Absence of carbamazepine-induced hyponatremia among patients also given lithium

Of 33 chronically psychotic patients in a state hospital, 17 received carbamazepine, 13 received carbamazepine and lithium, and three received carbamazepine and then the combination. There was a significant difference in serum sodium level between the patients receiving carbamazepine alone (mean +/- SD = 138.4 +/- 4.3 meq/liter) and those also receiving lithium (141.8 +/- 1.6 meq/liter). (A similar difference was seen for the patients who received the two treatments serially.) Age, sex, diagnosis, age at diagnosis, seizure disorder, antipsychotic drugs, and serum carbamazepine level did not explain this difference. The protection against hyponatremia provided by the carbamazepine-lithium combination occurred despite lithium's tendency to increase polyuria.     

Effect of abrupt change from standard to low serum levels of lithium: a reanalysis of double-blind lithium maintenance data

OBJECTIVE: Growing evidence suggests that abrupt lithium discontinuation increases the risk of recurrence for patients with bipolar disorder. To assess the effect of abrupt change in lithium dose, the authors reanalyzed data from a previously reported, randomized, double-blind trial of standard- versus low-dose lithium for maintenance therapy in bipolar disorder. METHOD: In the original study, serum lithium levels were obtained during a 2-month open stabilization period for 94 patients with bipolar disorder who were then randomly assigned to be maintained on a low (serum level=0.4-0.6 meq/liter) or a standard (0.8-1.0 meq/liter) level of lithium therapy. Patients were then followed for up to 182 weeks. This reanalysis examined the potential confounding influence of prerandomization lithium level and change in lithium level on the outcome of subjects assigned to a standard or low maintenance dose of lithium. RESULTS: In a Cox proportional hazards model incorporating pre- and postrandomization lithium levels and the interaction of these factors, only the interaction term remained significantly associated with time to recurrence. CONCLUSIONS: The findings indicate that change in serum lithium level may be a more powerful predictor of recurrence of bipolar disorder than the absolute assignment to a low or a standard dose of lithium and suggest that an abrupt decrease in lithium level should be avoided. This reanalysis did not directly address optimal maintenance lithium levels but raises questions about the original study's finding of superiority for lithium levels > or =0.8 meq/liter. The results underscore the importance of accounting for the possible confounding effects of changes in the intensity of pharmacotherapy in studies of maintenance therapies for bipolar disorder.     

Double-blind, placebo-controlled comparison of imipramine and paroxetine in the treatment of bipolar depression

OBJECTIVE: This study compared the efficacy and safety of paroxetine and imipramine with that of placebo in the treatment of bipolar depression in adult outpatients stabilized on a regimen of lithium. METHOD: In a double-blind, placebo-controlled study, 117 outpatients with DSM-III-R bipolar disorder, depressive phase, were randomly assigned to treatment with paroxetine (N=35), imipramine (N=39), or placebo (N=43) for 10 weeks. In addition to lithium monotherapy, patients may have received either carbamazepine or valproate in combination with lithium for control of manic symptoms. Patients were stratified on the basis of trough serum lithium levels determined at the screening visit (high: >0.8 meq/liter; low: 相似文献   

Effects of drugs on the initiation and maintenance of status epilepticus induced by administration of pilocarpine to lithium-pretreated rats

The ability of various drugs to prevent the onset of status epilepticus induced by administration of the muscarinic agonist, pilocarpine, to lithium-pretreated rats was determined. Motor limbic seizures and status epilepticus occurred in 100% of rats administered pilocarpine (30 mg/kg, s.c.) 20 h after pretreatment with lithium (3 meq/kg, i.p.). The latency to spike activity and to status epilepticus was 20 +/- 1 min and 24 +/- 1 min, respectively. Atropine, diazepam, phenytoin, carbamazepine, phenobarbital, paraldehyde, and L-phenylisopropyladenosine (L-PIA) prevented all phases of seizure activity induced by lithium/pilocarpine treatment. The initiation of status epilepticus was significantly prolonged by pretreatment with sodium valproate. These findings indicate that the seizures induced by administration of lithium and pilocarpine accurately model generalized tonic-clonic epilepsy. The anticonvulsant activity of L-PIA was prevented by prior treatment with the adenosine antagonist, theophylline. The latency to spike and seizure activity was decreased by theophylline, indicating that endogenous adenosine may have a tonic inhibitory influence on cholinergic neurons. Atropine, diazepam, phenobarbital, phenytoin, sodium valproate, L-PIA, and carbamazepine did not interrupt seizure activity when administered 60 min after pilocarpine (approximately 35 min after initiation of status epilepticus). When rats were administered paraldehyde at this time, status epilepticus was rapidly terminated and all rats survived. Thus, status epilepticus induced by lithium and pilocarpine provides a seizure model that is not responsive to conventional anticonvulsants.     

