating neuroinflammation^70,71.

Traditionally regarded as a structure primarily responsible for motor coordination, the cerebellum is now increasingly recognized for its integral role in cognitive processing. Emerging evidence suggests that cognitive aging is associated with widespread cerebellar dysrhythmias^72,73, highlighting the cerebellum’s susceptibility to age-related functional decline^74–76. Cerebellar circuits are significantly implicated in age-related changes in brain function^77,78, reflecting the cerebellum’s extensive connectivity across the whole brain through closed-loop networks that support integrative motor and cognitive functions⁷⁹. Resting-state studies have further demonstrated that genetically and epigenetically regulated networks within the neocerebellum contribute to cognitive function in aging populations^80,81. Consistent with previous hypotheses and empirical data, the present findings suggest that alterations in cerebellar connectivity may serve as a key neural substrate linking normative aging to cognitive decline. Aging is commonly associated with structural deterioration within the cerebellum, including significant volumetric reductions that may impair intra-cerebellar information processing and integration^76,82. This impairment is reflected in decreased cerebellar connectivity, which may underline observed deficits in cognitive performance. Intriguingly, the present findings suggest that cannabis use may confer a protective effect against such age-associated reductions in cerebellar connectivity. Specifically, cannabis use was associated with increased FC within cerebellar circuits, particularly in regions previously implicated in higher-order motor control and cognitive functioning in older adults. These findings raise the possibility that cannabis may modulate cerebellar network integrity, potentially contributing to the preservation of functional performance in aging populations.

Previous research has proposed that age-related cognitive decline may, in part, stem from alterations in the structure of whole-brain FC^83,84. Healthy aging may indirectly impact cognitive performance through age-related modifications in FC patterns. In the current study, we observed that, except for the DMN, most within-domain FNC exhibited a declining trend with age. This suggests a reduction in modular organization across the brain and supports the theory that cognitive aging is characterized by decreased network segregation^85,86. Such segregation refers to the degree to which specific brain regions maintain specialized functional roles, and its reduction is considered a hallmark of age-related functional decline⁸⁷. More importantly, our findings further reveal that cannabis use may exert opposing effects on the FC structure compared to normative aging. Specifically, cannabis use in older adults was associated with increased within-domain FC in subcortical, cerebellar, and sensorimotor networks, along with decreased between-domain FC involving these same regions. This pattern indicates enhanced network segregation, which may support more specialized processing within functional domains while preserving selective inter-network communication—an organization thought to promote efficient cognitive performance⁸⁸. Interestingly, prior research has shown that chronic cannabis use in younger individuals is associated with reduced network segregation, often manifesting as diminished cognitive control, increased mind-wandering, and impaired task-switching⁸⁹. Taken together with our findings, this suggests that cannabis may have age-dependent effects on brain functional organization, potentially due to differences in neurodevelopment, metabolism, immune response, and neurodegenerative processes. We speculate that cannabinoids and endocannabinoids may exert neuroprotective effects during aging by preserving an optimal balance between functional segregation and integration—an essential feature for maintaining specialized processing and efficient information transfer across brain networks. The enhanced cognitive functioning observed in older cannabis users may be partially attributed to cannabisdriven functional differentiation, as reflected in greater network specialization⁹⁰.

The current study contributes to the emerging body of evidence suggesting that cannabis use may confer neurocognitive benefits in older adults by modulating the organization of functional brain networks. The observed effects imply that cannabinoids may exert neuroprotective influences in aging populations, potentially through their regulatory roles in maintaining or enhancing functional brain segregation and integration. These findings align with theories of neural differentiation and compensatory adaptation, which suggest that age-related changes in brain function can be mitigated by the recruitment or preservation of specialized network structures, partly supported by the endocannabinoid system.