Epilepsy is one of the most common neurological disorders (1) and its treatment is life-long in nearly one-third of the patients (2). Several large epidemiological studies indicated that patients with epilepsy are at two- to six-folds increase in the risk of fractures compared to the general population (3). It has been reported that more than 15% of fractures are pathological due to metabolic bone disease (3, 4). Early reports suggested that institutionalization, inadequate dietary intake with vitamin D deficiency/insufficiency, reduced sunlight exposure (5), physical inactivity (6), aging (7), and female sex (8) are risk factors for osteopathy with epilepsy. However, many studies reported that antiepileptic drugs (AEDs) are major risk factors for bone disease in non-institutionalized patients and well-nourished ambulatory adults with epilepsy (9–12). Early reports incriminated AEDs with enzyme induction activity or their metabolism through hepatic cytochrome P450 enzyme system, as causes of metabolic bone disease due to increasing metabolism of vitamin D with secondary hypocalcemia and hyperparathyroidism (13–15). However, several studies reported that both enzyme inducing and non-enzyme inducing AEDs are potentially implicated for the comorbid metabolic bone disease (10, 12).

Bone mineral density (BMD) tests can identify osteopenia/osteoporosis, determine the risk for fractures, and measure the response to osteoporosis treatment. BMD values depend upon bone mineral content (BMC) and bone size. The most widely recognized BMD test is called a dual-energy X-ray absorptiometry (DEXA) (9, 16–18). DEXA is a non-invasive, painless, easy to use, and with less exposure to irradiation. Biochemical indices of bone and mineral metabolism include measurement of blood concentrations of calcium, phosphorus, vitamin D, parathyroid hormone (5, 19–23), and bone remodeling markers [e.g., the receptor activator of nuclear factor-kappa B ligands (RANKL) and osteoprotegerin (OPG)] (24). Such markers may indirectly identify patients with bone disease.

The severity of bone disease with AEDs is heterogeneous. In cases with mild severity, there may be more subtle changes in bone quality and hypovitaminosis D, hypocalcemia, and hyperparathyroidism. Cases of intermediate severity may exhibit characteristic features of osteopenia/osteoporosis. Some patients with severe bone disease may manifest features of osteomalacic disorder (5, 13, 25).

The exact pathophysiological mechanisms of secondary or pathologic bone loss with different AEDs are still uncertain. However, some mechanisms have been suggested, which include hepatic enzyme induction with increase in the metabolism of vitamin D (13–15), impaired calcium absorption (26), direct effect of AEDs on bone cells (9), and resistance to parathyroid hormone and inhibition of calcitonin secretion (27). Identification of exact pathogenic factors involved in development of bone disease with epilepsy and its medications is of primary importance for preventive and treatment strategies.

This is a comparative cross-sectional case–control study. This study included 75 adults with age ranged between 20 and 50 years and duration of illness of 6–25 years. The patients’ seizure types were diagnosed according to the International League Against Epilepsy (ILAE) criteria (28). Table 1 showed the demographical and clinical features of the studied patients. Patients were recruited from the out-patient epilepsy clinic of the Department of Neurology and Psychiatry of Assiut University Hospital, Assiut, Egypt. All patients were treated with conventional AEDs (CBZ, VPA, or CBZ + VPA) for at least 6 months before participation in this study. Forty healthy subjects recruited from the general population and matched for age- (range: 20–50 years; mean: 30.36 ± 7.59), sex-, and socioeconomic status were also included in this study for statistical comparisons.

Exclusion criteria (for patients and control subjects) included: (1) a history of an overt medical disorder before the onset of epilepsy which likely affect bone metabolism [e.g., thyroid (hyperthyroidism and hypothyroidism), hepatic, renal hematological, rheumatologic and gastrointestinal diseases, hypogonadism, and malabsorption], (2) medications known to affect bone turnover or can contribute to bone loss (e.g., glucorticoids, bisphosphonates, thiazides, anticoagulants, and steroids), (3) intake of calcium or vitamin D supplements, (4) risk factors for osteoporosis [e.g., family history of osteoporotic fracture, estrogen deficiency resulting from early menopause (before age 45 years), either naturally from surgical removal of the ovaries or as a result of prolonged amenorrhea (abnormal absence of menstruation) in younger women, and (5) overt bone deformities (e.g., osteogenesis imperfecta).

