A Systematic Review and Meta-analysis of Prevalence of Epilepsy, Dementia, Headache, and Parkinson Disease in India
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.314588
Source of Support: None, Conflict of Interest: None
Keywords: Dementia, epilepsy, headache, meta-analysis, Parkinson's disease, prevalenceKey message: This study presents the most updated pooled prevalence of epilepsy, dementia, headache, and PD in India that provides evidence to guide policies. Efforts should be directed at using uniform and standardized criteria's for diseases for conducting prevalence studies.
Mental and neurological disorders (MNDs) are increasingly indicated as a public health concern due to ongoing epidemiological transition from communicable to non-communicable diseases over the past decade and inadequate availability of treatment in MNDs. As per Global Burden of Disease study (2016), India contributes to 15% of the global burden of diseases due to MNDs and by 2025, it is estimated that the burden of MNDs will increase by 23% in India. MNDs like epilepsy, dementia, headache, and Parkinson's disease (PD) is associated with high morbidity and accounts for 18% of the totalYears Lost to Disability (YLDs) in 2017 as per a new study published by the Indian Council of Medical Research (ICMR). Owing to the huge populationand inadequate medical and diagnostic facilities in many states, there is a tremendous treatment gap in MNDsin India. Only one out of 10 patients with MNDs receives evidence based interventions in India. In view of importance of addressing MNDs, estimating the prevalence is crucial to target action on their clinical and socioeconomic impact.
Variability in the prevalence rates of neurological disorders has been observed across various epidemiological studies in the country. For example, the prevalence of epilepsy have been reported between 2.2 and 10.4 per 1,000 population across different regions in India. This variation in the prevalence rates of neurological disorders attributes to differences in the manner the cases are defined and identified in each study, and the inclusion and exclusion criteria's used, which poses a problem to understand the actual burden of neurological disorders. A meta-analysis of 20 studies to estimate the prevalence of epilepsy was published in 1999 but thereafter no meta-analysis has been published on the prevalence of common neurological disorders in India.The aim of this study was to conduct a systematic review and meta-analysis of the available studies to estimate the prevalence of epilepsy, dementia, headache, and PD in India.
Protocol and registration
The review protocol was registered in PROSPERO (https://www.crd.york.ac.uk/PROSPERO, registration number CRD42017078492), which is an international database of prospectively registered systematic reviews.
We reviewed PubMed for peer-reviewed studies from January 1980till March 2019 using search terms “disease name,” that is,”epilepsy” and “seizure,””dementia,””headache,”and “Parkinson's disease” combined with each of the terms “epidemiology,””prevalence,””incidence,””mortality,”and “morbidity.” Each phrase was combined with “India” using Boolean operator “AND” and the search was done in advanced search drop down menu under “title/abstract” category for retrieving the relevant studies. In addition to PubMed, a thorough literature search was done in Global Health Data Exchange (GHDx) data catalogue of Institute of Health Metrics and Evaluation (IHME) at http://ghdx.healthdata.org/geography India. Additional searches were also done in Google Scholar using above mentioned terms under “with the exact phrase” category. Hand searching of relevant grey literature like government websites and conference proceedings was also undertaken.
Study inclusion criteria
Prevalence studies were done in India between January 1980 and March 2019 (inclusive) were included. We searched for cross-sectional studies and community-based surveys published in English language that reported the prevalence rates of the disorders. For each disorder, different case definitions were adopted for inclusion of studies. For example, in epilepsy, studies that used either the definition as “two or more unprovoked seizures or seizure episodes” or that used International League Against Epilepsy (ILAE) definitions, were included. In addition, we also included studies that did not clearly report the definition criteria but were done on large representative samples. Editorials, brief communications, and conference proceedings, where relevant data was available were included. Studies that were not prevalence studies and those not conducted in India were excluded from the review. The methodological quality was independently assessed by two reviewers (VD and GRM) using Joanna Briggs Institute Critical Appraisal tool for prevalence. Disagreements were resolved by discussion with a third reviewer (DJ).
For each included study, information on the case definitions used, the total population screened, the number of cases detected or the prevalence rates (whichever was available) were extracted independently by two reviewers (VD and GRM) using an excel data extraction sheet. All disagreements were resolved through detailed discussion and consensus between the two authors (VD and GRM) and an independent review by the third reviewer (DJ).
This systematic review is compliant with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement, which is an evidence-based minimum set of items for reporting in systematic reviews and meta-analyses.
