16b.1    Assessment of protein status

A comprehensive assessment of protein status requires dietary intake data, and selected anthro­pometric and bio­chem­ical measures. In chronic protein deficiency, if facilities are available, technological methods such as dual-energy X-ray absorptiometry (DXA), ultra­sound, and bioelectrical impedance analysis (BIA) can also be used to quantify loss of skeletal muscle mass. No single param­eter is suffi­ciently reliable to assess protein nutri­tional status.

Protein turnover is a contin­uous process by which all body proteins are broken down and resyn­the­sized. The balance of break­down and synthesis reaches a steady state when there is no net gain or loss of protein. With inad­equate protein intake and/or a diet limited in indis­pens­able amino acids, the balance shifts towards tissue catabolism to provide sufficient amino acids for endoge­nous require­ments. Short-term inad­equate intake leads to protein losses from the visceral protein pool (i.e. serum proteins, erythro­cytes, granulo­cytes, lym­pho­cytes, etc (de Blaauw et al., 1996). Traditionally, the visceral protein pool has been quantified through the measure­ment of circulating visceral serum proteins, details of which are outlined below.

In chronic protein deficiency, however, muscle proteins undergo proteolysis to supply the necessary amino acids for enhanced protein synthesis. There­fore, there is loss of both skeletal muscle mass and muscle function (Hansen et al., 2000). Several methods are available for measuring skeletal muscle mass and function; they vary in complexity, availability and cost and are also discussed below.

Charac­ter­istics of the visceral serum proteins used traditionally to assess protein nutri­tional status in both com­mu­nity settings in low‑income countries and in hospitalized patients are summarized in Box 16b.1. The main site of synthesis for most of the visceral proteins is the liver.

Theoretically, visceral protein bio­markers have three potential roles, namely diagnosis of protein-energy malnu­trition, monitoring of therapeutic inter­ven­tions , and the provision of information on prognosis (i.e., predicting outcome) on which to decide whether to intervene therapeutically (Livingstone, 2013). To accomplish these roles, an ideal bio­marker should have the following characteristics listed in Box 16b.2 and summarised by Livingstone (2013).

Features of an ideal protein-energy visceral protein bio­marker

Unfortunately, none of the visceral protein bio­markers listed in Box 16b.1 meet all the criteria listed in Box 16b.2. This has led to controversy over their use, espe­cially in clinical settings where acute and chronic disease is common. In such settings visceral serum proteins act as negative-acute phase proteins, declining in response to inflam­mation due to the influence of cytokines, regardless of malnu­trition. As it is not possible to separate the influence of inflam­mation from that of malnu­trition, the American Society for Parenteral and Enteral Nutrition has deemed the available visceral serum proteins of limited use for assessing protein-energy malnutrition. Instead, they emphasize that visceral serum proteins should only be recog­nized as inflam­matory bio­markers asso­ciated with "nutrition risk" (Evans et al., 2019).

The current consensus is that visceral serum proteins should always be used alongside bio­markers of inflam­mation and as a complement to other methods of nutri­tional assess­ment such as a nutrition-focused physical examination (Bharadwaj et al., 2016). Several nutrition-focused physical examination assess­ment tools are available for diagnosing malnu­trition in hospitalized patients. See Keller (2019) for more details. Whether visceral serum proteins can be used to diagnose protein-energy nutri­tional status in the absence of inflam­mation is the subject of active investigation.

Currently, the bio­marker most widely used to identify the presence of inflam­mation in both com­mu­nity and hospital settings is serum C‑reactive protein (CRP). It is a useful bio­marker because it responds rapidly to changing clinical conditions, is widely available, and has high sensitivity (Evans et al 2019). Measuring CRP concurrently with visceral serum proteins allows for some differen­tiation between the contribution of inflammation versus protein-energy malnutrition when interpreting decreased visceral serum protein levels (Bharadwaj et al., 2016). The importance of adjusting for the confounding effects of inflam­mation on bio­markers in low‑income countries has led to a new approach using regres­sion correction to adjust for the confounding effects of inflam­mation termed the "BRINDA inflam­mation adjust­ment method." For the BRINDA method, CRP alone, alpha‑1‑acid glyco­protein (AGP) alone, or both AGP and CRP together can be used. So far, the BRINDA inflam­mation adjust­ment method has been applied only to bio­markers of micronutrient status (i.e., iron, zinc, and vitamin A). However, in the future this method may be modified for use with visceral serum proteins when used in settings where inflam­mation co‑exists. For details of a practical guide on applying the BRINDA adjust­ment method, see Luo et al. (2023).

All the serum visceral proteins are affected to varying degrees by several other factors besides inflam­mation, which together reduce their specificity; these are summarized in Box 16b.2 and discussed in more detail below.

16b.1.1 Total serum protein

This bio­marker measures the total protein content in blood serum and includes albumin and globulins. It is easily measured and in the past was used an index of visceral protein status in national nutrition surveys, espe­cially in low-income countries (Evans et al 2019). However, it is now known to be a poor bio­marker of protein status. Many factors influence total serum protein concen­tra­tions as shown in Box 16b.2, compromising their specificity and sensitivity. An additional factor often impacting on total serum protein levels in severely ill hospital patients is the administration of blood products such as albumin. A recent system­atic review and meta-analysis of older adults found only a moderate associ­ation between total serum protein levels and risk of malnu­trition in a non‑acute population, defined by validated nutrition assess­ment tools, such as subjective global assess­ment (SGA); although other bio­markers evaluated were said to be superior (Zhang et al., 2017).

