Central and peripheral mechanisms regulating energy homoeostasis and their implications in spinal cord injury

The central nervous system plays a vital role in modulating energy status, and the hypothalamus is the integrating, superordinate principal regulator of whole-body energy homoeostasis (Fig. 1). The arcuate nucleus within the hypothalamus plays a critical role in the regulation of feeding and metabolism. It integrates hormonal and nutritional signals from the peripheral circulation, as well as peripheral and central neuronal inputs, to generate a coordinated feedback response. The arcuate nucleus projects to second-order neurons in the paraventricular, dorsomedial, lateral and ventromedial nuclei of the hypothalamus. The second-order neurons further process the received information and project to multiple extrahypothalamic neurocircuits, leading to an integrated response that regulates energy intake and energy expenditure^{(Reference Timper and Brüning69)}. These centres jointly summate influences from various circulating substrates, hormones, neuropeptides and neurotransmitter signals that regulate food intake.

Fig. 1. The neuroendocrine components involved in the regulation of energy balance relevant to spinal cord injury. Organs and systemic signaling pathways are represented with green lines (circulating hormonal signals), red (voluntary neurological signals) and blue (autonomic neurological signals). The pop-out shows the action of these signals on regions in the hypothalamus and brainstem. Legend: 3V, third ventricle; ARC, arcuate nucleus; Carbs, carbohydrates, CCK, cholecystokinin; DMH, dorsomedial hypothalamic nucleus; FGF, fibroblast growth factors; FFA, free fatty acids; GLP-1, glucagon-like peptide 1; lat, lateral nucleus; n., nerve; PVN, paraventricular nucleus; P-YY-_3–36, peptide YY_3–36; SN, substantia nigra; and VMH, ventromedial hypothalamic nucleus. , Humoral; , Autonomic; , Voluntary

Gastrointestinal hormones also have a principal role in regulating central nervous system-dependent energy control. Ghrelin is mainly secreted from the stomach during a fasted state and stimulates body weight gain, adiposity and central feeding centres by activating neurons in the hypothalamus that stimulate food intake^{(Reference Timper and Brüning69)}. Various other hormones, such as peptide YY_3–36, cholecystokinin and glucagon-like peptide 1 are secreted from the small intestine upon the ingestion of foodstuff and exert appetite-suppressing effects in various brain regions, such as hypothalamic and brainstem nuclei, and by modulating vagal afferents, the peripheral elements of the brain–gut axis (Fig. 1)^{(Reference Crespo, Cachero and Jiménez70)}. Gut peptides provide information on ‘real-time’ food consumption and modify electrical activity of the vagal afferent pathway by attaching to vagal receptors that extend into the digestive tract mucosa. These intestinal-derived signals are sent by the vagus nerve to the nucleus of the solitary tract, with further projection to hypothalamic regions (Fig. 1)^{(Reference Miller71)}.

Disruption of the central mechanisms modulating energy metabolism has been previously recognised as the aetiology of obesity in non-disabled persons^{(Reference Timper and Brüning69,Reference Miller71)} . Obesity and cardiometabolic disorders are frequently associated with diminished production or resistance to the production of central and peripheral regulators of energy homoeostasis, including food intake and energy expenditure^{(Reference Elmquist and Joel72)}. Naznin et al. ^{(Reference Naznin, Toshinai and Waise73)} and Waise et al. ^{(Reference Waise, Toshinai and Naznin74)} reported that obesity-induced systemic inflammation spreads to the vagus nerve and subsequently the hypothalamus leading to the dysregulation of central and peripheral mechanisms governing satiety, energy regulation and fuel metabolism. With obesity and high-fat, energy-dense diets, increases in the concentration of saturated fatty acids from the periphery cross the blood–brain barrier and induce an inflammatory response on hypothalamic neurons^{(Reference Thaler, C-X and Schur75)}. Vinik et al. ^{(Reference Vinik, Maser and Mitchell76)} reported that obesity and type 2 diabetes mellitus-induced neuropathies alter vagal nerve neurotransmission, preventing bidirectional crosstalk between the central nervous system and the gut.

Although supraspinal centres remain intact following SCI, several neurological and endometabolic factors are influenced by the disruption of the central nervous system. With an SCI, compromised afferent and efferent signals from central and peripheral locations lead to dysregulation of the intricate equilibrium of energy metabolism. Physiological cues that are present in persons with SCI and influence appetite, satiety/satiation and energy balance are disrupted, further contributing to an imbalance in energy homoeostasis^{(Reference Holmes41)}. Besecker et al. ^{(Reference Besecker, Blanke and Deiter77)} have proposed an SCI-induced gastric vagal afferent neuropathy as a cause for homoeostatic dysregulation of energy balance in experimental SCI in rats. The authors hypothesise the disruption of the reflex transmission of chemical feeding-related signals from the gastrointestinal tract to the CNS^{(Reference Besecker, Blanke and Deiter77)}. The central and hypothesised peripheral dysregulation of energy homoeostasis and a deterioration of body composition that results in physical deconditioning produce the ‘perfect storm’ for the onset of neurogenic obesity and cardiometabolic risk in persons with SCI.

Energy expenditure after spinal cord injury

Energy balance reflects a dynamic relationship between energy expenditure and energy intake. TDEE represents the number of energies burned over 24 h and is the sum of basal metabolic rate (BMR), the thermic effect of physical activity (TEPA) and the thermic effect of food digestion (TEF)^{(Reference Farkas, Pitot and Gater78)}.