Details of the protocol were approved by the local ethics committee and all patients gave written informed consent to their participation.

All patients and healthy control subjects underwent the same research protocol, which included medical, neurological, and endocrinological histories and examinations.

Data collection included age, sex, weight, height, and body mass index (BMI). Seizure variables recorded were: age at onset, duration of illness, type, frequency, type of utilized AED(s) (monotherapy or polytherapy), duration of treatment, degree of patients’ control on AED(s), and obvious side effects from medications. Regarding the degree of control on AEDs, patients were considered controlled on AEDs, when seizure free for ≥1 year.

This study included 75 adults with primary epilepsy, with mean age of 31.90 ± 5.62 years and duration of illness of 12.65 ± 5.42 years. Patients were treated with one or more conventional AED(s) with nearly three quarter of the patients were on CBZ mono- or polytherapy for 10.57 ± 3.55 years. Nearly, two-thirds of the patients were controlled on the prescribed AEDs. Tables 2 and 3 showed the DEXA results of the studied groups and the frequency of patients with DEXA abnormalities. Compared to control subjects, patients had significantly lower BMD (p < 0.001; p < 0.0001), BMC (p < 0.001; p < 0.001), Z-score (p < 0.0001; p < 0.0001), and T-score (p < 0.0001; p < 0.0001) at the femoral neck and lumbar spine, respectively. Abnormalities in more than one site [i.e., femoral neck (unilateral or bilateral) and lumbar spine] were seen in affected patients. The majority of male patients had osteoporosis of the femoral neck and lumbar spine while the majority of females had osteopenia regardless of the type, dose, or serum level of AEDs [versus 20% (n = 8) for control subjects with osteopenia in the lumbar spine]. Bone pain (particularly low back pain) was reported in 45 patients (33.75%) particularly in female (n = 36, 80%). Twenty patients (15%) had history of limb fracture (males = 13, 65%; females = 7, 35%) before participation in this study. Tables 4 and 5 showed the laboratory results of the studied groups and the frequency of patients with abnormal biochemical parameters. Compared to control subjects, patients had significantly lower serum calcium (p < 0.0001), 25(OH)D (p < 0.0001), and OPG (p < 0.001) and higher ALP (p < 0.001), sRANKL (p < 0.0001) levels, and sRANKL/OPG (p < 0.0001). Nearly, 52% of the patients had hypocalcemia [versus 25% or n = 12 control subjects], 93% of the patients had vitamin D deficiency/insufficiency [versus 30% or n = 12 control subjects with vitamin D insufficiency], and 31% had high levels of sRANKL and 49% had low levels of OPG. No significant differences were identified between the levels of different laboratory and DEXA results in relation to the age, gender, weight, type, dose, or serum levels of AEDs, mono- versus polytherapy and the degree of control on AEDs. Correlations between demographical-, clinical-, laboratory-, and DEXA parameters showed that in patients with epilepsy (Table 6): (1) duration of illness and BMD at the femoral neck and lumbar spine respectively, (2) the duration of treatment with AEDs were correlated with BMD at the femoral neck and lumbar spine, respectively, levels of calcium, 25(OH)D, sRANKL, and OPG, (3) BMD at the femoral neck and lumbar spine and serum levels of 25(OH)D, sRANKL, and OPG, (4) levels of calcium and 25(OH)D, (5) sRANKL and OPG. In linear regression analysis and after adjusting for gender, age, weight, duration of therapy, and number of AEDs, we observed an association between BMD, 25(OH)D (partial R² = 1.5, p = 0.04), and sRANKL (partial R² = 1.7%, p = 0.03) concentrations.

