Keyword: Epilepsy
2 results found.
Review Article
Epidemiology and Health Data Insights, 1(6), 2025, ehdi021, https://doi.org/10.63946/ehdi/17420
ABSTRACT:
Background. There is growing interest in the association between antiepileptic drugs (AEDs) exposure and subsequent Parkinson’s disease (PD).
Methods. We conducted a literature search in the PubMed, SCOPUS, and Web of Science databases. We identified studies using an observational design and performed a meta-analysis to evaluate the association between AEDs exposure and incident PD. We assessed the quality of the studies and identified the pooled odds ratio (OR) for those exposed to AEDs compared to those who were not.
Results. Of the 1,775 unique studies identified, 55 were selected for full-text review. Five studies (n = 127,324) were included. Quality assessment revealed moderate-to-high methodological quality in the studies included. The overall OR for a PD was 1.82 (95% CI: 1.35-2.45) in AEDs recipients. When considering each drug individually, the magnitude of association was highest for valproate (OR 3.94, 95% CI: 3.15-4.92) and lowest for carbamazepine (OR 1.32, 95% CI: 1.16-1.49). Further interaction tests revealed higher odds for lamotrigine than for carbamazepine and valproate than for carbamazepine and lamotrigine.
Conclusion. This study revealed potential associations between AEDs and incident PD. However, existing evidence remains insufficient, making it premature to draw inferences on this matter.
Methods. We conducted a literature search in the PubMed, SCOPUS, and Web of Science databases. We identified studies using an observational design and performed a meta-analysis to evaluate the association between AEDs exposure and incident PD. We assessed the quality of the studies and identified the pooled odds ratio (OR) for those exposed to AEDs compared to those who were not.
Results. Of the 1,775 unique studies identified, 55 were selected for full-text review. Five studies (n = 127,324) were included. Quality assessment revealed moderate-to-high methodological quality in the studies included. The overall OR for a PD was 1.82 (95% CI: 1.35-2.45) in AEDs recipients. When considering each drug individually, the magnitude of association was highest for valproate (OR 3.94, 95% CI: 3.15-4.92) and lowest for carbamazepine (OR 1.32, 95% CI: 1.16-1.49). Further interaction tests revealed higher odds for lamotrigine than for carbamazepine and valproate than for carbamazepine and lamotrigine.
Conclusion. This study revealed potential associations between AEDs and incident PD. However, existing evidence remains insufficient, making it premature to draw inferences on this matter.
Methodological Paper
Epidemiology and Health Data Insights, 1(5), 2025, ehdi018, https://doi.org/10.63946/ehdi/17368
ABSTRACT:
Background. This study examined the reporting practice of subgroup effects of meta-analytic research published in the leading journal of the ILAE.
Methods. We selected studies that used ratio measures and employed subgroup analyses. Subgroup differences were calculated as the difference between the log-transformed estimates over the square root of the sum of the squared standard errors. For the calculated test scores, a corresponding two-tailed p-value was calculated using the standard normal cumulative distribution function. The authors also conducted additional analyses accounting for multiple comparisons.
Results. The literature search identified 55 publications, of which 14 (25 %) were included. Neither study used a formal test to compare the subgroups. The number of reported subgroup estimates ranged from 2 to 20, and the number of pairwise comparisons ranged from 1 to 53. Overall, there were 187 comparisons, resulting in a median log difference of 0 (IQR 0.12 – 0.17) and a range from -2.92 to 2.32. The median p-value was 0.54 (IQR 0.21-0.85) with 18 (9%) comparisons showing p-values lower than the conventional significance level, whereas 6 and 21 contrasts were 0.05 < p ≤ 0.10 and 0.10 < p ≤ 0.20, respectively. Seven (4%) comparisons resulted in a p-value lower than the corrected significance level when adjusted for multiple comparisons.
Conclusion. There was a lack of compliance with the reporting guidelines. The findings from the subgroup analyses were commonly interpreted without employing a formal test. There is need to emphasize the importance of adherence to established reporting standards when presenting the subgroup effects.
Methods. We selected studies that used ratio measures and employed subgroup analyses. Subgroup differences were calculated as the difference between the log-transformed estimates over the square root of the sum of the squared standard errors. For the calculated test scores, a corresponding two-tailed p-value was calculated using the standard normal cumulative distribution function. The authors also conducted additional analyses accounting for multiple comparisons.
Results. The literature search identified 55 publications, of which 14 (25 %) were included. Neither study used a formal test to compare the subgroups. The number of reported subgroup estimates ranged from 2 to 20, and the number of pairwise comparisons ranged from 1 to 53. Overall, there were 187 comparisons, resulting in a median log difference of 0 (IQR 0.12 – 0.17) and a range from -2.92 to 2.32. The median p-value was 0.54 (IQR 0.21-0.85) with 18 (9%) comparisons showing p-values lower than the conventional significance level, whereas 6 and 21 contrasts were 0.05 < p ≤ 0.10 and 0.10 < p ≤ 0.20, respectively. Seven (4%) comparisons resulted in a p-value lower than the corrected significance level when adjusted for multiple comparisons.
Conclusion. There was a lack of compliance with the reporting guidelines. The findings from the subgroup analyses were commonly interpreted without employing a formal test. There is need to emphasize the importance of adherence to established reporting standards when presenting the subgroup effects.