intracranial pressure, suggesting that nutrition was most effective in those likely to have the greatest injury response, with attenuation of that response a potential mechanism (Hartl et al., 2008). These findings should be contrasted with the earlier general recommendations of the Brain Trauma Foundation regarding nutrition in TBI that suggested full feeding to metabolic expenditure, which might approach 50 kcal/kg by seven days postinjury (Bratton et al., 2007), with the goal of maintaining lean tissue.
Traditionally, total daily energy expenditure has been considered to be composed of basal energy expenditure, often estimated by the Harris-Benedict equations; activity energy expenditure, which is generally quite limited in hospitalized, critically ill patients; and a small component due to the thermal effect of feeding, representing about 10 percent of the total. The Harris-Benedict equations estimate energy expenditure based on height, weight, sex, and age. A rough estimate of basal energy expenditure for most patients who are not severely malnourished and who are in the same age group is approximately 22–24 kcal/kg/day. Basal energy expenditure increases with injury in proportion to the degree of systemic inflammatory response, ranging usually from 0–100 percent increases from basal expenditure (Duke et al., 1970), with a similar range in TBI (McEvoy et al., 2009). It is likely that all military personnel experiencing TBI will be of normal body composition without malnutrition or obesity. In normal subjects, total energy intake must meet energy expenditure, and provision of adequate protein (0.8 g/kg) and other essential micronutrients is needed to maintain the protein content of the body, the lean body mass. However, for the critically ill, permissive underfeeding (i.e., the modest restriction of nutrient intake, specifically in critically ill patients, over a short term) is recommended. A recent study concluded that, for critically ill patients, permissive underfeeding (60–70 percent of calculated requirement) may be associated with lower mortality rates than underfeeding (90–100 percent of calculated requirement) (Arabi et al., 2011). With severe injury such as TBI, energy intakes in the range of 25–30 kcal/kg/day are generally recommended (Cerra et al., 1997; McClave et al., 2009). Harris-Benedict values are similar at 22–24 kcal/kg, but their calculation includes the additional factors of height, sex, and age. Both values are likely to be underestimates of total energy expenditure, and less likely to promote hyperglycemia for the first two weeks of injury. It should be noted that this concept of permissive underfeeding is different from the concept of underfeeding. Underfeeding is defined as a cumulative total caloric deficit of greater than 10,000 kcal (Bartlett et al., 1982) or a caloric intake of less than nine kcal/kg/day in the first seven days postinjury (Krishnan et al., 2003); both of these situations are associated with higher mortality rates. Furthermore, with the grossly inadequate caloric intake in underfeeding, protein intakes are often even more severely limited. For instance, in the trial of intensive insulin therapy in medical patients, while the caloric intakes were in the range of 1,500 kcal/day, protein intakes were less than 15 g/day in the first three days (Van den Berghe et al., 2006). For reasonable retention of lean tissue with injury, both energy sufficient to meet at least basal energy expenditure and a greater amount of protein (up to 1.5 g/kg/day) are required (Bistrian and Babineau, 1998); even then, total sparing of lean tissue is often impossible in the acute phase of injury, due to the impact of the systemic inflammatory response on the protein catabolic rate (Jensen et al., 2010; Ling et al., 1997). The maximal amount of protein that can be utilized for protein synthesis (i.e., about 1.5 g/kg/day when accompanied by these modest energy intakes) was determined by isotope studies in severely burned individuals (Wolfe et al., 1983) and sophisticated measures of body composition change using in vivo neutron activation analysis (Ishibashi et al., 1998). A second and very important aspect of the relationship between nutrient intake and body composition during injury, including TBI, is that increasing energy intake adversely affects glucose homeostasis. When glucose intakes as TPN exceed 30 kcal/kg (or about 5 mg/kg/min,