In the non-disabled population, TEPA and TEF account for approximately 20 % and 8% of TDEE^{(Reference Farkas, Pitot and Gater78)}, whereas after SCI they account for about 5 % and 6 %, respectively^{(Reference Monroe, Tataranni and Pratley34,Reference Buchholz, McGillivray and Pencharz79)} . To date, limited research has tested differences in the TEPA and TEF between persons with and without SCI. Monroe et al. ^{(Reference Monroe, Tataranni and Pratley34)}, measured TEPA using a respiratory chamber. They observed significantly less TEPA in men with SCI compared with men without SCI^{(Reference Monroe, Tataranni and Pratley34)}. This finding is likely because movement is restricted to the upper limbs. Consequently, the energy cost of exercise and activities of daily living is significantly lower in SCI compared with a non-disabled person^{(Reference Collins, Gater and Kiratli80)}. Aksnes et al. ^{(Reference Aksnes, Brundin and Hjeltnes81)} and Buchholz et al. ^{(Reference Buchholz, McGillivray and Pencharz79)}did not find differences in TEF between persons with chronic SCI and non-disabled controls, possibly because of a 2-h post-prandial testing window that only captured the parasympathetic-controlled obligatory phase of TEF. Alternatively, Monroe et al. ^{(Reference Monroe, Tataranni and Pratley34)} and Asahara and Yamasaki^{(Reference Asahara and Yamasaki82)} reported significant differences between persons with and without SCI for a 24- and 3-hour test, respectively. The authors of both studies incorporated a longer testing duration, therefore, capturing both the obligatory and the sympathetic and skeletal muscle mass-mediated facultative phases of TEF^{(Reference Asahara and Yamasaki82)}. Both TEF and TEPA present unique assessment challenges attributed to the scarcity of literature. Therefore, TEF and TEPA should remain an activeresearch focus given their influence on energy intake and TDEE.

BMR typically accounts for 60 to 70 % of TDEE in the non-disabled population^{(Reference Melanson83)}, but in persons with SCI, it accounts for 70 to 80 %^{(Reference Gorgey, Caudill and Sistrun84)}. Fat-free mass, composed of bone, muscle and organs, contributes the most to BMR, and of fat-free mass, skeletal muscle mass accounts for 85 % of the variance. Attenuation of BMR following SCI originates from a significant reduction in metabolically active tissue^{(Reference Farkas, Gorgey and Dolbow9)}, sympathetic nervous system dysfunction^{(Reference Buchholz and Pencharz85)} and altered hormonal milieu^{(Reference Gorgey, Khalil and Gill86)}.

Most SCI literature measures resting metabolic rate (RMR) or resting energy expenditure rather than the more precise measurement of BMR (Table 3). The available literature indicates that the mean measured BMR for persons with SCI ranges from 1022 to 1943 kcal/d, and mean measured RMR ranges from 959 to 2519 kcal/d (Table 3)^{(Reference Farkas, Gorgey and Dolbow9,Reference Monroe, Tataranni and Pratley34,Reference Buchholz, McGillivray and Pencharz79–Reference Aksnes, Brundin and Hjeltnes81,Reference Gorgey, Caudill and Sistrun84,Reference Bauman, Spungen and Wang87–Reference Aquilani, Boschi and Contardi120)} . Buchholz et al. ^{(Reference Buchholz, McGillivray and Pencharz79)} reported resting metabolism was significantly lower in persons with paraplegia (1472 kcal/d) compared with BMI-matched non-disabled controls (1677 kcal/d). In another study by Buchholz et al. ^{(Reference Buchholz, McGillivray and Pencharz121)}, authors examined RMR in twenty-seven persons with SCI by injury completeness (complete: 1417 v. incomplete: 1480 kcal/d) and sex (men: 1555 v. women: 1245 kcal/d), only observing significant differences by sex. Gorgey et al. ^{(Reference Gorgey, Farkas and Dolbow110)} recently reported non-significant differences in BMR by sex (men: 1421 v. women: 1367 kcal/d) and Farkas et al.^{(Reference Farkas, Gorgey and Dolbow9)} observed significant differences by LOI (tetraplegia: 1224 and paraplegia: 1517 kcal/d) in motor complete SCI. Collins et al. ^{(Reference Collins, Gater and Kiratli80)}, however, did not report significant differences in RMR by LOI. These differences may stem from population demographics as Farkas et al. ^{(Reference Gorgey, Farkas and Dolbow110)} examined chronic motor complete SCI, whereas Collins et al. ^{(Reference Collins, Gater and Kiratli80)} included complete and incomplete SCI. Additionally, Farkas et al. ^{(Reference Gorgey, Farkas and Dolbow110)} measured BMR and Collins et al. ^{(Reference Collins, Gater and Kiratli80)} assessed RMR in their respective studies.

Table 3. BMR/RMR in spinal cord injury literature

LOI, level of injury; AIS, American Spinal Injury Association Impairment Scale; TSI, time since injury; Para, paraplegia; PI, pressure injuries; C, complete; I, incomplete; Tetra, tetraplegia.

Blank spaces indicate data were not provided in the study; data are presented as mean ± standard deviation.

* RMR/resting energy expenditure was measured.

† Standard deviation not provided.

** BMR was measured.

RMR and BMR primarily differ in testing procedures, but both are non-invasively measured with indirect calorimetry using a metabolic cart^{(Reference Farkas, Pitot and Gater78)}. The participant lies in a supine position in a dark room with minimal movement following at least an 8-h fast for RMR or a 12-h fast for BMR^{(Reference McMurray, Soares and Caspersen122)}. Because RMR is not at a basal state, it is usually higher than BMR for persons with and without SCI as only a short quiescent period is required (10 to 20 min^{(Reference Compher, Frankenfield and Keim123)}) prior to data acquisition^{(Reference Bauman, Spungen and Wang87)}. Rather for BMR, the participant is awakened in the morning following an overnight stay, refrains from exercise, caffeine and alcohol for the previous 24 h, is free from emotional stress, and familiar with the apparatus^{(Reference McMurray, Soares and Caspersen122,Reference Henry124)} . Bauman et al. ^{(Reference Bauman, Spungen and Wang87)} examined BMR and RMR in pairs of monozygotic twins with and without SCI. The authors reported lower basal metabolism (SCI twin: 1387 and non-SCI twin: 1660 kcal/d) in both groups compared with RMR values (SCI twin: 1682 and non-SCI twin: 1854 kcal/d). Additionally, both values were significantly lower in SCI compared with individuals without the injury^{(Reference Bauman, Spungen and Wang87)}. There is an approximate 20 % difference between BMR and RMR values in persons with SCI compared with an 11% difference in persons without SCI. Considering that resting/basal metabolism are the largest components of TDEE in persons with and without SCI and it is significantly influenced by fat-free mass, several studies have reported that RMR can be used as a strong predictor of energy intake^{(Reference Blundell, Caudwell and Gibbons125–Reference Hopkins, Finlayson and Duarte128)}. When using RMR rather than BMR as a predictor of energetic intake in persons with SCI, dietary need can be overestimated by nearly 400 kcal/d (assuming approximately a 1900 kcal/d diet^{(Reference Farkas, Pitot and Berg19)}). These data indicate BMR is a more sensitive indicator of energetic need, and less reliance should be placed on RMR in SCI research. However, TDEE remains superior as it accounts for the multiple components of daily energy expenditure.

