Ketogenic Ratio, Calories and Fluids: Do They Matter?

Summary

The ketogenic diet (KD) traditionally was initiated using specified ketogenic ratios, limited calories and fluids. Recent work has shown that lower ketogenic ratio diets are frequently as efficacious as higher ones and result in fewer adverse effects. In animals, calorie restriction is anticonvulsant. In children however, the need for calorie restriction is less clear, but avoidance of excessive calories may improve efficacy of the diet. There is no evidence that fluid restriction is a necessary component of the KD. Given the higher risk of nephrolithiasis, adequate fluid intake should be encouraged.

The ketogenic diet (KD) has been utilized for nearly a century to treat refractory seizures. The original description, published in 1921 from the Mayo Clinic, is similar to the traditional KD used today (Wilder, 1921). This diet was initiated with a fast, used specified ketogenic ratio calculations, emphasized monitoring of ketosis, limited fluids and restricted calories. However, more recently reported lower ketogenic ratio variants (Kossoff et al. 2003, Pfeifer and Thiele 2005) and less aggressive approaches to fluid or caloric restriction now raise questions regarding the importance of these factors. Over the last decade, work done in animal models to elucidate mechanisms of action of the KD has provided further information on ketogenic ratio, fluid and calorie restriction.

Ketogenic Ratio

The ketogenic ratio is defined as the ratio of grams of fat to grams of carbohydrate plus protein. Higher ratios result in greater degrees of ketosis. Traditionally, the KD has been calculated based on specific ratios, based on patient age. Infants and teens were generally started on a 3:1 ratio and other children on a 4:1 ratio. The dietary ratio was adjusted to maintain urinary ketones in the moderate to high range (80-160 mg/dL). While higher ratios may confer better seizure control, they may also result in poorer tolerability of the diet (Nylen et al. 2005).

A survey of worldwide use of the KD has shown that centers in India and Asia use lower ratios with good success (Kossoff and McGrogan JR, 2005). Furthermore, “newer” variations of the diet, such as the modified Atkins diet and the low glycemic index diet have significantly lower ratios, yet similar efficacy to the traditional diet (Kossoff et al. 2003, Pfeifer and Thiele, 2005).

Studies have evaluated the association between higher ketogenic ratios and improved seizure control. In animals, higher ratios correlate with greater efficacy. Bough fed groups of rats KDs with ratios varying from 1:1 to 9:1 (Bough et al. 2000). All diets were calorie-restricted to approximately 90% of the normal daily requirement. Animals were maintained on the diet from P37 to P57-58, when testing to determine seizure susceptibility to pentylenetetrazole (PTZ)-induced seizures was performed. Weight gain and degree of ketosis was assessed for each group. Higher ratios correlated significantly with poorer weight gain and higher mean B-hydroxybutyrate levels (p<0.05 for both). The efficacy was significantly greater for animals fed diets exceeding a 6:1 ratio, compared to those fed 4:1 or 5:1 ratios (p=0.009 and p=0.02) as evidenced by greater resistance to seizures.

Seo compared efficacy and tolerability of a 3:1 and 4:1 diet in 76 children with intractable epilepsy (Seo et al. 2007). Groups were comparable for age, gender and seizure frequency, but the 3:1 group had more patients with partial seizures. Efficacy was higher with a 4:1 ratio (p<0.05) - at 3 months, 55% in the 4:1 group versus 31% in the 3:1 group were seizure-free. Ten of 22 children who were seizure-free on the 4:1 diet at 3 months switched to a 3:1 diet and all remained seizure-free. Twelve of 22 children who were not seizure-free on a 3:1 diet at 3 months switched to the 4:1 diet; while 75% experienced a further reduction in seizures, none became seizure-free. Interestingly, no significant differences in ketone levels were found between the 4:1 and 3:1 groups. Regarding tolerability, children experienced significantly less gastrointestinal intolerance on the 3:1 compared to the 4:1 ratio (p<0.05), although there was no significant difference in maintenance rates between the groups. The authors recommended starting a higher ratio diet, and decreasing the ratio in those with significant gastrointestinal intolerance.

The modified Atkins diet has no restrictions on protein, calories or fluids, and may be an option for patients unable to tolerate the more restrictive traditional KD. Twenty children, aged 3-8 years were prospectively treated for six months with this diet (Kossoff et al. 2006). Carbohydrates were initially limited to 10 grams per day and fats were encouraged. At six months, 80% remained on the diet, 65% experienced a >50% reduction and 35% a greater than 90% reduction in seizures. These results were comparable to a large prospective study of the traditional KD, in which 51% had a >50% and 32% had a greater than 90% reduction in seizures (Freeman et al. 1998). All children on the modified Atkins diet achieved moderate ketosis within 4 days of diet initiation. While only 29% maintained large ketosis long-term, 80% who lost large urinary ketosis did not lose seizure control, suggesting that degree of ketosis may be less important than previously thought.

While the modified Atkins diet appears equally efficacious to the traditional KD, the ideal starting carbohydrate limit was not known. Kossoff performed a randomized, crossover comparison of daily carbohydrate limits in 20 children (randomized to either 10 or 20 grams/day) (Kossoff et al. 2007). At three months, 60% of the 10 gram versus only 10% of the 20 gram group achieved a greater than 50% reduction in seizures (P=0.03). Ketone levels and ketogenic ratios were not significantly higher in the 10 gram group and neither was predictive of efficacy. Sixteen children crossed over to the opposite arm at 3 months - 82% experienced no change in seizure frequency. Overall, the majority (82%) felt that the 20 gram diet was more tolerable. In conclusion, greater carbohydrate restriction appeared to improve efficacy early on, however carbohydrates could be liberalized at three months without worsening of seizures.