Differential effects of acute and short-term lithium administration on dialysate glutamate and GABA levels in the frontal cortex of the conscious rat

In the present study, we employed in vivo microdialysis in the frontal cortex of the awake rat to investigate the effects of acute and short-term (twice daily, 3 days) lithium chloride administration (1, 2, and 4 meq/kg, s.c.) on local dialysate glutamate and GABA levels. Acute lithium (1 meq/kg) failed to influence cortical glutamate levels while the higher (2 and 4 meq/kg) doses increased (+38 +/- 6% of basal levels) and reduced (-27 +/- 4%) cortical glutamate levels, respectively. Cortical GABA levels were affected by acute lithium only at the highest 4 meq/kg dose (+62 +/- 6%). Furthermore, these effects were prevented by tetrodotoxin (1 microM) and low-calcium (0.2 mM) medium perfusion. Following short-term administration, lithium increased (+58 +/- 4%) cortical dialysate glutamate levels at the 1 meq/kg dose, was ineffective at 2 meq/kg, while the effect of the 4 meq/kg dose was similar to that observed after acute administration. Interestingly, intracortical perfusion with the GABA(B) receptor antagonist CGP 35348 (100 microM) reversed the acute lithium (4 meq/kg)-induced decrease in glutamate levels. Taken together, these findings indicate a differential dose and duration dependent effect of lithium on cortical dialysate glutamate levels involving both a direct enhancement and an indirect inhibition that is mediated via an activation of local GABA(B) receptor. These findings may be relevant for the therapeutic effects of the drug.     

The effects of lithium carbonate on healthy volunteers: mood stabilization?

A 2-month lithium-placebo double-blind cross-over study was carried out with 17 healthy volunteers. Their mood was self-rated: twice daily (AM, PM) with the Visual Analogue Mood Scale (VAMS); weekly with the analogue scales for subjective states and body symptoms; and three times (basal and at the end of each treatment period) with the Profile of Mood States (POMS). Memory and reaction time were also assessed, but did not show any change. The mean VAMS score decreased during lithium treatment, but the mean mood variability, a measure of the mean successive differences between consecutive mood ratings (delta squared), did not change significantly. There was a tendency toward decreased mood variability on lithium, both during the full 1-month treatment period and in the last week of treatment, when all volunteers had a lithium serum level ranging from 0.6 to 1.0 mEq/liter. The lower mean VAMS scores on lithium could be attributed to lithium-induced dysphoric mood as recorded on the analogue scales and POMS. However, very large inter- and intraindividual differences in response to lithium were observed. Actually, lithium even had an opposite effect on some volunteers' mood. The data and problems involved with assessment of mood and its changes are discussed.     

Salivary free concentrations of anti-epileptic drugs: an evaluation in a routine clinical setting

OBJECTIVE: This study was aimed at correlating the salivary and serum free concentrations of anti-epileptic drugs (carbamazepine, phenytoin and sodium valproate) in a population of neurological patients in a routine clinical setting. METHOD: Twenty-seven paired serum/saliva specimens from 22 patients: 10 for carbamazepine, 8 for phenytoin and 9 for sodium valproate were obtained to study these correlations. Salivary and serum free concentrations of anti-epileptic drugs, anti-epileptic drug dosing history, and associated information were collected prospectively. The salivary and serum free concentrations of the anti-epileptic drugs were simultaneously quantified using fluorescence polarization immunoassay (TDx analyzer). RESULTS: For both carbamazepine and phenytoin there was a strong correlation between the salivary and serum free concentrations, 0.99 and 0.98, respectively. The mean ratio of salivary to serum free carbamazepine concentration was 1.02 +/- 0.11 and 0.82 +/- 0.15 for phenytoin. A poor correlation between salivary and serum free concentration was observed for sodium valproate (0.70) with a mean ratio of salivary to serum free concentration of 0.48 +/- 0.27. CONCLUSION: Monitoring of free salivary concentrations of anti-epileptic drugs, particularly phenytoin and carbamazepine proved to be a realistic alternative in this routine clinical setting.     