According to published studies, during the acute phase of SCI, TDEE ranges from 2030 to 3344 kcal/d (Table 4)^{(Reference Cox, Weiss and Posuniak89,Reference Barco, Smith and Peerless93,Reference Desneves, Panisset and Rafferty94,Reference Rowan and Kazemi97,Reference Wouda, Lundgaard and Becker101,Reference Wouda, Lundgaard and Becker102,Reference Rodriguez, Benzel and Clevenger129)} . In chronic SCI, TDEE is from 1332 to 2834 kcal/d (Table 4)^{(Reference Farkas, Gorgey and Dolbow9,Reference Monroe, Tataranni and Pratley34,Reference Desneves, Panisset and Rafferty94,Reference Shea, Shay and Leiter98–Reference Tanhoffer, Tanhoffer and Raymond100,Reference Buchholz, McGillivray and Pencharz121,Reference Mollinger, Spurr and el Ghatit130)} . TDEE is reduced in persons with chronic SCI by as much as 54 % in persons with tetraplegia^{(Reference Mollinger, Spurr and el Ghatit130)} and nearly 20 % in individuals with paraplegia^{(Reference Buchholz, McGillivray and Pencharz121)}. TDEE can be assessed by measuring average daily energy expenditure using direct, or whole body, calorimetry, doubly labelled water, or mechanical ventilation. Of the SCI literature that assessed TDEE, only 33 % measured TDEE. Direct calorimetry measures the amount of heat produced while enclosed within a respiratory chamber and is the gold standard for measuring energy metabolism^{(Reference Farkas, Pitot and Gater78)}. A test participant is completely enclosed in the chamber where there are no social interactions during the measurements and audiovisual contact with investigators^{(Reference Schutz131)}. In cross-sectional study designs, the participants spend a minimum of 24 h continuously, and up to a week (or more) in dietary intervention studies^{(Reference Schutz131)}. This method has several limitations, including the cost of highly specialised equipment, space to house the equipment, confinement of the participant and the need to exclude anything emitting heat other than the research subject. For persons with paralysis, and especially high injury levels, direct calorimetry is unrealistic because of the need for caregiver assistance, power wheelchairs and/or assistive electronic devices. To date, Monroe and colleagues^{(Reference Monroe, Tataranni and Pratley34)} are the only investigators to use direct calorimetry with a respiratory chamber in persons with SCI. The authors demonstrated a TDEE of 1870 kcal/d in chronic complete SCI compared with 2376 kcal in persons without SCI^{(Reference Monroe, Tataranni and Pratley34)}, a 24 % difference.

Table 4. TDEE after SCI

SCI, spinal cord injury; LOI, level of injury; AIS, American Spinal Injury Association Impairment Scale; TSI, time since injury; TDEE, total daily energy expenditure; Para, paraplegia; C, complete; I, incomplete; Tetra, tetraplegia;

Blank spaces indicate data were not provided in the study; data are presented as mean ± standard deviation.

* Standard deviation are not provided.

Less labour- and time-intensive methods to measure energy expenditure are with doubly labelled water or a metabolic cart during mechanical ventilation. Doubly labelled water is centred around the difference between the apparent turnover rates of the hydrogen and oxygen of body water as a function of carbon dioxide production. The procedure encompasses enriching a research participant with heavy oxygen and heavy hydrogen and then determining the difference in washout kinetics between the isotopes. The oxygen isotope is lost as water and as carbon dioxide due to exchange in the bicarbonate pools. The hydrogen isotope is lost only as water^{(Reference Westerterp132)}. The strength of doubly labelled water is that it is a non-invasive and inconspicuous free-living evaluation of TDEE with no constraint or restriction for the participant. The total number of variables in the equations used to calculate TDEE from doubly labelled water is nine, plus two additional constants, from which the equation for isotope dilution spaces calculation includes five variables and one constant^{(Reference Schutz131)}, thus, making the technique mathematically complex and prone to miscalculation. Double labelled water is infrequently used (Desneves et al. ^{(Reference Desneves, Panisset and Rafferty94)} and Tanhoffer et al. ^{(Reference Tanhoffer, Tanhoffer and Raymond99,Reference Tanhoffer, Tanhoffer and Raymond100)} ) to measure TDEE after SCI. In 2012 and 2015, Tanhoffer et al. ^{(Reference Tanhoffer, Tanhoffer and Raymond99,Reference Tanhoffer, Tanhoffer and Raymond100)} reported a mean TDEE of 2346 and 2406 kcal/d, respectively, in chronic SCI, while Desneves et al. ^{(Reference Desneves, Panisset and Rafferty94)} reported 2354 kcal/d during the acute stage. These observed values far exceed the reported values by Monroe et al. ^{(Reference Monroe, Tataranni and Pratley34)} In the acute phase, mechanical ventilators provide a unique option to continuously measure respiratory gases with the addition of a metabolic monitor, similar to indirect calorimetry^{(Reference Weissman, Sardar and Kemper133)}, as published by Barco et al. ^{(Reference Barco, Smith and Peerless93)} Because nutritional risk is associated with ventilatory support after SCI and high levels of injury often need respiratory management^{(Reference Wong, Derry and Jamous134)}, the use of metabolic monitoring with mechanical ventilators can provide an easy method to determine energetic need during the initial hospitalisation. However, it is likely because of the lack of availability of the specialised equipment, cost and trained personnel required to measure TDEE that most studies rely on predicting TDEE, a method prone to error^{(Reference Farkas, Pitot and Gater78)}. Of the literature that assessed TDEE, 71 % estimated it through various methods (Table 4). Several of these methods are examined in the next section given their clinical use in defining energetic targets.