Although the knowledge regarding disease- and AEDs-related bone abnormalities are rapidly increasing; however, data regarding the mechanisms of bone loss in patients with epilepsy are not clearly understood. In this study, the potential relationships between the BMD findings in a group of ambulatory adults with epilepsy and treated with AEDs and markers of bone remodeling, which include calcium, 25(OH)D, ALP, and OPG/RANKL systems were discussed. The results of this study showed that in ambulatory adults with epilepsy and on chronic treatment with AEDs, there is an increased risk of (1) bone loss (osteopenia/osteoporosis) and fracture, (2) hypocalcemia (52%) and 25(OH)D deficiency/insufficiency, (3) increased bone turnover as indicated by lower serum levels of OPG and higher levels of sRANKL and sRANKL/OPG ratio (markers of bone remodeling), and (4) metabolic bone compromise with prolongation of the duration of treatment with AEDs due to disturbed mineral metabolism associated with compromise of bone remodeling with accelerated bone turnover. These changes occur regardless of patients’ age, gender, weight, type, dose, or serum levels of AEDs, number of utilized AEDs, and the degree of control on AEDs, but BMD loss was also associated with prolongation of the disease duration which also indicated prolonged duration of treatment.

First, the findings of this study showed increased markers of bone loss as evidenced by lower BMD, BMC, Z-score, and T-score at the femoral neck and lumbar spine regardless of the type, dose, or serum level of the utilized AED. Osteoporosis of the femoral neck was reported in 21.33% while that of the lumbar spine was reported in 72%. Osteopenia of the femoral neck was reported in 29.33% while that of the lumbar spine was reported in 13.33%. No significant differences in DEXA results were identified between patients on CBZ, VPA, or polytherapy. Bone mass changes reported in the literature vary greatly for patients undergoing treatment with AEDs and ranged between 13 and 75%. However, both enzyme inducing and non-enzyme inducing AEDs are potentially implicated for bone mass loss with epilepsy (10, 12–15, 20). The majority of male patients had osteoporosis of the femoral neck and lumbar spine while the majority of females had osteopenia. The variations between the results of different studies may depend on the type of AEDs used, and the AED blood level required for effective treatment. Other factors may include the length of time of treatment, the existence of associated pathologies, the age and sex of the patients, their level of physical activity, and where and how the measurements are made. We reported history of fracture in 15%, which was predominant in males (7, 32). It has been reported that gender plays a major role in the pathogenesis of fractures. In accordance, Cummings et al. (33) reported that fractures commonly occurred in males aged <40 years (between 30 and 49) and then rapidly declined while the peak rate occurred 20 years later (between age 50 and 59 years) in females, and declined subsequently.

Second, the findings of this study showed that biochemical abnormalities of bone-related mineral metabolism included hypocalcemia (52%) and hypovitaminosis D (93.33%). 25(OH)D levels were positively correlated with calcium levels but negatively correlated with BMD. 25(OH)D deficiency/insufficiency was reported irrespective to the type of AED (17). Hypovitaminosis D was previously reported in 50% of ambulatory patients on AEDs. The other common laboratory abnormalities described in relation to AED are elevated levels of ALP and hyperparathyroidiam (5, 13, 17, 19, 20, 34). It has been suggested that hepatic induction of cytochrome P450 enzymes by enzyme inducing AEDs (e.g., phenobarbital, phenytoin, and CBZ) increases the metabolism and clearance of vitamin D with secondary hypocalcemia and hyperparathyroidism (13–15). The mechanism of disturbed mineral metabolism due to VPA is not completely understood. VPA may cause renal tubular dysfunction, which could indirectly influence bone and mineral metabolism with increased urinary loss of calcium and phosphorus. In support, VPA has been associated with reversible Fanconi syndrome (35). It is known that approximately 99% of calcium in the body is stored in the skeleton. Calcium is stored in or removed from the skeleton to regulate serum calcium levels. 25(OH)D is the major circulating form of vitamin D. 25[OH]D is biologically inert and transported to the kidney, where renal 1a-hydroxylase converts 25(OH)D to 1,25(OH)2D. 1,25(OH)2D is an important regulator of serum calcium concentration. 1,25(OH)2D directly enhances the entry of calcium through the plasma membrane into intestinal absorption cells and transfers calcium across the basolateral membrane into the circulation to maintain calcium homeostasis. 1,25(OH)2D promotes mineralization of osteoid and, with appropriate levels of calcium and phosphorus, induces deposition of calcium hydroxyapatite into the bone matrix (36). Controversial results reported lack of any correlation between vitamin D and BMD in patients with epilepsy in some studies (4, 37). Some suggested that AEDs may inhibit vitamin D-mediated calcium absorption (26). Filardi et al. (38) reported that adults who were on chronic anticonvulsant therapy and who lived in low latitude regions had normal bone mineral density as well as vitamin D serum level. Mattson and Gidal (14) reported low BMD irrespective of vitamin D levels. Verrotti et al. (22) reported that VPA doses do not seem to change bone turnover in adult epileptic patients. This could be explained by vitamin D-independent pathways such as impaired calcium absorption (26), resistance to parathyroid hormone and inhibition of calcitonin secretion (27), direct effects of AEDs on bone cells (9), hyperparathyroidism and calcitonin deficiency (39).

