HARRIS-BENEDICT EQUATIONS DO NOT ADEQUATELY PREDICT ENERGY: DISCUSSION
In this group of elderly African-American hospitalized patients, resting energy expenditure was significantly greater than Harris-Benedict equation predictions. If the Harris-Benedict predictions were applied to caloric delivery, weight loss as a result of underfeeding might be expected. The reasons for this difference may be myriad, but consideration should be given to issues of age, metabolic stress, and race.
Age
The use of Harris-Benedict equations for predicting energy requirements in the elderly has been criticized because of the limited representation of elderly subjects in the original study populations.
In the 1919 series of 239 healthy subjects, the age for men was 27+/-9 (range= 16-63) and for women was 31+/-14 (range=15-74) years. In 1935, Benedict reported on data from older, healthy subjects (eight men age=84.1+/-6.8 and 35 women age=75.9+/-6.1) with a range of 66-91 years, but the original equations were not revised, even though 5/8 men and 9/36 women had measured RMR that was >10% different from the earlier predictions. In fact, the ages of men in this current cohort all exceeded the range of ages included in the Harris-Benedict equation, and the mean age for men and women was greater than the upper boundary of age range. Clearly, on the basis of age difference alone, the Harris-Benedict predictions may not be appropriate for such an elderly group of patients. canadian pharmacy viagra
The Harris-Benedict equations have also been criticized for use in hospitalized patients. Benedict made a clear statement against using “hospital normals” to determine physiological standards. He stated in 1928 that ill subjects would likely have reduced basal energy needs due to malnutrition, whereas the opposite is likely true.
Hypermetabolism has been defined as a measured RMR >110% of Harris-Benedict BEE. In the current cohort, the mean RMR was 120% of the Harris-Benedict prediction and consistent with hypermetabolism. Our patients had energy expenditure measured during an unstable, stressed period requiring hospitalization. Hypermetabolism has frequently been described in patients with critical illnesses, including sepsis, postoperative states, and medical disorders with an inflammatory component. For example, Long et al. documented an increase in RMR of 120% beyond resting energy predictions due to surgery, 135% for trauma, 160% for severe sepsis, and 210% for severe burns. Clearly, Table 1 verifies by clinical conditions the likelihood of hypermetabolism. Our undernourished patients, those with BMI <18.5 kg/m2, were similar to a previously reported group of 14 elderly, hospitalized, severely underweight (BMI 15.8+/-1.8 kg/m2) patients. That group’s mean RMR was 126% of the Harris-Benedict prediction, and mean caloric expenditure was 31.4 kcal/kg. It was therefore not surprising that our cohort of elderly patients with a wide variety of medical illnesses would have elevated energy needs; however, the range in kcal/kg is broad. Race Energy expenditure has been evaluated to a very limited degree in elderly community-dwelling African-American subjects. To our knowledge, these data are the first reported in hospitalized African-American patients of this age. In two trials of elderly African-American subjects (28 males, 37 females in each trial),the mean age was 15 years younger, BMI 6-10 units higher, and weight 20 kg greater than the current cohort. Although the absolute resting energy expenditure is essentially the same as our current measures (1,631 kcal in men, 1,576 kcal in men and 1,390 kcal in women, 1,431 kcal in women, it is greatly increased in our cohort when considered relative to body weight in kg (approximately 17 kcal/kg in reference 23 versus the current cohort’s 25 kcal/kg/day. The weight is not available in reference 27). Perhaps the most pertinent reason for the difference between the current data and predictive equations is that Harris and Benedict did not study African-American subjects in the original data used to develop their predictive equations”. Benedict later measured basal energy needs of Jamaican blacks in 1928 using a “field respiration apparatus” to measure 23 young (18-21 years) men and 19 young (16-24 years) women. While the mean deviations from Harris-Benedict predictions were -3.4+/-1.3%o in men and -5.4+/-0.8% in women, 26/45 measurements varied by >10%. Whether these differences predominantly reflected a population difference or change in measurement methodology is not clear. Recent studies, however, have suggested that a difference in RMR between African-American and Caucasian subjects may be based on the expression of uncoupling protein genes. suhagra
Limitations
The study reported herein has several limitations. First, total energy expenditure was not measured due to the unavailability of doubly labeled water methodology, and activity expenditure was not accounted for. For the community-dwelling population, total energy expenditure is estimated at 150% RMR. By contrast, in ill, ventilated patients, average total energy expenditure is only 110% RMR, though the RMR may be considerably elevated from basal needs. Perhaps actual total energy needs for this group of patients lie between these two measures, as activity energy expenditure during hospitalization is typically very limited due to the added impact of illness on old age and frailty.
Secondly, the energy intake side of the energy balance equation and serial body weights were not measured, so we have no information on the adequacy of caloric intake or actual weight loss. Significant weight loss during hospitalization may have a profound impact on mortality in elderly patients. The elderly have excess mortality with reduced BMI and with loss of >10% of maximum lifetime weight within 10 years. In this cohort, 31% were malnourished by BMI at admission. By contrast to the general population, overweight in the elderly does not confer excess risk of mortality until the BMI is >27 or 28 kg/m2. In this cohort, 23% were overweight or obese.
Implications
The clinical question that remains is how best to determine a given elderly individual’s caloric needs in the setting of acute illness. This is particularly critical if supplemental enteral or parenteral feedings are required for a prolonged period of time, as underfeeding may lead to malnutrition and poor outcomes, and the significant risks of overfeeding include refeeding syndrome, hepatic steatosis, and excess CO2 production. prescription drugs from canada
Using our group’s mean estimate of 25 kcal/kg would be a place to begin, though this figure conceals the actual broad range of caloric need (14-39 kcal/kg/day). To estimate caloric needs for weight maintenance, an additional factor for caloric expenditure of physical activity is needed. A variable multiplier of the resting energy estimate between 1.1 (very inactive) to 1.5 (fully and frequently ambulatory) should be considered. It is advisable to add 500 kcal/day to promote a safe rate of weight gain, if that outcome is desired, such as in patients whose BMI is <18.5 kg/m2. Clearly, patients at greater nutritional risk should have individual RMR measurements using indirect calorimetry. Future trials of measured RMR in a large cohort of elderly, hospitalized African-American patients are needed to develop improved predictive equations, though it may be that the variability among hospitalized patients in general makes predictive equations inherently limited.
Summary
This study underscores the limitations of predictive equations in current use to estimate energy requirements in very elderly, ill, African-American patients. Since no adequate estimate has been clini cally validated in this population, and the actual range in energy needs is quite variable, the safest plan for longer-term nutrient prescription and for all patients with nutritional risk remains the measurement of energy expenditure.
buy levitra uk