An even less restrictive form of the KD, the low glycemic index treatment (LGIT), was utilized in 20 patients with refractory epilepsy (Pfeifer and Thiele, 2005). Carbohydrates with a glycemic index <50, relative to glucose were allowed, and total carbohydrate was restricted to 40-60 grams per day. Eleven patients were started de novo, eight (73%) of whom experienced a greater than 50% reduction in seizures, with 4 (36%) achieving seizure freedom. Nine patients were switched to the LGIT because of inability to tolerate a traditional KD - only 2 (22%) experienced a worsening of seizures.

The above work suggests that many patients achieve seizure control with lower ketogenic ratios than were traditionally used, with improved tolerability of the diet. However, in a minority, higher ratios result in better efficacy.

Calorie Restriction

Traditionally, calorie restriction was felt to be an integral part of the ketogenic diet with calories restricted to approximately 75% of daily requirements. The modified Atkins or LGI diets do not mandate calorie restriction, and even when using a traditional KD, most centers no longer restrict calories to this degree.

Calorie restriction alone reduces seizures in animals. In a genetic mouse model of multifactorial idiopathic epilepsy, animals fed calorie-restricted diets showed delayed onset and reduced incidence of seizures compared to those fed ad libitum (Greene et al. 2001). Furthermore, in juvenile animals, a 15% calorie-restricted diet of regular chow was more effective than a ketogenic diet fed ad libitum in delaying seizure onset and reducing seizure susceptibility. To determine the effects of calorie restriction, ketosis and carbohydrate intake, Eagles compared seizure susceptibility to PTZ at P57 in rats fed a high carbohydrate diet that was calorie-restricted to 90, 65 or 50% to those fed a standard KD that was calorie-restricted to 90% (Eagles et al. 2003). Seizure threshold was elevated in proportion to calorie restriction, and animals fed a high carbohydrate diet calorie-restricted to 50% had thresholds similar to those fed a KD calorie-restricted to 90%, suggesting that calorie restriction alone has a beneficial anticonvulsant effect.

In a study designed to examine the anticonvulsant and antiepileptic effects of the KD, Bough (2003) studied both network excitability and kindling in the dentate gyrus of animals fed one of three diets: (1) ketogenic 80-90% calorie-restricted, (2) normal 80-90% calorie-restricted, and (3) normal ad libitum. Decreased network excitability, as manifested by greater paired pulse inhibition, elevated maximal dentate activation thresholds and an absence of spreading depression-like events was seen in both calorierestricted groups. However, only animals fed the ketogenic calorie-restricted diet showed resistance to kindling, manifested by a reduced rate of increase in electrographic seizure duration after repeated stimuli. These results confirm the anticonvulsant effect of calorie restriction, but also suggest that the KD may have an additive anti-epileptogenic action.

Several mechanisms have been suggested to explain the anticonvulsant action of calorie restriction. Calorie restriction results in increased glutamic acid decarboxylase-65 and 67 expression, enhancing conversion of glutamate to GABA, thus diminishing excitation (Cheng et al. 2004). Limitation of glucose also activates K_ATP channels in the central nervous system, which lead to membrane hyperpolarization, making cells less excitable (Schwartzkroin 1999).

In humans, no study to date has shown a benefit of calorie restriction. While excessive weight gain is perceived to correlate with poorer efficacy, no link was found between either ideal body mass index or change in body mass index over time and seizure control in children treated with the KD (Hamdy et al. 2007). However, in adults starting the Atkins diet, efficacy appeared greatest in those who lost weight (Kossoff et al. 2008).

Fluid Restriction

Traditionally, fluids have been restricted to 80-90% of daily requirements. Early studies from the 1920's and 1930's suggested that tissue hydration was one of the mechanisms by which the KD worked, and created a perception that overhydration reduces efficacy.

The KD may predispose to nephrolithiasis because of hypercalciuria, acid urine, low urinary citrate and low fluid intake. Overall approximately 2-4% of patients treated with a traditional diet develop stones; those with hypercalciuria tend to be a higher risk (Sampath et al, 2007). While concerns have been raised that concurrent use of carbonic anhydrase inhibitors such as topiramate, zonisamide or acetazolamide may exacerbate stone formation in children on the diet, a recent study refuted this theory (Sampath et al. 2007).

There is no scientific evidence to suggest that fluid restriction is needed or beneficial. Because of concerns of possible nephrolithiasis, most centers no longer restrict fluids.

Summary

Lower ketogenic ratios are frequently as effective as higher ones at controlling seizures, and result in fewer adverse effects. However, a minority of patients experience improved seizure control at higher ratios. There is evidence, both from studies on the traditional (Seo et al. 2007) and the modified Atkins diet (Kossoff et al. 2007) to suggest that starting at higher ratios may result in better control, but that ratios can often be weaned over time without deterioration if efficacy.

In animals, calorie restriction has an independent anticonvulsant effect, above simply increasing ketosis. In children, the need for calorie restriction is less clear. While avoidance of too many calories may improve efficacy, restriction to 75% of daily requirements is probably excessive.

There is no evidence that fluid restriction is a necessary component of the ketogenic diet. Given the higher risk of nephrolithiasis, adequate fluid intake should be encouraged.