Third, this study showed reduced levels of OPG and higher levels of sRANKL and sRANKL/OPG ratio in patients with epilepsy, the duration of treatment with AEDs were positively correlated with sRANKs and negatively correlated with OPG, BMD at the femoral neck, and lumbar spine and serum levels of sRANKL and OPG and in linear regression analysis and after adjusting for gender, age, weight, duration of therapy, and number of AEDs, we observed an association between BMD, 25(OH)D, and sRANKL concentrations.

Bone is a complex dynamic tissue. It undergoes remodeling, a continuing process of bone resorption and formation in discrete areas throughout the skeleton. BMD represents a coordinated action or the balance between bone resorption or the bone-depleting action of osteoclasts and the bone-formative action of osteoblasts (bone remodeling). Osteoclast and osteoblast (main bone cells) dynamics, body weight, exercise, vitamin D, calcium, and phosphorous homeostasis as well as connective tissue arrangements can alter bone structure and architecture. Several hormones, polypeptides, cytokines, growth factors, nutrients, minerals, and essential cofactors, regulate bone remodeling. The active metabolite of vitamin D [1,25(OH)2D] and adequate amounts of calcium and phosphorus are required for bone formation (40). Bone remodeling is intimately related to changes in expression of members of tumor necrosis factor (TNF) receptor superfamily on bone marrow stromal cells [e.g., the macrophage colony-stimulating factor (M-CSF); the receptor activator of nuclear factor-kappa B ligands (RANKL), RANK and OPG] and various cytokines and calciotropic hormones (41). RANKL binds the RANK receptor on osteoclast precursors and induces the formation of osteoclasts by a special signaling pathway (42, 43). A soluble form of RANKL (sRANKL) is released from its membrane bound state by metalloproteinases. OPG blocks the RANKL–RANK interaction by binding to and neutralizing both soluble and cell-bound forms of RANKL thus inhibits osteoclast differentiation and activation (44). It is the balance between the expression of RANKL and OPG that determines the extent of bone resorption. It has been suggested that in patients with epilepsy and in order to improve calcium absorption when the 25-OH Vitamin D level is low, 1,25OHD is increased, a conversion step mediated by parathormone. Elevation of the active metabolite of vitamin D, 1,25OHD, can stimulate bone resorption through the receptor-activated NF-KB ligand (RANKL), promoting maturation of osteoclasts (24). In support, bone biopsies from patients with epilepsy revealed diminished volume of trabecular bone and increased amount of osteoclastic resorptive activity and osteoid in 10 and 40% of patients receiving long-term AED therapy (37).

Despite the strength of this study, it has some limitations: (1) the recruitment of the study group from a tertiary care center with more severe cases and this explains the high percentage of observed bone loss evidence by DEXA and the biochemically related abnormalities, (2) because of the cross-sectional nature of the present study, it was not possible to know temporal relationship between bone loss to the use of AEDs. However, such limitations could only be overcome through a longitudinal and multi-center study design.

Ambulatory adults with epilepsy and chronic treatment with AEDs are at greater risk of metabolic osteopathy as evidenced by osteopenia/osteoporosis, fracture, hypocalcemia, and hypovitaminosis D. The metabolic bone loss and mineral abnormalities are associated with accelerated bone turnover as evidenced by higher levels of sRANKL and sRANKL/OPG ratio and lower levels of OPG (markers of bone remodeling). Thus, patients receiving long-term AED have to be monitored for indices of bone health, including BMD and vitamin D status and to be provided with proper preventive and treatment strategies.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.