Energy (energetic) intake relative to energy expenditure after spinal cord injury

Energy intake reflects energetic gain by the ingestion of foodstuff of different energetic densities. Defining optimal nutrient intake and its management is challenging after SCI because practical guidelines for determining energy requirements for this niche population are limited. Energetic need is dependent on several factors after SCI, including the injury phase^{(Reference Perret and Stoffel-Kurt117)} and the physical activity/therapy within phase^{(Reference Mollinger, Spurr and el Ghatit130)}, the level^{(Reference Farkas, Gorgey and Dolbow9,Reference Cox, Weiss and Posuniak89)} and completeness^{(Reference Mollinger, Spurr and el Ghatit130)} of the injury, sex^{(Reference Groah, Nash and Ljungberg135)}, body composition and its post-injury changes^{(Reference Buchholz and Pencharz85,Reference Yilmaz, Yasar and Goktepe103)} , presence of infection or pressure injuries^{(Reference Aquilani, Boschi and Contardi120)}, and frequency and accuracy of reporting on the energetic intake used to assess nutrition^{(Reference Farkas, Pitot and Berg19,Reference Gorgey, Caudill and Sistrun84)} . Consequently, tremendous variability in energetic intake (Table 5) is reported across the literature, even though most studies report it is within or exceeds daily AHA, ADG, DGA and PHE recommendations (Table 6).

Table 5. Total energetic and macronutrient intake in SCI literature

SCI, spinal cord injury; LOI, level of injury; AIS, American Spinal Injury Association Impairment Scale; TSI, time since injury; SR, self-report; Para, paraplegia; Tetra, tetraplegia; C, complete; I, incomplete; PI, pressure injuries; FFQ, food frequency questionnaire; RD, registered dietitian.

Blank spaces indicate data were not provided in the study; data are presented as mean ± standard deviation.

* Standard deviation not provided.

Table 6. Comparison of authoritative, evidence-based non-spinal cord injury dietary guidelines

AMDR, Acceptable Macronutrient Distribution Range; DGA, Dietary Guidelines for American; AI, adequate intake; UL, tolerable upper intake level.

* Guidelines were established to prevent CVD and manage overweight and obese adults.

† Estimated energy requirement equation for males = 662 – (9·53 × age (years)) + PA × ((15·91 × weight (kg)) + (539·6 × height (m))), where PA is the physical activity coefficient.

‡ Estimated energy requirement equation for females = 354 – (6·91 × age (years)) + PA × ((9·36 × weight (kg)) + (726 × height (m))), where PA is the physical activity coefficient.

§ Values in parentheses are an example of the total g/d of total fibre calculated from 14 g/1000 kcal multiplied by the median energy intake (kcal/1000 kcal/d) from the Continuing Survey of Food Intakes by Individuals (1994–1996, 1998).

The IOM encourages establishing energetic intake using a sex-specific prediction equation, relying on age, height, weight and physical activity. While less precise than measuring energy requirements, the IOM is the only guideline that makes such a recommendation rather than providing an acceptable macronutrient distribution range. A limitation of IOM’s equation is that it neglects to include resting or basal metabolism as the largest determinant of TDEE, and therefore energetic intake. TDEE can also be estimated using the product of BMR (or RMR) and the common activity correction factor of 1·2^{(Reference Farkas, Gorgey and Dolbow9)}. Several studies have predicted TDEE using 1·2 and other previously published activity, stress, injury, and/or trauma correction factors to determine energetic intake for persons with SCI^{(Reference Monroe, Tataranni and Pratley34,Reference Cox, Weiss and Posuniak89,Reference Rowan and Kazemi97,Reference Shea, Shay and Leiter98,Reference Wouda, Lundgaard and Becker101,Reference Wouda, Lundgaard and Becker102,Reference Buchholz, McGillivray and Pencharz121,Reference Rodriguez, Benzel and Clevenger129)} .

In the acute phase of SCI, the literature indicates total energetic intake ranges from 755 to 2290 kcal/d (Table 5)^{(Reference Cox, Weiss and Posuniak89,Reference Rowan and Kazemi97,Reference Kearns, Thompson and Werner119,Reference Aquilani, Boschi and Contardi120,Reference Iyer, Beck and Walton136–Reference Laven, Huang and DeVivo138)} . Over the first 4 weeks of the SCI, this value increases by over 400 kcal/d^{(Reference Kearns, Thompson and Werner119)} and likely results from the thermic effect of voluntary respiration (diaphragmatic and intercostal muscle activation) for those weaned from mechanical ventilators. The conversion from the catabolic state to declining energetic needs is not well researched, though RMR has been shown to decrease 10 weeks post-SCI^{(Reference Felleiter, Haeberli and Schmid139)}. In seminal work by Cox et al. ^{(Reference Cox, Weiss and Posuniak89)}, the authors reported that persons in the early rehabilitation phase of the injury require up to 54 % less energy content than would be predicted by most standard formulae. The authors further determined that in the rehabilitation phase, persons with tetraplegia need 22·7 kcal/kg/d and persons with paraplegia need 27·9 kcal/kg/d^{(Reference Cox, Weiss and Posuniak89)}, guidelines that are still widely used today⁽¹⁴⁰⁾. Of note, this calculation published by Cox et al. ^{(Reference Cox, Weiss and Posuniak89)} was developed in fifteen persons with tetraplegia and five with paraplegia and mostly men (86 %) whom typically expend^{(Reference Gorgey, Farkas and Dolbow110,Reference Buchholz, McGillivray and Pencharz121)} and consume^{(Reference Iyer, Beck and Walton136,Reference Sabour, Soltani and Latifi141)} more than women. The equations also do not account for the serial weight loss and weight regain that occurs during the early phase of the SCI and a drop in energy expenditure that persists through the rehabilitation phases of treatment.

