Prevalence and correlates of epileptic seizure in substance-abusing subjects
*Surendra K. Mattoo, MD, Department of Psychiatry, Postgraduate Institute of Medical Education and Research, Chandigarh 160012, India. Email: firstname.lastname@example.org
Life-time prevalence of epileptic seizures was assessed in 626 consecutive patients treated for substance abuse. Seizures were reported in 8.63% (9.2% in alcohol abusers, 12.5% in opioid abusers). A total of 64.8% of the seizures were associated with substance use. These occurred during withdrawal in the alcohol cohort and during intoxication with dextropropoxyphene and withdrawal from heroin or poppy husk in the opioid cohort. Results indicate that seizures may be more common in older patients with longer duration of dependence among those abusing alcohol.
THE ASSOCIATION BETWEEN epileptic seizure (ES) and substance use has been recognized. The prevalence of seizures in patients with alcohol dependence is approximately 9.9%.1 Opioids have been described as convulsant and anti-convulsant.2 In adults, recent heroin use may be a risk factor for seizure development.3 There are case reports of seizures during heroin withdrawal,4 and overdose of dextropropoxyphene.5
The aims of the present study were therefore to clarify the lifetime prevalence of ES and its correlates in a population of treatment-seeking substance-abusing subjects.
The study was conducted at the Drug De-addiction and Treatment Center (DDTC), Postgraduate Institute of Medical Education and Research, Chandigarh, India. Almost all patients at DDTC are men.
Consecutive new outpatient registrants and admitted patients (first or subsequent admissions) from 1 June to 30 November 2006 were assessed after obtaining informed consent. Sociodemographic data were recorded and clinical diagnoses made as per the ICD-10.6 ES was defined as per the International League against Epilepsy guidelines.7‘Lifetime prevalence of ES’ was defined as the occurrence of at least one ES of any etiology at any time in the lifetime of the subject8 and was expressed as percentage. The clinical diagnosis of ES was confirmed on arrival.7 The information was gathered from the patient and attendants. It included the history of diagnosis and treatment for epilepsy. The interview started with the following screening questions: have you ever lost consciousness; have you ever fallen on the ground; have you ever injured yourself without knowing it; have you ever had episodes in which you lost contact with surroundings; have you ever had uncontrollable shaking of arms and legs; and have you ever lost control of your bowels or bladder.9 If the answer to any of these was affirmative, a detailed clinical interview was conducted by one of the authors to arrive at a diagnosis of the presence/absence of ES. The temporal context of ES was coded as: during intoxication (I; F1X.06), withdrawal (W; F1X.31),6 or independent of substance influence. Duration of dependence was defined as the duration in years from intake into the study to the time when the patient first fulfilled criteria for dependence (F1x.24).6 Statistical analysis was done using SPSS version 10.0 (SPSS, Chicago, IL, USA).
Table 1 lists details on 626 patients (including two female patients, neither reporting ES) assessed; there was no refusal to participate in the study. Main substances of dependence were alcohol and opioids. The category ‘others’ included those with diagnoses such as harmful use (F1x.1).6 Buprenorphine was abused as an i.v. cocktail with benzodiazepines or antihistaminics, dextropropoxyphene as a combination with acetaminophen and diphenoxylate as a combination with atropine sulphate.
Table 1. Subject profile
|Opioids (by type)|
A total of 97% of ES-positive patients were codependent on nicotine. Two patients reported a myoclonic and partial type of ES each; the rest had generalized tonic–clonic seizure (including all substance-associated ES). ES with dextropropoxyphene was associated with the use of the same or 1.5 to 3–4-fold the normal dose.
Independent-sample t-test indicated that the patients in the alcohol group compared to those in the opioid group had a significantly longer mean duration of dependence (10.10 years vs 4.69 years, P < 0.01, F = 74.30, d.f. = 496, 95% confidence interval [CI]: 3.38–5.48) and were significantly older (38.73 years vs 28.28 years, P < 0.01, F = 127.05, d.f. = 496, 95%CI: 8.73–11.77). The age and duration of dependence, however, did not significantly differentiate ES-positive and -negative patients in either the alcohol or the opioid groups. Pearson χ2 test in the alcohol group showed that there were significantly more patients reporting ES than expected in the fifth and sixth decades of life (P < 0.01, F = 46.48, d.f. = 8).
Dependence on opioids develops more quickly than that on alcohol,10 thus treatment may be sought earlier in the former. This may explain the significant age differential of the two cohorts.
The prevalence and characteristics of ES in the alcohol cohort (withdrawal-related, generalized tonic–clonic and in patients significantly older with longer durations of dependence) are similar to earlier reports.1,11,12 This age and dependence and duration-related vulnerability may come about as a result of greater physical comorbidity in older patients13 or development of pro-epileptic kindling mechanisms.14
The prevalence of ES in the opioid cohort was higher than expected. Putative causal pathways for opioid-induced ES include effects of opioids and contaminants on opiate and non-opiate receptor-dependent epileptogenesis,15 or a synergistic action of the opioid, contaminants and comorbidities.3 Most ES-positive patients in the opioid-abusing cohort were in the third decade of life, which is when unprovoked seizures occur commonly in the general adult population.16 Thus, opioid-induced epileptogenesis is probably distinct from that of alcohol-induced epileptogenesis, in which increasing age is important. The relationship of ES to intoxication or withdrawal may reflect interactions between levels of opioids in the body, their pro- and anticonvulsant effects, effects of contaminants and individual propensity.
Certain limitations were inevitable. The diagnoses generated were clinical and not supported by investigations such as electroencephalography or neuroimaging for determination of etiology. Therefore, patients with seizures as a result of infection, trauma and metabolic imbalance may have been included. The conclusions derived are thus suggestive rather than confirmatory. We also did not have a control group with which to compare the findings. The present findings warrant further investigation in larger both-gender multi-location samples.