Changes in platelet function and excessive bruising were investigated by chart review and prospective screening in 51 patients treated with the ketogenic diet for epilepsy at RUSH-Presbyterian-St Luke’s Medical Center, Chicago, IL A significant increase in bruising or other minor bleeding was reported and/or observed in 16 (31%) patients. The complication was more frequent in younger patients but was independent of sex and number of concurrent anticonvulsants (AED). The average age of patients with bleeding was 5 years compared to 8 years for patients without bleeding. The possibility of an interaction with a specific AED was suggested by a 25% use of lamotrigine in patients with bruises compared to 8% use in the nonbruising group (not significant p=.ll). Five of 6 patients tested had prolonged bleeding times and all had diminished response to platelet aggregating agents. One had a latent von Willebrand disease. The mechanism of the diet-induced bruising may be complex, involving interaction between the diet and individual platelet dysfunction. A possible bleeding tendency should be evaluated in patients on the ketogenic diet who are candidates for surgery or anticoagulant therapy. 
COMMENT. Despite the absence of serious bleeding in this series of patients treated with the ketogenic diet, a 30% incidence of diet-induced bruising deserves further study and evaluation. A possible interaction with lamotrigine is suggested in some patients receiving concurrent drug and diet.
This is not the first observation of platelet dysfunction and anemia as a complication of the ketogenic diet. Complications in 10% of 52 children treated by Ballaban-Gil et al, 1998 (see Ped Neur Briefs August 1998;12:60) included thrombocytopenia and hemolytic anemia. Valproate interaction could not be excluded in 29 (56%). The proportion of ketogenic/antiketogenic foods was 4:1 in this study but was not specified in the above Pres St Luke’s study. This report of diet-induced bruising is another reason to endorse the Mayo Clinic method of slow initiation of the diet with lower ratios, in place of the Hopkins recommended ratio of 4:1. Using the Mayo Clinic method, I have not encountered this or other serious complication as reported with the Hopkins regimen (Ped Neur Briefs 1998; 12:61).
A fat-overload syndrome with neurologic complications is reported in 2 children receiving fat emulsion therapy. Both patients died and autopsy showed cerebral intravascular lipid deposition and areas of necrosis and hemorrhage. (Progress in Pediatric Neurology III, PNB Publ, 1997;p98) . A rapid rise in triglyceride levels was invoked as a factor in this complication.