When evaluating the energy intake and expenditure data, persons with acute SCI appear to be in a negative energy balance (TDEE: 2030 to 3344 kcal/d v. energetic intake: 755 to 2290 kcal/d). Distinct from other trauma conditions, persons with acute SCI do not demonstrate hypermetabolism following injury^{(Reference Kearns, Thompson and Werner119,Reference Rodriguez, Benzel and Clevenger129,Reference Kolpek, Ott and Record142,Reference Rodriguez, Clevenger and Osler143)} . While a negative nitrogen balance does occur, this is obligatory. While the underlying mechanism contributing to a negative nitrogen balance following SCI remains poorly understood^{(Reference Kolpek, Ott and Record142–Reference Tator, van der Jagt and Malkin145)}, efforts made to shift the obligatory loss of nitrogen by increasing energetic intake can lead to overfeeding. Confounding this matter is that several studies and registered dietitians use correction factors to increase energetic intake. Kaufman et al. ^{(Reference Kaufman, Rowlands and Stein137)} and Barco et al. ^{(Reference Barco, Smith and Peerless93)} used an activity factor of 1·1. However, a stress factor of 1·2 to 1·75 is routinely used in in the literature^{(Reference Cox, Weiss and Posuniak89,Reference Rodriguez, Benzel and Clevenger129,Reference Kaufman, Rowlands and Stein137,Reference Rodriguez, Clevenger and Osler143)} . Rodriguez et al. ^{(Reference Rodriguez, Benzel and Clevenger129)} examined the Harris-Benedict equation with an activity factor of 1·2 and an injury factor of 1·6 and identified that the equation overestimated energetic requirements in twelve persons with acute SCI. The same authors reported that establishing nutritional management upon serial indirect calorimetry measurements with higher stress and activity factors result in overfeeding^{(Reference Rodriguez, Benzel and Clevenger129)}. Kearns et al. ^{(Reference Kearns, Thompson and Werner119)} compared measured RMR with the Harris-Benedict equation in five individuals with tetraplegia and reported that the use of the equation leads to an overfeeding by nearly 70 %. While the time frame of the investigation relative to SCI was not specified, the authors suggested administering 80 % of predicted energetic needs^{(Reference Kearns, Thompson and Werner119)}.

The 2008 Paralyzed Veterans of America (PVA) Early Acute Management in Adults with Spinal Cord Injury: A Clinical Practice Guideline for Health-Care Professions state ‘Provide appropriate nutrition when resuscitation has been completed and there is no evidence of ongoing (spinal) shock or hypoperfusion’. The PVA recommendations do endorse the determination of energetic requirements for nutritional support using a 30-min energy expenditure measurement by indirect calorimetry. These guidelines for acute SCI do not define or provide a reference for what ‘appropriate nutrition’ entails, and hospitals and inpatient rehabilitation facilities do not use or often have access to metabolic carts, and insurance plans do not cover the cost of indirect calorimetry. Similarly, the Academy of Nutrition and Dietetics (AND) recommends the use of indirect calorimetry during the acute phase of SCI. They state ‘actual energy needs are at least 10 % below predicted needs⁽¹⁴⁰⁾’. But the AND recommends in the absence of indirect calorimetry to use the Harris-Benedict formula using admission weight and an injury factor of 1·1 and an activity factor of 1·2⁽¹⁴⁰⁾. This type of prediction method overestimates TDEE and subsequently energetic intake in persons with SCI, thereby leading to overfeeding^{(Reference Rodriguez, Benzel and Clevenger129,Reference Rodriguez, Clevenger and Osler143)} . Persons with SCI remain in an obligatory negative nitrogen balance, but providing excess energy content should be avoided. Overfeeding carries unique complications, such as hyperglycaemia, hypercapnia, hypertriacylglycerolaemia, uremia and obesity^{(Reference Nash, Groah and Gater8,Reference Todd, Gonzalez and Turner146)} .

With regard to chronic SCI, the PVA Consortium for Spinal Cord Medicine recently assembled an expert panel to compile the Clinical Practice Guidelines on Identification and Management of Cardiometabolic Risk after SCI (PVA guidelines; Table 7; Fig. 2)^{(Reference Nash, Groah and Gater10)}. The inaugural guidelines recommended when establishing energetic targets, all persons with chronic SCI should undergo an energetic assessment using indirect calorimetry to estimate energy expenditure and assess energy needs^{(Reference Nash, Groah and Gater10)}. Given indirect calorimetry is used to measure resting and basal metabolism, an important consideration is how to determine TDEE. In 2019, Farkas et al. ^{(Reference Farkas, Gorgey and Dolbow9)} developed a novel SCI-specific correction factor of 1·15 to estimate TDEE from BMR (or RMR) using 2·7 ml of oxygen/kg of body weight/min^{(Reference Collins, Gater and Kiratli80)}, a MET (metabolic equivalent of task) for SCI. The SCI-specific TDEE prediction equation requires validation against the gold standard respiratory chamber but provides promise. It is a novel method to estimate TDEE, and to accurately determine energetic needs in chronic SCI.

Table 7. Practical dietary recommendations with example foods to consume and avoid for persons with SCI

* The PVA guidelines do not mention seeds; however, the authors are including seeds in the nut category.

† Limit to special occasions (i.e., birthdays, weddings, holidays, etc.)

‡ For persons with limited upper extremity function, ask the butcher to trim the fat at the supermarket.

§ For persons with hypertension, although the authors recommend adopting ≤ 2400 mg of Na for all individuals regardless of hypertension status given the elevated consumption of Na-dense foods reported in the literature.

Fig. 2. Sequential dietary recommendations for persons with a spinal cord injury (SCI). First, BMR or RMR should be annually measured with indirect calorimetry or estimated using SCI-specific predictions equations (Nightingale and Gorgey^{(Reference Nightingale and Gorgey116)}, Chun et al. ^{(Reference Chun, Kim and Shin105)} or Buchholz et al. ^{(Reference Buchholz, McGillivray and Pencharz79)}) when indirect calorimetry is unavailable. Second, total daily energy expenditure should be estimated as the product of BMR or RMR and 1·15 for persons with SCI using the Farkas et al. ^{(Reference Farkas, Gorgey and Dolbow9)} equation. Third, a registered dietician should oversee a healthy dietary pattern following the Clinical Practice Guidelines on Identification and Management of Cardiometabolic Risk after SCI and additional recommendations provided in this review^{(Reference Nash, Groah and Gater10)}.