Lithium placental passage and obstetrical outcome: implications for clinical management during late pregnancy

OBJECTIVE: Lithium has been used during pregnancy for more than four decades, but quantification of fetal lithium exposure and clinical correlations of such exposure are limited. The study objectives were to 1) quantify the rate of lithium placental passage, 2) assess any association between plasma concentration of lithium at delivery and adverse perinatal events, and 3) determine whether lithium concentrations can be reduced by briefly suspending therapy proximate to delivery. METHOD: Maternal blood and umbilical cord blood were obtained at delivery for assay of lithium concentrations, and obstetrical outcome data were collected prospectively for 10 participants. These data were combined with results from MEDLINE and PsycINFO searches that identified 32 cases in which maternal lithium was administered throughout delivery. Statistical analysis of the pooled data was conducted. RESULTS: The ratio of lithium concentrations in umbilical cord blood to maternal blood (mean=1.05, SD=0.13) was uniform across a wide range of maternal concentrations (0.2-2.6 meq/liter). Significantly lower Apgar scores, longer hospital stays, and higher rates of CNS and neuromuscular complications were observed in infants with higher lithium concentrations (>0.64 meq/liter) at delivery. Withholding lithium therapy for 24-48 hours before delivery resulted in a 0.28 meq/liter reduction in maternal lithium concentration. CONCLUSIONS: Lithium completely equilibrates across the placenta. Higher lithium concentrations at delivery are associated with more perinatal complications, and lithium concentrations can be reduced by brief suspension of therapy proximate to delivery. Treatment guidelines are proposed to improve neonatal well-being when lithium use is indicated in late pregnancy.     

Lithium lengthens circadian period in a diurnal primate, Saimiri sciureus

Lithium lengthens the period of free-running circadian rhythms in a variety of species, but this effect has not been demonstrated unequivocally in primates. Because of the possible link between lithium's action on the circadian clock and its therapeutic action in human mood disorders, we tested the ability of lithium to lengthen circadian period in a diurnal primate with circadian properties similar to those of humans. Lithium carbonate was administered in food pellets to 8 adult male squirrel monkeys (Saimiri sciureus) for at least 27 consecutive days. Serum lithium levels on the last day of lithium administration ranged from 0.76 to 2.02 mEq/liter, comparable to the therapeutic range for treatment of bipolar disorder in humans (0.6-1.2 mEq/liter). Circadian periods of perch-hopping activity were longer during lithium treatment than during baseline in 7 of the 8 monkeys (changes of -0.08 to +1.41 hr, mean +0.55 hr, p = 0.01), and returned toward baseline values when lithium was discontinued. In most cases, the period change was evident within a few days after beginning full lithium dose, and was not accompanied by changes in level or pattern of activity, nor in amplitude of the circadian rhythm. Food consumption and body weight were reduced during lithium treatment, and rebounded on return to lithium-free diet. Period change was related to lithium dose (p less than 0.05), but did not correlate with food consumption, body weight, or baseline circadian period. These results, by establishing that lithium lengthens circadian period in primates, suggest that studying the cellular mechanisms of this circadian effect may be relevant to understanding lithium's therapeutic effect on mood in humans.     

Diurnal variation in water homeostasis among schizophrenic patients subject to water intoxication

Among seven schizophrenic patients subject to water intoxication (six men and one woman, mean age 39.1 +/- 6.9 years), we measured serum sodium, plasma arginine vasopressin, and urine osmolality at 7 a.m. and 4 p.m. on eight consecutive Thursdays. On the days of greatest diurnal change in serum sodium, the 7 a.m. serum sodium was 141.1 +/- 1.8 mEq/l and the 4 p.m. value was 129.9 +/- 3.2 mEq/l. Plasma vasopressin also tended to be lower at 4 p.m. but, in many cases, was inadequately suppressed for the level of hyponatremia. The urine was dilute at both 7 a.m. and 4 p.m. and mean urine osmolality did not differ at the two times. In three patients, urine osmolality was consistently subnormal relative to plasma vasopressin at both 7 a.m. and 4 p.m. This abnormality was consistent with nephrogenic diabetes insipidus secondary to lithium and, possibly, phenytoin which the patients received to protect them against hyponatremia. We conclude that the combination of polydipsia and abnormal osmoregulation of vasopressin secretion contributes importantly to the afternoon hyponatremia found in schizophrenic patients subject to water intoxication.     

