This cross-sectional study investigated associations between anticonvulsant use and bone health in a large population-based sample of men and women. After adjustment for confounders, BMD and QUS parameters were generally lower in men using anticonvulsants compared to non-users. In women, this was not clearly the case, with the difference in BMD and QUS parameters between anticonvulsant users and non-users not reaching statistical significance.
Only two studies have looked at how anticonvulsant use affects QUS measures albeit in children and young adults. A small Italian study (n = 164, aged 2–21 years) investigating how anticonvulsants affect DXA and QUS measures in girls with or without Rett Syndrome (RS) found that anticonvulsant therapy was associated with lower bone measures, although fracture risk was not elevated [21]. Similarly, a cross-sectional study conducted in Spain found that valproate, but not carbamazepine, phenobarbital, lamotrigine, topiramate, vigabatrine or phenytoin was associated with lower QUS measures in children (n = 65, aged 6.5 ± 3.1 years) compared to non-users?; despite having 27.3% of their treated group taking two or more anticonvulsants [20]. On the other hand, several studies have looked at associations between anticonvulsant use and BMD [31–33], with most studies conducted primarily in patients with epilepsy and in paediatric populations [15]. It has been estimated that over 50% of anticonvulsant users develop bone anomalies [31], and studies conducted over time have shown that long-term use is associated with increased fracture risk and associated bone disease [32, 34]. In a large, multiethnic, postmenopausal cohort (n = 138,667, aged 50–79 years), anticonvulsant use was associated with an increased fracture risk, which was further increased with polypharmacy and enzyme-inducing anticonvulsants rather than the non-enzyme inducing type [32]. Interestingly, anticonvulsant use was not associated with BMD at the spine, hip and total body, but was associated with falls risk. While our results reflect the association between anticonvulsants and bone health in adults, it appears that our findings are concordant with the existing literature, that while anticonvulsant use may be broadly associated with bone deficits, associations in girls and women are less clear.
Chronic anticonvulsant use has been shown to increase fracture risk by 1.2–2.4 fold [35]; explained in part by hepatic enzyme-inducing anticonvulsants known to increase conversion of 25-hydroxyvitamin D into inactive metabolites [36]. Mezuk et al [37] observed that although enzyme-inducing anticonvulsants may pose a greater fracture risk (HR 2.19, 95% CI: 1.97–2.43), non-enzyme-inducing anticonvulsants are also associated with an increased fracture risk (HR 1.66, 95% CI: 1.54–1.79). Additionally, Lee et al [35] observed that first generation anticonvulsants such as valproic acid, carbamazepine, phenytoin and phenobarbital are associated with an increased rate of fragility fractures, when compared to newer agents.
Confounding by indication is another factor possibly playing a role [38]. Anticonvulsants are used in the treatment of epilepsy and bipolar disorder, with both having been shown to be independently associated with increased fracture risk, although the mechanism of action has been thought to be due to falls related to seizures in epilepsy [39]. In a recent systematic review, however, bipolar disorder was associated with a 20–80% increased risk of fracture, independent of age, sex, medication and comorbidities [14]. Decreased physical activity and other modifiable risk factors, such as diet, substance use, smoking, SES, sun exposure [40], medical comorbidities, polypharmacy and drug-induced metabolic imbalances [41, 42] may also contribute to the overall decline in bone health.
The results of this study, when taken in context of research showing that anticonvulsant use independently increases fracture risk, add to existing clinical findings, and may assist treatment decisions in managing an already vulnerable population. Adequate calcium supplementation could counteract the deleterious effects of anticonvulsants [43], while clinical trial evidence suggests that the association between calcium supplementation and improved bone health is weak [44]. The antiepileptic drug and osteoporosis prevention trial (ADOPT) reported similar findings, where over 69% of their cohort (n = 80 men, aged ≥ 58 years) had improved BMD measures with adequate calcium supplementation [45].
Both strengths and limitations need to be taken into consideration. The strengths of this study include age range and representative nature of the sample, and number of confounding variables tested. A limitation of this study is its cross-sectional nature, preventing conclusions on how BMD and QUS measures varied over time. Secondly, confounding by indication has not been considered, and the study may have been underpowered to detect a significant difference in women. Other factors for consideration are a healthy participant bias, given that patients approach clinics for assessment, and the small number of anticonvulsant users is an additional consideration, limiting subgroup analyses of specific agents. Lastly, any unidentified confounding may affect our findings.