Across the chronic SCI literature, energetic intake ranges from 1212 to 2732 kcal/d (Table 5)^{(Reference Farkas, Gorgey and Dolbow9,Reference Monroe, Tataranni and Pratley34,Reference Gorgey, Caudill and Sistrun84,Reference Gorgey, Khalil and Gill86,Reference Liu, Spungen and Fink92,Reference Gorgey, Mather and Cupp108,Reference Gorgey, Martin and Metz109,Reference Nightingale, Williams and Thompson115,Reference Perret and Stoffel-Kurt117,Reference Mollinger, Spurr and el Ghatit130,Reference Groah, Nash and Ljungberg135,Reference Sabour, Soltani and Latifi141,Reference Allison, Beaudry and Thomas147–Reference Barboriak, Rooney and El Ghatit157)} , seemingly appearing to be in energy balance when evaluating TDEE (1332 to 2728 kcal/d). At face value, an energy balance does not appear to be congruent with the reported high rates of obesity^{(Reference Nash, Groah and Gater8,Reference Farkas, Gorgey and Dolbow9,Reference Gater, Farkas and Berg11,Reference Gater, Farkas and Dolbow29,Reference Farkas, Gorgey and Dolbow60)} . However, in a recent meta-analysis by Farkas et al. ^{(Reference Farkas, Pitot and Berg19)}, the authors reported in a sample of 606 persons with chronic SCI, a pooled energetic intake of 1876 kcal/d and a pooled RMR of 1492 kcal/d. Estimating a TDEE of 1716 kcal/d (using RMR × 1·15^{(Reference Farkas, Gorgey and Dolbow9)}), there is a positive energy balance of over 150 kcal/d. This is further supported by the additional work by Farkas and colleagues^{(Reference Farkas, Gorgey and Dolbow9)}. The authors reported a greater energetic intake in persons with tetraplegia compared with paraplegia when adjusting energy intake by body weight, thereby accounting for body composition that is significantly different by injury level^{(Reference Farkas, Gorgey and Dolbow9,Reference Gorgey and Gater107,Reference Inskip, Plunet and Ramer158,Reference Spungen, Adkins and Stewart159)} . The authors concluded that their findings may explain why persons with tetraplegia had significantly great percentage body fat relative to paraplegia^{(Reference Farkas, Gorgey and Dolbow9)}. Collectively, these data provide support that persons with chronic SCI may overconsume relative to their need, and this may contribute to the high rates of neurogenic obesity. However, the findings are subject to how energetic intake was operationalised.

While dietary recalls, dietary diaries/records/logs and food frequency questionnaire (FFQ) have demonstrated a strong agreement amongst themselves, research has identified systematic misreporting errors for all of the self-reported dietary instruments^{(Reference Ravelli and Schoeller160)}. Most of the literature in both acute and chronic SCI describe using dietary assessments without indicating self-report or whether a registered dietitian administered or reviewed the instrument (Table 5). In fact, 40 % of the investigators specify the use of self-report dietary assessment techniques, whereas 11 % and 9 % indicated measuring the food consumed off the plate or if a registered dietitian performed the assessment, respectively (the remaining studies provided insufficient detail to clearly determine assessment methods). It is well established that dietary assessment methods underreport true energetic consumption in persons without SCI^{(Reference Ravelli and Schoeller160,Reference Subar, Freedman and Tooze161)} , and a similar phenomenon is likely present in the population with SCI^{(Reference Farkas, Gorgey and Dolbow9,Reference Gorgey, Caudill and Sistrun84)} . Moreover, self-report after SCI becomes a challenge, especially with higher levels of injury. Persons with tetraplegia may have difficulty writing down, especially in detail, intake data and may limit, and in some cases omit, what they ate or drank on their assessment instruments. Portion size, food preparation and cooking details may be omitted and are details that can greatly influence the energy content of food. Family members or caregiver(s) may introduce a source of error by recording their food. Accordingly, it is likely that, as with the population without SCI^{(Reference Subar, Freedman and Tooze161)}, energetic intake is being underreported. Future large-scale studies with more stringent dietary assessment and testing methods are needed to examine the energetic need relative energy expenditure after SCI.

Macronutrient intake

Macronutrients are dietary constituents that provide energy. They include protein, carbohydrates, fats and alcohol. Although alcohol is considered a macronutrient and provides energy, it is not needed for survival.

Micronutrients

Micronutrients include vitamins and minerals that are required for cellular communication, water and nutrient transport, the structural integrity of bones, wound healing, and acid–base balance^{(Reference Gater5)}. Several micronutrient intakes are within inadequate ranges in persons with SCI according to the recommended guidelines established by the DGA, IOM, PHE and ADG (Table 6)^{(Reference Farkas, Pitot and Berg19,Reference Perret and Stoffel-Kurt117,Reference Groah, Nash and Ljungberg135,Reference Walters, Buchholz and Martin Ginis152,Reference Levine, Nash and Green162,Reference Lieberman, Goff and Hammond164–Reference Krempien and Barr166)} . Although others have reported below recommended intake values of vitamins A, B₅, B₇, C, D and E in individuals with chronic SCI^{(Reference Perret and Stoffel-Kurt117,Reference Groah, Nash and Ljungberg135,Reference Walters, Buchholz and Martin Ginis152,Reference Doubelt, de Zepetnek and MacDonald163,Reference Tomey, Chen and Wang165,Reference Sabour, Larijani and Vafa175,Reference Javidan, Sabour and Latifi187)} , as well as below-recommended intake in the minerals Ca, Mg and K^{(Reference Perret and Stoffel-Kurt117,Reference Groah, Nash and Ljungberg135,Reference Walters, Buchholz and Martin Ginis152,Reference Levine, Nash and Green162,Reference Doubelt, de Zepetnek and MacDonald163,Reference Sabour, Larijani and Vafa175,Reference Javidan, Sabour and Latifi187)} . In the meta-analysis by Farkas et al. ^{(Reference Farkas, Pitot and Berg19)}, the authors reported below recommended intakes for vitamins A, B₅, B₇, B₉, D and E, and the minerals K and Ca in persons with chronic SCI according to the DGA report. The authors also found excess intake of vitamins B₁, B₂, B₃, B₁₂ and K, and the minerals Cu, P, Zn and Na according to recommendations^{(Reference Farkas, Pitot and Berg19)}. Na is one of the most widely studied micronutrients after SCI with an average consumption ranging from 2402 to 4300 mg^{(Reference Perret and Stoffel-Kurt117,Reference Groah, Nash and Ljungberg135,Reference Silveira, Winter and Clark150,Reference Walters, Buchholz and Martin Ginis152,Reference Levine, Nash and Green162,Reference Tomey, Chen and Wang165,Reference Krempien and Barr166)} . PVA guidelines recommend Na consumption ≤ 2400 mg/d for all persons with SCI and hypertension. We argue that the high consumption of Na and the prevalence of hypertension in the population^{(Reference Nash, Groah and Gater8,Reference Gater, Farkas and Berg11,Reference Gater, Farkas and Dolbow29)} necessitate the need to implement a Na intake ≤ 2400 mg/d for all with SCI.