Cognitive functions, epileptic syndromes and antiepileptic drugs.

Cognitive function of patients on monotherapy specific for their epileptic syndrome has been studied infrequently. We evaluated 7 patients with symptomatic localised epilepsies (SEL) on phenytoin aged 30 +/- 12 (mean +/- standard deviation) years, 8 with idiopathic generalised epilepsies on sodium valproate aged 18 +/- 4 years, 16 with SEL on carbamazepine aged 28 +/- 11 years, and 35 healthy controls aged 27 +/- 11 years. All subjects were of normal intelligence, educated appropriately to age, and led productive lives in the community. Two of the patients on carbamazepine and one on valproate had less than five partial, absence or myoclonic seizures monthly, the remaining were controlled. Carbamazepine serum concentrations were 12 +/- 5 micrograms/ml, phenytoin were 23 +/- 7, and valproate were 62 +/- 23 (mean +/- sd). Tests included immediate recall and recognition for pictures, Stroop test, delayed recall and recognition of pictures. Patients on phenytoin and valproate performed significantly worse than controls on immediate recall, and patients on carbamazepine performed significantly worse than controls in Stroop test (p < 0.01). The results indicate relatively minor effects of the epileptic syndromes and of phenytoin, carbamazepine and valproate on cognition of patients with controlled epilepsy leading productive lives in the community. We conclude that the cognitive deficit found in chronic epileptic patients on poly-therapeutic drug regimen must be multifactorial, and that future studies need to control for all possible variables in order to achieve meaningful results.     

Effect of Felbamate on Phenytoin and Carbamazepine Serum Concentrations

Felbamate (FBM) is a novel antiepileptic drug (AED) undergoing clinical trials in the United States. During a double-blind, cross-over clinical trial, patients received concomitant phenytoin (PHT) and carbamazepine (CBZ). Dosages of PHT and CBZ were adjusted to maintain serum concentrations +/- 20 and +/- 25% of baseline values. All patients required a PHT dosage decrease of 10-30% during active FBM treatment to maintain stable concentrations. CBZ serum concentrations decreased significantly in patients receiving active FBM. The mean decrease was 1.3 micrograms/ml and occurred in 30 of 32 patients. Therefore, FBM apparently causes a bidirectional effect on the serum concentrations of PHT and CBZ when all three drugs are taken concomitantly.     

Polydipsia, hyponatremia, and seizures in psychotic patients.

Case histories are presented for four psychotic patients who ingested large quantities of water and subsequently developed grand mal seizures and serum sodium levels of less than 121 meq/liter. The physiology of psychogenic polydipsia and related disorders is reviewed. The relation of this disorder to temporal lobe seizures and to the use of phenothiazines is considered.     

Lack of an effect of topiramate on lamotrigine serum concentrations

PURPOSE: Pharmacokinetic interactions between the older antiepileptic drugs (AEDs) and topiramate (TPM) were assessed during the clinical development of this drug. Lamotrigine (LTG) has become established as an important new drug in treating a wide spectrum of seizure types, but there are no published data on whether LTG serum concentrations change when TPM is added to treatment. METHODS: Escalating doses of TPM were added to stable LTG treatment in 24 young patients (8-21 years) with epilepsy. Blood samples taken before the morning dose were collected for drug-concentration measurement in all patients before starting treatment with TPM and after stabilisation at each dose escalation. Several patients had been maintained on unchanged therapy with drug-concentration monitoring for many months before introducing TPM, and a sequence of baseline LTG serum concentrations were available on these patients. RESULTS: The mean of all baseline LTG concentrations for the group as a whole was 10.4 +/- 4.4 mg/L compared with 9.7 +/- 4.3 mg/L after addition of TPM. A comparison of last baseline LTG concentration with first test LTG concentration (i.e., after 2 weeks' TPM treatment) gave mean values of 10.7 +/- 4.7 and 10.8 +/- 4.6 mg/L, respectively. The mean LTG concentration for patients while taking their highest TPM dose was 9.5 +/- 4.3 mg/L. The analysis-of-variance modeling for the effect of TPM on LTG concentration yielded a mean LTG concentration ratio (with TPM vs. without TPM) of 94.2%, with a 90% confidence interval of 89.5-99.1%. CONCLUSIONS: TPM did not cause a significant change in LTG serum concentration in this group of patients.     

