In brief, the obese ICU patient should be slightly underfed, and should have an excess of dietary protein. According to the current consensus, one should manage the nutritional demands of the obese ICU patient in the following manner:
This strategy minimises the complications of overfeeding, while minimising protein catabolism and maximising protein synthetic function. This is asked about in Question 24 from the first paper of 2012; and the identical Question 25 from the first paper of 2019; each time the college gave us a patient with a BMI of 40.
In summary, from Coeffier et al (2016), the following metabolic matters trouble the obese patient in the ICU:
The ASPEN Guidelines are the only major statement which makes recommendations in this setting. The ESPEN Guidelines and the Canadian Clinical Practice Guidelines make no specific recommendations about the morbidly obese patient (at least none that I could find).
Ravussin et al (1982) measured the daily energy expenditure of both normal and obese subjects in "a comfortable airtight respiration chamber". The resting energy expenditure of the obese test subjects was much higher than what might have been predicted from their ideal body weight. In fact, only 8% of the "extra " energy expenditure could be attributed to the increased cost of movement, invested in shifting their fatty limbs to and fro. Most of the extra energy was spent at rest, by the "fat free mass".
The reason for this increased metabolic rate is the surprising muscularity of the morbidly obese patient. The increase in body weight, even though mainly fatty tissue is gained, also results in the increase in the ratio of lean body tissue to the total body mass. The lean tissue proportion must increase inevitably, as even minor efforts of mobility require substantially more muscle. The morbidly obese person is constantly doing weight training against the resistance of their own gargantuan bulk. The consequence of this increased muscle mass is an increase in the resting metabolic rate, out of proportion to what might be expected from the ideal body weight.
Yes, fat people have diabetes. This is not surprising to anybody, and is the expected feature of managing a morbidly obese patient in the ICU. Why is it relevant and why does it get a mention in this chapter? Well, it is all to do with the normal survival-promoting adaptations to severe illness. As Port et al (2010) put it, "obesity is a proinflammatory state and probably lowers the threshold at which these mechanisms become overwhelmed or exaggerated during critical illness." Alterations in glucose metabolism associated with the stress response to critical illness are already going to put you into a hyperglycaemic state, with decreased insulin levels and decreased insulin sensitivity. To add a further impairment to this state results in hyperglycaemia which is more difficult to control. Complications of hyperglycaemia ensue (these are discussed elsewhere). In brief, it increases mortality, predisposes the patient to infection, and facilitates the proinflammatory state.
At least in the studied context of trauma, Jeevanandam et al (1991) found that the obese subjects mobilized relatively more protein and less fat compared with non-obese subjects. Lean patients relied largely on fatty acid oxidation for energy (about 61% of their resting energy expenditure was burning fat) whereas obese patients derived most energy from protein catabolism (only 39% of energy came from free fatty acids).
This seems dysfunctional, as fat is the form in which energy is stored, whereas all body protein is structural and is therefore essential. Why do these people preferentially dismantle important structural protein? There appears to be some sort of relative block to the mobilisation of lipid. It is thought that this block is the result of poor catecholamine response to stress among the obese patients: catecholamine release is the major stimulus to stress-induced lipolysis in the non-obese patient. Moreover, there is an increased rate of insulin synthesis, all due to insulin resistance, which has the same lipolysis-dampening effect. The morbidly obese metabolism resorts to protein catabolism to stimulate gluconeogenesis as a consequence of these hormonal changes.
The following theoretical, pragmatic and logistic obstacles arise to frustrate the appropriate delivery of nutrition to the morbidly obese patient:
Normal equations break down at the extremes of weight. For example, the usual "25kcal/kg/day" shortcut fails miserably in somebody whose weight exceeds 100kg. If you use the actual body weight to calculate metabolic requirements, you end up grossly over-feeding the patient; if you use the ideal body weight then you end up under-feeding them (having failed to take into account their exaggerated resting metabolic rate). To say nothing of the changed metabolic requirements in critical illness.
Clearly, there should be some better way of estimating their nutritional needs.
Indirect calorimetry is one such way. it is after all the gold standard. However, it is not available to everybody. Which predictive equations come closest? Frankenfield et al (2013) measured the resting energy expenditure of both morbidly obese and severely malnourished patients, and then compared the measured values to the predictions of such equations as the Penn State equation, Faisy equation, Ireton-Jones equation, Mifflin–St Jeor equation, Harris-Benedict equation, and American College of Chest Physicians standard (the ACCP are the people who stand behind that "25kcal/kg/day" shortcut). The equations differed wildly in their predictive value. For instance, the ACCC standard was accurate in exactly 0% of the patients (that's right - zero). The best performer was the Penn State equation - 75% of the predictions fell within 10% of the measured resting energy expenditure. To celebrate this achievement, the Penn State equation is reproduced below:
REE = (BEE ×1.1) + (VE × 32) + (Tmax × 140) - 5340
- BEE is the basal energy expenditure as predicted by the Harris-Benedict equation,
- VE is the minute volume in L/min,
- Tmax is the maximum temperature in degrees Celsius
Having said all this, if you have no indirect calorimetry and you are too lazy to perform the necessary calculations for the Penn State equation, you may still resort to using crude approximations based on measured weight. Ireton-Jones et al (1991) have demonstrated that actual body weight (frequently 30% above the ideal body weight) is the better parameter to use.
The ASPEN guidelines recommend high-protein hypocaloric feeding. (i.e. they recommend you do not exceed 60-70% of calculated energy requirements (i.e. intentionally underfeed them) - this is based on expert consensus rather than any specific trial. Choban et al (1997) had investigated hypocaloric vs. eucaloric feeding and found little difference in nitrogen balance between the two. From such data it is surmised that we should underfeed these patients. If the nitrogen balance is the same, there should be no protein catabolism, and with fewer calories you are more likely to mobilise fat stores and less likely to develop complications of overfeeding.
According to ASPEN, you should also oversupply protein: 2.0-2.5g/kg of ideal body weight, or 1.3-1.5g/kg of actual body weight. This is also based on the same study by Choban et al (1997) which demonstrated that 2.0g/kg/day of protein was not sufficient for maintaining a neutral nitrogen balance.