Our findings are important, as they demonstrate that the high rates of anticholinergic burden experienced by adults with intellectual disabilities at all ages (not just older ages) are due to several classes of medications, not only psychotropic, and which may be prescribed by several different clinicians. Whilst the general practitioner has an overview, it is important for all clinicians to be cognisant of this issue. Of the 38.4% of adults with intellectual disabilities on long-term anticholinergic prescriptions, the medications most frequently prescribed were central nervous system medications (62.6%), gastro-intestinal medications (46.7%), and cardiovascular medications (28.4%), and polypharmacy (both anticholinergic and psychotropic) was more common than monotherapy.

Examining psychotropic contributions to anticholinergic burden showed that adults with intellectual disabilities were more likely to be prescribed antipsychotics, antiepileptics and anxiolytics/hypnotics compared to adults without intellectual disabilities, and had similar levels of antidepressant prescriptions. Results on antipsychotic medication use support previous evidence of an age-related increase for those with intellectual disabilities [18, 23,24,25], in contrast to the general population who showed a decreased use across age (Fig. 1). Whilst we find statistically significant differences between the groups from ages 25 onwards, the smaller numbers in older age groups incur large confidence intervals (Additional File 1: Table S1). The current results and previous evidence [18, 26], indicate long-term antipsychotic use, despite associated health risks, including anticholinergic effects, and best practice guidelines advocating non-pharmacological approaches in the long-term [16]. These findings suggest antipsychotics comprise a significant contribution to high anticholinergic burden.

Regarding anticholinergic antiepileptics, the greatest group difference in the most affluent neighbourhoods (Additional File 1: Table S2, Figure S1). This reflects the higher rates of prescribing in more deprived areas in the general population, which is a typical finding [10]. Prescribing is high across all neighbourhoods for the intellectual disabilities population, with no obvious deprivation gradient. This highlights the need for medication reviews and service provision in all neighbourhoods regardless of extent of deprivation.

Anticholinergic polypharmacy across all drug classes was present, and at 1.6 greater odds for the adults with intellectual disabilities. The combination of multiple lower risk drugs (ARS score 1) and high polypharmacy in adults with intellectual disabilities contributes to the overall clinical risk of high anticholinergic burden. Non-anticholinergic licensed alternatives could be explored in place of the more commonly used anticholinergic medicines. A realistic medicine patient centred approach is key and the prescription of psychotropic medication should be “informed by a comprehensive biopsychosocial assessment” [19]. Rates of multimorbidity (physical and mental ill-health) are high in people with intellectual disabilities [10], so polypharmacy (including psychotropic) can be indicated and appropriate. For example, around 25% of adults with intellectual disabilities have epilepsy requiring antiepileptic medications [12]. Although, antiepileptic drugs are commonly prescribed as mood-stabilisers and not only seizure-control [27], which can increase the risk of high anticholinergic burden.

Clinical services are typically organised around single conditions, which can pose additional risks for polypharmacy. Comprehensive, regular, and targeted medication reviews would help to identify excess prescribing. To see where medications can be rationalised one drug at a time, to establish support for treatment adherence, to enable good communication between the person’s different health care professionals/teams if they are accessing multiple services, and not focus simply on single-disease care pathways. This is well-recognised in the general population, e.g., recent NICE clinical guideline on multimorbidity [28], but has received little attention for people with intellectual disabilities. For example, a review on the effectiveness of medication reviews for those with intellectual disabilities included only 8 studies, with just 3 of good quality; all of which evidenced the reduction in medication-related problems after multidisciplinary medication reviews [29].

Inappropriate polypharmacy is a problem among adults with intellectual disabilities and has a high risk of adverse effects. Previous evidence from representative population samples of adults with intellectual disabilities report a high prevalence of medicine use (average of 4–7 prescriptions) and polypharmacy (overall between 21% and 38%, psychotropic polypharmacy between 23% and 41%) [30,31,32,33]. Higher rates of polypharmacy (54%) and psychotropic polypharmacy (66%) have been reported in older adults with intellectual disabilities (aged 40 +) [2, 34]. Our data are comparable to these results; similar to Axmon et al. [1], we find that most adults take at least 2 anticholinergic medications irrespective of age or sex. Our data include only anticholinergic prescriptions, therefore, are likely an underestimation of overall prescriptions and polypharmacy, e.g., many adults with intellectual disabilities take the antiepileptic sodium valproate, which is psychotropic but not anticholinergic. Results show adults with intellectual disabilities were 2–3 times more likely to experience long-term polypharmacy compared to their peers. Our findings support previous work reporting high psychotropic polypharmacy in this patient population to be predominantly due to antipsychotic and antiepileptic medication use. [3, 24]

Polypharmacy and anticholinergic burden are inter-related, but both show an independent dose–response relationship with all-cause hospital admissions and mortality in general population adults [9, 35]. Polypharmacy prevalence is increasing; a recent UK Biobank study reported anticholinergic burden to be 3–9 times higher between 1990 and 2015 [36]. Inappropriate prescriptions pose a health risk, adults with intellectual disabilities have a 2.7 times greater odds of hospitalisation due to psychotropic adverse medication events [37]. More generally, there is evidence that every prescribed drug leads to an increase in having a potential drug–drug interaction of clinical significance for this population (OR = 0.87 [0.72–1.00]) [31]. People with intellectual disabilities often experience multimorbidity which may necessitate polypharmacy, but this can also mean increased exposure to potential drug–drug interactions. Moreover, both polypharmacy and multimorbidity were independent significant predictors for mortality in older (aged 50 +) adults with intellectual disabilities [38]. The precise relationship between polypharmacy, anticholinergic burden, multimorbidity, and mortality is yet to be clarified.

The strength of the current study lies in the sample; ours was a large representative group of adults with intellectual disabilities age-, sex-, and neighbourhood deprivation-matched to general population adults without intellectual disabilities within Scotland’s largest NHS health board. The use of this record linkage allows for robust information on prescribed medication with anticholinergic effects, although only a year of data was analysed. The study’s specific focus is on anticholinergic medication, and not the total rates of encashed prescriptions; so conclusions on overall prescribing cannot be drawn. In addition, as this is prescription data, there is no clarity of patient usage, dosage, or clinical indication. Therefore, an important limitation is the lack of clinical data on the conditions, diseases, or indeed, actual anticholinergic burden as measured by serum activity assays. There is also a lack of information on the severity of the intellectual disabilities. Finally, we were unable to distinguish between regular or pro re nata (PRN) medication in our database. Whilst our conservative definition of only including medications with 3 + repeat prescriptions over the 12 months is highly likely to have included only long-term anticholinergic medication use, we cannot say with 100% certainty that it did not include some people with very high use of PRN (as required) medication. In addition, this means that we may have underestimated the extent of exposure to anticholinergic medication in view of the use of PRN antipsychotics in this population which did not meet our threshold definition.