Adjunctive quetiapine in bipolar patients partially responsive to lithium or valproate

Despite therapeutic treatment with lithium or valproate, patients with bipolar I disorder often require adjunctive therapy to treat persistent symptoms. In order to evaluate the effects of quetiapine for bipolar symptoms inadequately responsive to mood stabilizers, a retrospective chart review was undertaken at the Veterans Affairs (VA) Long Beach Mood Disorders Clinic for all bipolar I outpatients who had been prescribed adjunctive quetiapine during an 18-month study period. Among 75 lithium- or valproate-treated patients receiving quetiapine, 16 were identified in whom therapeutic treatment with lithium (> or =0.8 meq/l) or valproate (> or =75 microg/ml) could be verified during the 2-month period prior to quetiapine initiation. Chart notes were utilized by the principal investigator to assign Clinical Global Impression Bipolar ratings (CGI-BP) before and after 30-120 days of quetiapine treatment (mean=173+/-157 mg). Nine of 16 lithium- or valproate-stabilized bipolar patients (56%) were judged much or very much improved in CGI-BP overall ratings following adjunctive quetiapine. In addition, for the entire sample, quetiapine augmentation resulted in significant improvements in clinician-rated bipolar severity scores (CGI-BP) for mania, depression, and overall bipolar illness. The majority of quetiapine-related symptomatic improvement was due to diminished insomnia, agitation, irritability, and mood disturbance. Side effects were mild and transitory including sedation and dizziness. Low-dose quetiapine may be a useful and well-tolerated adjunct for some bipolar patients with incomplete response to lithium or valproate.     

Diurnal weight gain as a predictor of serum sodium concentration in patients with psychosis, intermittent hyponatremia, and polydipsia (PIP syndrome)

Ten male patients (mean age 37.3 +/- 6.4 years) with psychosis, intermittent hyponatremia, and polydipsia (PIP syndrome) underwent measurement of weight, sitting and standing blood pressure, and serum sodium concentration at 7 a.m. and 4 p.m. weekly for 8 consecutive weeks. Blood pressure was higher in the afternoon than in the morning. The diurnal decrease in serum sodium (141.4 +/- 2.8 to 134.2 +/- 4.8 mEq/l) was associated with a diurnal increase in weight (78.4 +/- 9.7 to 80.0 +/- 10.3 kg). When the weight increase was normalized by dividing by 7 a.m. weight (NDWG), the following relationship evolved: diurnal serum sodium decrease = 3.060 + [201.728 x NDWG]. Therefore, NDWG accounted for 63.1% of the variability of serum sodium. Using the known relationship of plasma water, total body water, and total body weight, we calculated that antidiuresis (afternoon weight gain) accounted for 62.5% of afternoon hyponatremia. Thus, two separate methods of calculating the relationship between antidiuresis and hyponatremia provided remarkably similar findings. We derived a table to predict 4 p.m. serum sodium values based on 7 a.m. weight, 7 a.m. serum sodium, and 4 p.m. weight.     

Lithium enhancement of central 5-HT transmission induced by 5-HT precursors

The effects of acute and chronic lithium chloride administration on synaptic transmission between bulbospinal norepinephrine (NE) or 5-hydroxy-tryptamine (5-HT) pathways and sympathetic preganglionic neurons were tested in unanesthetized, spinal cats. Discharges recorded from sympathetic preganglionic white rami were evoked by stimulation of spinal reflex pathways or descending excitatory pathways in the cervical spinal cord. Acute lithium administration (2 meq/kg) produced insignificant depression of the reflex pathway but markedly depressed transmission through the intraspinal pathway, an effect that was prevented by depletion or blockage of 5-HT. These observations and the failure of lithium to alter the typical effects of L-dopa on both pathways indicate that lithium does not affect transmission through the excitatory NE pathway. L-Tryptophan (1,0 mg/kg) alone produced little or no depression of either pathway, but 3--4 hr after lithium, this dose of L-tryptophan gradually depressed transmission through both pathways by about 20%. After chronic lithium pretreatment (1 meq/kg twice a day for 3 days), L-tryptophan rapidly depressed transmission through spinal reflex and intraspinal pathways by 40% and 50% respectively. Chronic lithium pretreatment also more than doubled the depression of transmission through both pathways produced by 30 mg/kg of 5-HTP. The average of plasma lithium levels 8--10 hr after the last chronic dose was 1.5 meq/liter. These results support the proposal that lithium increases the uptake of L-tryptophan and 5-HTP by central 5-HT terminals and thereby enhances 5-HT synthesis which is reflected in increased transmission at central 5-HT synapses.     