Three studies have evaluated vitamin and mineral supplementation after SCI. Opperman et al. ^{(Reference Opperman, Buchholz and Darlington188)} reported that nutritional supplementation was common in individuals with long-standing SCI, but no common characteristics (e.g., sex, LOI, age, education, etc.) distinguished users from non-users. According to the authors, 71 % of the sample reported using supplements at least once, with approximately 51 % being classified as consistent supplement users at least twice across the three time points assessed in the study. Both Opperman et al. ^{(Reference Opperman, Buchholz and Darlington188)} and Walters et al. ^{(Reference Walters, Buchholz and Martin Ginis152)} reported that participants with SCI consumed a micronutrient supplement in the form of Ca, a multivitamin, or vitamin D. The latter authors also observed vitamin C supplementation in their participants^{(Reference Walters, Buchholz and Martin Ginis152)}. Similarly, Wong and colleagues^{(Reference Wong, Graham and Green189)} reported that the three most prescribed supplements for persons with SCI were multivitamins, vitamins B and vitamin D at an SCI centre. The same authors noted that micronutrient supplementation was significantly associated with age, nutrition risk and serum albumin concentration^{(Reference Wong, Graham and Green189)}. Ca and vitamin D supplementation are important for bone health given the high prevalence of osteopenia and osteoporosis in persons with SCI^{(Reference McMillan, Nash and Gater190)}. Furthermore, vitamins B and C deficiencies are linked to anaemia and impaired wound healing^{(Reference Bird, Murphy and Ciappio191)}, both of which are reported at high rates after SCI^{(Reference Frisbie192,Reference Kruger, Pires and Ngann193)} . Multivitamin and mineral supplementation can correct deficiencies, but if there are no deficiencies, they can potentially be harmful. Persons with SCI should be prescribed vitamin and mineral supplements only if specific deficiencies have been detected or to prevent them (such as vitamin D and Ca for bone density), minimising toxicity risk. Health care professionals should place a greater emphasis on following a healthy dietary pattern (described below) to naturally consume vitamins and minerals rather than relying on supplements that may not completely correct deficiencies and carry some risk (e.g., toxicity).

Dietary recommendations after spinal cord injury

Persons with SCI are instructed to adopt a healthy diet^{(Reference Gorgey, Caudill and Sistrun84,Reference Khalil, Gorgey and Janisko194)} . But what is a healthy diet for this population? Evidence-based guidelines to ameliorate the risks of obesity and CMS did not exist for the SCI population until recently. The PVA guidelines are the first comprehensive publication to provide data-driven recommendations on healthy eating for individuals with SCI. When considering the PVA guidelines, one should note the small sample sizes in SCI literature relative to non-disabled research in diet and nutritional status. This inherent delimitation of clinical research targeting a niche patient population restricts the depth of conclusions that can be drawn from a truly evidence-based approach. Therefore, the inaugural iteration of the PVA guideline recommendations corresponds to the several current recommendations for identifying and managing CMS in the non-disabled population. However, the guidelines also factor in the alterations in the body composition and the unique endometabolic physiology that accompany SCI.

The PVA guidelines recommend energetic assessment utilising indirect calorimetry to determine energy expenditure and assess energy needs to implement a heart-healthy nutrition plan focusing on vegetables, fruits, poultry, fish, low-fat dairy products, whole grains, legumes, nuts and non-tropical vegetable oils, while limiting sweets, sugar-sweetened beverages and red meats (Table 7; Fig. 2). The PVA report also limits dietary saturated fat to 5 % to 6 % of the total energetic intake and limits daily Na intake to ≤ 2400 mg for individuals with hypertension (Table 7; Fig. 2)^{(Reference Nash, Groah and Gater10)}.

A reduced emphasis should be placed on limiting macronutrients in diets with persons with SCI, but rather focus on providing a healthy dietary pattern (Table 7; Fig. 2). The authors of this review further recommend adopting ≤ 2400 mg of Na for all individuals, regardless of hypertension status given the elevated consumption of Na-dense foods reported in the literature. We also emphasise the importance of lean poultry consisting of a moderate 3 to 4 oz portion, and the consumption of fish two times per week. Vegetables should be consumed between three and four servings per d. They should consist of the five vegetable subgroups (including dark green, red and orange, legumes (beans/peas), starchy, and others). Fruits should favour whole fruits at three least servings per d and significantly limit 100 % fruit juice because of their added/high sugar content and limited fibre content. Emphasis should be placed on low-fat dairy products in the form of milk, yogurt and cheese in small amounts while limiting saturated fat intake below 5–6 %. High-fat, sugary-based sweets and drinks should be replaced with fresh fruit and water, respectively. Flavoured and unflavoured carbonated water and zero energy liquid water enhancers can be used to provide variety and flavour to drinks. Red meat and sweets should be consumed only on special occasions such as birthdays, weddings, holidays, etc. By following the above-referenced dietary patterns, persons with SCI will naturally limit their intake of refined/simple carbohydrates, Na, and saturated fat and increase the consumption of unsaturated fats and fibre. Such dietary patterns will also promote optimal ingestion of micronutrients.

