Prediction equation for estimating total daily energy requirements of special operations personnel

Special Operation Forces (SOF), are an elite subset of the military population that regularly engage in prolonged strenuous field operations, resulting in high daily physical activity levels [1, 2] and daily energy expenditures that typically exceed that of the average Service member [3, 4]. Achievement of energy balance is compromised during field operations as energy intake is often confined to short dispersed time periods, and the availability of food is limited to what can be carried [5]. With sustained periods of elevated physical activity and limited food supply it becomes increasingly difficult to consume an adequate amount of energy to match energy expenditure, leading to a net negative energy balance (energy expenditure > energy intake) [6]. This negative energy balance will result in reductions in body mass, and more importantly fat-free mass [7, 8]. Ultimately, if these periods of negative energy balance are of sufficient duration, there will be heightened risk of lower body peak power loss and strength decrements [9, 10], potentially compromising Service members’ ability to optimally perform military tasks and subsequently may impact mission success. Understanding daily energy expenditure is thus a critical factor in developing feeding regimens to minimize the severity of negative energy balance under operational conditions to reduce body mass and fat-free mass losses in an attempt to mitigate declines in physical performance [11].

The capability to accurately predict energy requirements during various SOF training operations is an important first step in developing adequate field feeding programs aimed at minimizing negative energy balance. Given that physical activity level is the main variable dictating daily energy expenditure in SOF personnel [1, 2], it is unlikely that energy expenditure equations developed for use by civilian populations, which are based primarily on body mass, would be accurate for this population [12, 13]. Over the past decade the energy expenditures elicited by a wide range of US SOF training operations have been captured [1, 2, 14]. Compiling data from these multiple studies are thus a source to create an algorithm to predict energy cost of SOF training missions.

The objective of this study is to integrate the raw data collected during 12 SOF training operations into a single dataset to develop an algorithm capable of predicting the energy requirements of personnel engaged in SOF training operations. Additionally, this analysis provides the opportunity to examine the occurrence and severity of energy deficits during SOF training. The outcomes from this analysis will provide useful information to develop appropriate feeding regimens aimed at minimizing negative energy balance during SOF operations.

The primary finding of this integrative data analysis was that physical activity level accounted for the majority of the variation between the 12 SOF training operations for both daily energy expenditure (r = 0.91), energy balance (r = − 0.83), and change in body mass (r = − 0.47). Predictive equations for energy expenditure, separating physical activity level into quartiles coupled with body mass (Model A) or FFM (Model B), were correlated (r = 0.74 and r = 0.76, respectively) to measured energy expenditures with a standard error of the estimate of < 650 kcal·d^− 1 for both Models. The predictive energy expenditure equation generated in this analysis will allow for development of interventions and appropriate military doctrine to encourage increased energy intake when physical activity level is anticipated to be high in an attempt to minimize negative energy balances and their associated physical performance decrements.

Energy expenditure during SOF training is quite variable depending on training activity, but averages ~ 4500 Kcal·d^− 1 (range: 3700 to 6300 Kcal·d^− 1). Moreover, SOF Service members experienced an average energy balance of − 1400 Kcal·d^− 1 (range: 250 to − 3900 Kcal·d^− 1) during training operations, resulting in a daily energy deficit of 30% total energy needs and declines in body mass that averaged 1.75 kg (range: 0.24 to − 4.47 kg) over the course of the training operation (4–10 days). Physical activity level was the factor that accounted for the majority of the variance in energy expenditure amongst the factors examined. These observations are not unique, as high physical activity levels and periods of negative energy balances are common during strenuous military training operations [8, 9, 15, 23‐28]. The failure for SOF Service members to adequately increase energy intake to match energy expenditure is likely multifactorial. Standard field feeding protocol is to provide Service members with 3 MREs (~ 3600 Kcal·d^− 1) per day [29]. If every component of the MREs is consumed, with an average energy expenditure of ~ 4500 Kcal·d^− 1, SOF Service members would still be in a negative energy balance of ~ 900 Kcal·d^− 1. As energy is limited by food availability, it is logical that personnel participating in SOF training with higher physical activity demands, and thus energy expenditures, would be the populations with a more severe negative energy balance. Mission objectives may also curtail time available to eat. In the present investigation, Urban Combat and Raider Spirit, trainings that produced the highest measured energy expenditures, also had the lowest energy intakes. These findings suggest that even if additional food is provided, if feeding is not appropriately integrated as part of training operations, it is unlikely that energy balance will be achieved [24]. Chronic or persistent under-eating and negative energy balance will not only result in undesired reductions in body mass, but can also compromise physical performance, and thus potentially mission readiness [30]. Therefore, there is interest in knowing which training events are the most energy demanding and where energy imbalance is most pronounced.