Lithium in breast milk and nursing infants: clinical implications

OBJECTIVE: Current practice guidelines discourage use of lithium during breast-feeding, despite limited data. This study aimed to quantify lithium exposure in nursing infants. METHOD: In 10 mother-infant pairs, the authors obtained assays of lithium in maternal serum, breast milk, and infant serum and indices of infant renal and thyroid function. RESULTS: Maternal serum, breast milk, and infant serum daily trough concentrations of lithium averaged 0.76, 0.35, and 0.16 meq/liter, respectively, each lithium level lower than the preceding level by approximately one-half. No serious adverse events were observed, and elevations of thyroid-stimulating hormone, blood urea nitrogen, and creatinine were few, minor, and transient. CONCLUSIONS: Serum lithium levels in nursing infants were low and well tolerated. No significant adverse clinical or behavioral effects in the infants were noted. These findings encourage reassessment of recommendations against lithium during breast-feeding and underscore the importance of close clinical monitoring of nursing infants.     

Effect of indomethacin on edema following single and repetitive cerebral ischemia in the gerbil

BACKGROUND AND PURPOSE: Repetitive periods of cerebral ischemia result in more severe injury than a single period of ischemia of similar total duration. We investigated the possibility of prostaglandin mediation of this increased injury by attempting to modify brain edema formation with indomethacin pretreatment. METHODS: Under halothane/N2O anesthesia, groups of gerbils underwent bilateral carotid occlusion to induce forebrain ischemia. Group I underwent a single 15-minute period of carotid occlusion. Group II underwent three 5-minute periods of occlusion at hourly intervals. Groups III and IV were similar to groups I and II, respectively, but received 0.2 mg/kg indomethacin before carotid occlusion. Cortical and cerebellar water and sodium contents were determined in control animals (n = 6) at time zero and in experimental animals 24, 48, and 72 hours after ischemia (n = 6-10 gerbils/group at each time point). RESULTS: Cortical water and sodium contents in group II peaked 48 hours after insult (82.15 +/- 0.31% and 420 +/- 14 meq/kg dry wt, respectively) and were significantly higher than control and group I values at both 24 and 48 hours. Cortical water did not change from control in group I animals. Indomethacin pretreatment significantly attenuated increases in water and sodium content seen at 48 hours in gerbils undergoing repetitive ischemia (peak 80.02 +/- 0.45% and 300 +/- 39 meq/kg dry wt), but did not affect mortality. CONCLUSIONS: Indomethacin lessens edema after repetitive cerebral ischemia, suggesting that elevations of cyclooxygenase products are responsible, at least in part, for severe brain edema following repetitive ischemia.     

Increase of low serum concentrations of high-density lipoprotein (HDL) cholesterol in TIA-patients treated with phenytoin

Serum high density lipoprotein (HDL) cholesterol and other lipoproteins were measured in 27 TIA-patients with a mean age of 49 +/- 10 years before and during phenytoin therapy. The pretreatment concentrations of HDL-cholesterol (mmol/l, mean +/- SD) were lower (p less than 0.001) in male (1.03 +/- 0.25) and in female patients (1.15 +/- 0.44) than in healthy male (1.28 +/- 0.34) and female controls (1.52 +/- 0.31) respectively. After one month's phenytoin therapy HDL cholesterol concentrations reached normal levels (men 1.33 +/- 0.38, women 1.61 +/- 0.27) and after 9 months of therapy even surpassed them (men 1.47 +/- 0.27, p less than 0.05; women 1.91 +/- 0.33, p less than 0.01). Percent increase of HDL cholesterol after 9 months of therapy was 42 +/- 25 in men and 68 +/- 46 in women. There was a positive correlation (r = 0.43, p less than 0.05) between serum phenytoin level and increase of HDL cholesterol. HDL/LDL cholesterol ratio increased (p less than 0.01) also during 9 months of therapy (men from 0.26 +/- 0.05 to 0.36 +/- 0.10, women from 0.26 +/- 0.07 to 0.43 +/- 0.13) and showed a positive correlation (r = 0.91, p less than 0.001) with increase of serum HDL cholesterol. The HDL cholesterol levels achieved have been maintained with a mean serum phenytoin level of 5.6 +/- 3.6 mg/l. Phenytoin induced increase in serum HDL levels should not yet be equated with protection against atherosclerosis.