We recommend and recognise the significance of annual dietary assessments (minimally) and nutrition education with a registered dietitian as part of the medical assessment and standard of care for individuals with SCI (Fig. 2). It is our recommendations that in addition to assessing body composition by 3- or 4-compartment modelling^{(Reference Gater, Farkas and Dolbow29)}, registered dietitians should: (1) assess resting metabolism through indirect calorimetry, or, when unavailable, assess RMR/BMR with the Nightingale and Gorgey^{(Reference Nightingale and Gorgey116)}, Chun et al. ^{(Reference Chun, Kim and Shin105)} or Buchholz et al. ^{(Reference Buchholz, McGillivray and Pencharz79)} SCI-specific prediction equation (2) calculate TDEE using the prediction equation and correction factor by Farkas et al. ^{(Reference Farkas, Gorgey and Dolbow9)} to determine accurate energy needs, (3) encourage adherence to the SCI dietary guidelines detailed above and balanced nutrition (Table 7, Fig. 2) as an overall healthy lifestyle choice, (4) prescribe dietary supplements when specific vitamin and/or mineral deficiencies have been detected or to prevent them when appropriate nutrition/healthy dietary patterns are unsucessful, and (5) explore dietary anomalies specific to SCI, such as avoidance of food groups that may affect bowel/bladder function post-injury. Periodic assessments with the health care team, including a registered dietitian, should be implemented to manage neurogenic obesity, cardiometabolic risk and nutrition status after SCI and allow the individual themselves to take an active role in their overall health.

Future research

Expert advice advocates for pairing diet and physical activity to mitigate neurogenic obesity and reduce the cardiometabolic risk after SCI. However, the interaction of feeding and activity in SCI has only recently begun to be studied and is likely unique in this population. Furthermore, several barriers, such as transportation, overuse injuries, pain, access to facilities, financial restraints, educational knowledge, disability/SCI-specific resources on exercise, and fear of musculoskeletal or integumentary injury, can limit exercise/physical activity engagement. To mitigate barriers to physical exercise and facilitate improvements in overall health, research focused exclusively on dietary intervention may provide a large-scale ‘cure’ to chronic diseases in persons with SCI.

Rather than a nutrient (i.e., low-fat diets) focus, nutrition research is turning its attention to dietary patterns as we consume a whole range of foods and food groups, not just nutrients. Evidence suggests that vegetarian and/or plant-based diets create a negative energy balance and decrease the risk of obesity, diabetes and other chronic health ailments^{(Reference Tonstad, Butler and Yan195,Reference Fardet and Boirie196)} . Fardet and Boirie^{(Reference Fardet and Boirie196)} examined diet and chronic disease risk from over 300 meta-analyses and systematic reviews published in the last 63 years. The authors reported that plant-based foods were more protective against the risk of developing chronic disease compared with animal-based foods, reinforcing fundamental dietary patterns for good health. Among plant-based foods, whole-grain-based foods had a small edge over fruits and vegetables, while for animal-based foods, dairy products overall were considered neutral on health, and fish was considered protective. Red and processed meats were correlated with elevated chronic disease risk as were sugar-sweetened beverages and highly refined low-fibre grains^{(Reference Fardet and Boirie196)}. These data were further supported by evidence published by the AHA that diets higher in plant-based foods and lower in animal-based products were associated with a lower risk of cardiovascular morbidity and mortality^{(Reference Kim, Caulfield and Garcia-Larsen197)}. Plant-based diets may provide an alternative nutritional plan for cardiometabolic risk factors after SCI and minimise overeating by enhancing satiety relative to their need. Future research should focus on the effects of such diets in the SCI population but also attend to energy balance.

Final considerations

In summary, the loss of metabolically active tissue below the LOI, endometabolic pathophysiology, environmental and physical barriers, and poor dietary choices contribute to the suboptimal dietetics, poor body habitus and cardiometabolic risk in SCI^{(Reference Nash and Gater198)}. Individuals with SCI meet several evidence-based recommendations published by authoritative guidelines for non-disabled individuals, although they are likely underestimated or overestimated for this population. These guidelines do not account for SCI-induced changes in body composition^{(Reference Gorgey, Dolbow and Dolbow199,Reference Gater, Farkas and Taylor200)} , reduced metabolic requirements^{(Reference Farkas, Pitot and Berg19,Reference Monroe, Tataranni and Pratley34,Reference Buchholz, McGillivray and Pencharz79,Reference Bauman, Spungen and Wang87,Reference Buchholz, McGillivray and Pencharz121)} , gut dysmotility^{(Reference Holmes41,Reference Keshavarzian, Barnes and Bruninga201)} and sympathetic nervous system dysfunction^{(Reference Stjernberg, Blumberg and Wallin33,Reference Monroe, Tataranni and Pratley34)} and need to be interpreted and used with caution. The PVA guidelines are currently the strongest evidence-based dietary guidelines for the population with SCI and should be followed by persons with SCI and their health care team. Stakeholders and practitioners should have an understanding that the currently limited evidence-based means that guidelines contain might not fully capture the unique nutritional needs of persons with SCI. Therefore, clinical nutritional strategies should also rely on strong inferences from existing studies and use routine monitoring of individual responses to interventions. Because diet-related comorbidities are related to anatomical and physiological changes after SCI, annual nutritional analysis and indirect calorimetry (or, minimally using an SCI-specific BMR/RMR prediction method) with a registered dietitian are encouraged in clinical practice. Future multicentre controlled trials are needed (with less reliance on cross-sectional study design) to collect large data sets on energy expenditure, energy needs and total energetic intake after SCI. These data are needed to help develop comprehensive, evidence-based dietary reference values for nutrients specific for persons with SCI aimed at reducing the secondary complications of the injury.