Data analysis was done in Metafor—a free open source add-on for conducting meta-analysis with the statistical software R (http://www.metafor-project.org/doku.php). The package consists of a collection of functions that allow users to calculate various effect sizes or outcome measures, fit fixed-effects, random-effects, and mixed-effects models to such data and carry out moderator and meta-regression analysis. The prevalence data was analyzed to generate various types of meta-analytical plots like funnel plots (publication bias), sensitivity analysis, and sub-group analysis. Meta-analytic techniques for the analysis of observational studies to quantitatively integrate the data across multiple studies to derive pooled estimates of prevalence were used.
The method of analyzing the prevalence data is based on the meta-analysis of proportions that includes studies that are observational and non-comparative. The heterogeneity across studies were examined using Cochran's Q test and quantified by I2 statistic that describes the percentage of variation across studies resulting in heterogeneity rather than chance with its values of 25%, 50%, and 75% indicating low, moderate, and high heterogeneity respectively. The pooled prevalence of the disorders were calculated using a random effects model by DerSimonian and Laird method.
Out of a total of 723 articles (PubMed: 546, Google Scholar: 121 and GBD database: 56), 623 articles were duplicates and not-relevant based on the inclusion–exclusion criteria's and hence were excluded. A total of 100 articles (Epilepsy = 51, Dementia = 25, Headache = 16, PD = 8) were available for critical appraisal. Sixty articles were assessed for eligibility after removing 40 articles in which data was either insufficient or not available. A total of 50 studies were ultimately included for quantitative analysis of the data. The total numbers of studies found relevant were: epilepsy (22), dementia (19), headache (6), and PD (3). [Figure 1] shows PRISMA chart for screening and selection of studies retrieved from various databases using the search strategy.
The summary of the studies included in the systematic review are presented in [Table 1], [Table 2], [Table 3]. The prevalence rates for epilepsy were reported in all 22 studies that included 13, 52,074 individuals. The prevalence rates for dementia and headache were reported from 17 and 6 studies including a total of 2, 12,769 and 9,329 individuals, respectively. PD prevalence rates were reported in 3 studies covering 2, 17,369 individuals. Majority of the included studies were published in the past two decades. The articles on prevalence of epilepsy were published between the years 1998 and 2016. Twelve of these studies reported data from rural areas, six from urban, and four from both urban and rural areas. The included articles on prevalence of dementia were published between the years 1997 and 2014 and 12 of these studies reported data from urban areas, five from rural areas, and two from both urban and rural areas. Two of the six articles on the prevalence of headache published between 2003 and 2016, reported data from urban areas and four from both urban and rural areas. Of the three PD studies published during 1988–2007, two provided data from urban areas. More studies reported data from the southern region for dementia (South: 7, North: 5, East: 4, West: 3) and headache (South: 4, North: 2) while the north-south distribution for epilepsy studies (South: 7, North: 7, East: 6, West: 1) and PD studies (East = 1, South = 1, West = 1) was equal.
The Joanna Briggs Institute Critical Appraisal checklist for prevalence studies was used to critically appraise the included studies [Table 1]. The number of studies in each disorder with maximum score (i.e., 9) were epilepsy (n = 9/22), dementia (n = 3/19), headache (n = 3/6), and PD (n = 2/3)
Pooled prevalence estimates
The prevalence of epilepsy, dementia, headache, and PD was calculated from the studies included. The pooled prevalence of epilepsy was 4.7per 1,000 population (95% CI: 3.8–5.6) when all the 23 studies were included in the analyses [Figure 2]. Of the 23 studies, study by Krishnaiah et al. reported a prevalence of 8.8%, which was an outlier study. As expected the heterogeneity in the studies was high (P < 0.01, I2 = 98%). The prevalence of dementia was found to be 33.7 per 1,000 population (95% CI: 19.4–49.8) (P = 0, I2 = 99%) [Figure 3]. The pooled prevalence of headache and PD were found to be 438.8 per 1,000 population (95% CI: 287.6–602.3) (P < 0.0001, I2 = 97.99%) [Figure 4] and 0.8 per 1,000 population (95% CI: 0.4–1.3) (P < 0.01, I2 = 95%), respectively [Figure 5].
Funnel plots and Egger's regression tests revealed no publication bias in the estimates of epilepsy (z = -1.6785, P = 0.0933) and dementia (z = 0.0397, P = 0.9683) prevalence rates. The number of studies in headache and PD groups were too less to assess the publications bias.
Sensitivity analysis and cumulative meta-analysis
To evaluate the robustness of the prevalence results, we performed leave-one-out analysis for epilepsy and dementia by iteratively removing one study at a time and recalculating the pooled prevalence. In epilepsy, the outlier analysis showed that the outlier study (Krishnaiah et al. 2016) was not significantly associated with the outcome. The pooled prevalence rate after removing this study from the analysis was 4.2 per 1,000 population (95% CI: 3.5–5.1). Five studies (i.e., Seby etal., Rodriguez etal., Mathuranath etal., and Sengupta etal.) reported very high prevalence rates as compared to the other studies. The forest plot after excluding these outlier studies from the analysis yielded a pooled prevalence of 20.8 per 1,000 [95% CI 11.3–33.1].