The derivation of a prediction algorithm that can accurately estimate the energy demands of SOF operations, is an important advance for better aligning food and energy availability with mission energy demands. Using regression analysis modeling and capturing physical activity level using a factor-based input, we were able to generate two predictive equations, one using body mass and PAF (Model A) and the other using fat-free mass and PAF (Model B) that had acceptable predictive accuracy. The generation of models using either fat-free mass or total mass was done to see if one or the other appeared to increase prediction acceptability and total mass was desirable in that is a variable that is relatively easy for the end user to capture. Additionally, simplifying activity estimates from measured physical activity levels (daily energy expenditure / daily resting metabolic rate) into four PAF will likely also aid in the practical use of these equations. As data collection for these past investigations were centered around specific training events, such as weapons training, land navigation, squad raids, and ambushes, the four PAF encompass common military tasks relevant to SOF groups across all branches of the military. Despite simplifying physical activity level into four discrete PAFs, the derived equations were highly associated to measured energy expenditures with a standard error of the estimate being ~ 14% (< 650 kcal·d^− 1) of the mean for both Models. Though predictive algorithms derived in the current study provide estimates that overall were not different than measured daily energy expenditures during SOF training operations, the standard error of the estimate may be large enough to result in meaningful declines in performance [31]. Additionally, it is important to note that while both algorithms provided reasonable accuracy for prediction of average daily energy expenditure over the training periods in the data set, the PAF doesn’t capture the time domain, which is an important limitation of our algorithm that should be considered by SOF Service members using these equations. Furthermore, the resting metabolic rate equation used to calculate physical activity level was estimated from participant’s fat-free mass. Not having measurements of resting metabolic rate using indirect calorimetry may be a limitation in our model development and have contributed to the underestimated energy expenditure of the model for higher intensity operation. Use of fat-free mass in estimates of resting metabolic rate may also limit the generalizability to Soldiers with similar body compositions to those included in the current data set. Additional data are needed that contain more day-to-day variability in time spent in these military activities and heterogeneous sample size to properly estimate the accuracy and acceptability of these prediction algorithms.

The nutrition standards for the military personnel, known as the military dietary reference intakes (MDRIs), recommend the average male Service member consume 3400 Kcal·d^− 1 for moderate activity in order to match intake to daily energy expenditures [29]. In context of this recommendation, SOF Service members must consume on average 135% (range: 108 to 185%) more energy than the MDRI to match daily energy expenditures during training operations. The Basic Daily Food Allowance (BDFA) is the amount of money provided per Service member for each meal service. Our findings suggest that the current BDFA allotted to SOF for personnel feeding (BDFA × 1.25) may be modestly too low to support the energy demands of personnel engaged in these training courses, based on greater energy requirements when compared to non-SOF personnel [32]. Increased funding may allow greater flexibility in menu planning potentially improving food options to stimulate energy intake and minimize energy deficits [1, 33]. Lastly, understanding when energy expenditures are anticipated to be elevated during training operations should prove useful to leaders to instill good food intake discipline within their units so as to ensure combat readiness.

While both predictive equations provided reasonably accurate estimations of the measured energy expenditures, they are not without limitations. The current investigation included data collected from 12 training operations in US Army Soldiers and Marines. As US Navy Special Warfare Command (WARCOM) and Air Force Special Operations Command (AFSOC) Service members’ data were not included, the generalizability of these equations to all branches within the US Special Operations Command (USSOCOM) remains to be determined. That said, given the wide variation of energy expenditures and military tasks included in the current investigation, it is likely that many of the tasks included would fall within the scope of activities in these other services and/or their unique activities would fall within the range of activities studied. Additionally, while our predictive equations were highly correlated with measured energy expenditures, there was an under prediction of total daily energy expenditure in the training activity with the highest energy demand (Raider Spirit, ~ 6300 Kcal·d^− 1). Future studies will be needed to understand how best to resolve this error of estimation.

In conclusion, we provide predictive algorithms that include body mass/fat free mass and a physical activity factor that are capable of predicting total daily energy demands of personnel engaged in an array of SOF training missions with acceptable accuracy. Based on these data, SOF Service members have energy requirements that exceed 135% (range: 108 to 185%) of the MDRI, at least during SOF training operations. These predictive equations provide nutrition professionals a tool to predict SOF energy expenditure with reasonable accuracy and logisticians to better align food delivery with energy requirements. The outputs of said tool are also of value for designing appropriate feeding strategies to support the nutritional requirements of the Warfighter.