Subgroup analysis for epilepsy was performed to determine the prevalence of epilepsy for each decade. It is found that the pooled prevalence of epilepsy in India till the year 2000 was 3.8 per 1,000 population (95% CI: 3.0–4.8) (Q = 131.7, P < 0.0001, I2 = 94.68%), which increased to 5.3 per 1,000 population (95% CI: 3.5–5.0) (Q = 501.5, P < 0.0001, I2 = 97.01%) by 2010. The sub-group analysis could not be performed in other disorders due to less number of studies.
In addition to the calculations of prevalence, we performed a mixed effects model meta-regression analysis to see if the year of study, region of the study, and JBI score has any effect on the pooled prevalence of epilepsy. The analysis showed no significant association between these variables and the pooled prevalence of epilepsy.
Various epidemiological studies in neurology started in the 1980s, which marked the emergence of discipline of neuroepidemiology in India but there are wide variations in the reported prevalence rates of neurological disorders across India due to multiple factors such as non-standardization of diagnostic criteria's, case definitions, and questionnaires used to detect the positive cases. Hence the aim of this study was to calculate the pooled prevalence rates of common neurological disorders, for example, epilepsy, dementia, headache, and PD in India by doing a systematic review and meta-analysis of the existing prevalence studies in the literature.
A total of 51 studies were included in the present analysis covering a total population of 17, 91,541 for all four disorders. Epilepsy is the most common serious neurological disorder with significant morbidity and social impact. The prevalence of epilepsy has been reported as 6.2–7.6 and 5–10 per 1,000 population in developed and resource poor countries, respectively. In the present study, we found the prevalence of epilepsy as 4.3 per 1,000 population, which is similar to a meta-analysis of prevalence studies of epilepsy done by Bharucha et al. in 2003 (5.59 per 1,000 population). The prevalence of epilepsy found in this study is in accordance with most prevalence studies conducted in India. Epilepsy is associated with social stigma and as compared to dementia, headache, and PD, most of the prevalence studies in epilepsy were done in the rural settings, which provide important information on prevalence of this disorder in the rural areas where majority of the population resides. The meta-regression analysis in the present study did not show any significant difference between prevalence of epilepsy in urban vs rural areas.
Dementia is a disease of elderly and there are 24.2 million people live with dementia worldwide, with 4.6 million new cases every year. The crude prevalence rate of dementia in India varies from 8 to 35 per 1,000 among population above 55 years of age. The prevalence of dementia found in present meta-analysis is 23.7 per 1,000 population, which is higher than already reported prevalence rates in India. This may be due to the fact that we included adults with more than 50 years of age while previous studies included either >55 years or >60 years of age groups.
Headache disorders are one of the commonest disorders in the world affecting both genders and all socioeconomic strata of society and continue to remain an unrecognized health problem. They account for substantial disability and high economic burden but are given poor share of resource allocation. The reporting of prevalence of headache in India has been very wide, ranging from as low as 0.2% to a high of 58% mainly due to varied criteria's used to define different types of headaches. The pooled prevalence of headache in the present study was found to be 438.8 per 1,000 population, which is higher than previously reported rates. In the present meta-analysis, we included tension type headache, migraine, and chronic daily headache as single entity, which may be the reason for higher rates reported in this study. The prevalence rate of PD in India is considered low except in certain ethnic groups like Parsis in Mumbai. The reported crude prevalence rate of Parkinsonism in India varies from 33 to 328 per 100,000 populations. In this study, the prevalence of PD was found to be 0.7 per 1,000 population. This wide discrepancy in the prevalence rates of PD in India may be due to genetic diversity, ethnicity, environmental factors, and varied methodology.
Several limitations should be considered while interpreting the results of this study. There are no uniform definitions of diseases used for conducting prevalence studies across various centers. The studies included represented only a few areas of the country and hence the representation of the sample was inadequate. We found only a few studies on prevalence of headache and PD for meta-analysis. Although all published studies were included, there are chances of missing some studies from the grey literature.
This study presents the most updated pooled prevalence of epilepsy, dementia, headache, and PD in India that provides evidence to guide policies. The review has identified major lacunae in ascertainment and case definition of these disorders, selection of source population, and inclusion and exclusion criteria resulting in wide heterogeneity observed in the results. Efforts should be directed at preparing standardized definitions of diseases that will help in minimizing the inter-study variations, reduce possible selection bias of subjects and result in more reliable and comparable estimation of morbidity.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]
[Table 1], [Table 2], [Table 3]