16b.1.2 Serum albumin

The primary protein in human serum is albumin (between 3‑5g/kg body weight) which has been used in the past as a bio­marker of protein nutri­tional status. However, serum albumin concen­tra­tions only reflect changes within the intra­vascular space and not the total visceral protein pool. As such, it is a poor nutri­tional bio­marker. Albumin has a large pool size and a long half-life (14‑20d) so it is insensitive to short-term changes in protein status and thus not appro­priate for assessing acute changes in nutri­tional status. More­over, in cases of both acute and chronic illness, serum albumin concen­tra­tions may be main­tained due to a compen­satory reduc­tion in albumin catabolism. For example, in the anorexia nervosa population, normal or only slightly decreased serum albumin levels have been observed, despite the presence of severe protein-energy malnu­trition, low muscle mass and reduced muscle strength (Lee et al., 2015).

Albumin plays a signif­icant role in main­taining colloidal osmotic pressure (Levitt & Levitt, 2016) so in an inflam­matory state albumin leaks out of the intra­vascular space resulting in a measurable decrease in serum values (Evans et al., 2019). Even a well‑nourished person can develop hypo­albuminemia (low albumin) within hours of a traumatic injury or acute illness due to inflam­mation and fluid shifts (Soeters et al., 2019). As albumin leaves the intra­vascular space, interstitial edema can develop which further dilutes measured serum values.

Albumin also provides more than 50% of the antioxidant activity in plasma (Taverna et al., 2013) so it binds to pro‑oxidative substances and is rapidly degraded. The half‑life of albumin may be shortened during an inflam­matory period, further contributing to reduced serum levels.

The use of serum albumin concen­tra­tions as a prognostic index for the devel­op­ment of post­operative compli­cations has been extensively studied. Adults under­going spinal surgery with pre‑operative hypo­albuminemia were reported to experience an increase in post­operative complications (hospital length of stay, surgical site infec­tions and readmission rates) (Chaker et al., 2024). Examination of the National Surgical Quality Improvement Program (NSQIP) database has also revealed increased 30d morbidity and mortality rates in patients with pre‑operative hypo­albuminemia following a variety of other surgical procedures including pancreatic-duodenectomy surgery (Sawchuk et al., 2024), esophagectomy (Li & Zhu, 2024), and head and neck microvascular surgery (Xu et al., 2024). Whether compli­cations could be improved with pre‑operative nutri­tional and/or medical inter­ven­tions is uncertain due to the retro­spec­tive nature of these studies. In many cases, however, surgery was the only possible inter­ven­tion to improve the under­lying inflam­matory processes.

Several other studies attempting to evaluate whether intense nutri­tional inter­ven­tions in acute or chronically ill patients with hypo­albuminemia improved low serum albumin levels and/or clinical outcomes were unable to demonstrate any signif­icant improvements. The 2021 American Society of Parenteral and Enteral Nutrition (ASPEN) position paper discourages the use of serum albumin as a proxy measure of total body protein and as a bio­marker of nutri­tional status in clinical settings (Evans et al 2019).

Factors affecting serum albumin

Age influences serum albumin, with concen­tra­tions rising until the second or third decade of life and then declining thereafter, possibly related to a reduc­tion in the rate of albumin synthesis. In a 2017 review and meta-analysis of malnu­trition risk in older adults (n=111), albumin levels <35g/dL were asso­ciated with reduced nutri­tional status. However, the authors suggest that albumin was insensitive as a screening bio­marker in this population (Zhang et al., 2017).

Sex also influences serum albumin levels, with males tending to have higher values than females, the maximum difference occurring at age 25y (Richie et al., 1999).

Pregnancy reduces serum albumin due in part to hemodilution.

Inflammation, infection, and metabolic stress elicit the acute phase response mediated by pro‑inflam­matory cytokines. Albumin is a negative acute phase reactant, so its production is decreased if there is an active infection or inflam­mation resulting in low serum albumin concen­tra­tions (Evans et al 2019).

Disease states such as some gastro­intestinal diseases (e.g., ulcerative colitis and Crohn's disease) and renal diseases are asso­ciated with decreased serum albumin concen­tra­tions (i.e., hypo­albuminemia) due to an increase in albumin losses either via the gastro­intestinal tract or kidneys. In liver disease, low serum albumin concen­tra­tions arise from a reduc­tion in protein synthesis, whereas in hypo­thyroidism and congestive heart failure, fluid retention and dilutional effects contribute to reduced serum albumin levels. Degradation of albumin is also reduced in hypo­thyroidism further contributing to altered albumin levels

Semi-starvation is asso­ciated with low albumin levels (i.e., hypo­albuminemia) due to a reduc­tion in protein intake which in turn reduces the ability of the liver to synthesize albumin. In some cases, however, fluid loss may occur leading to higher serum albumin concen­tra­tions (hyper­albuminemia).

Dehydration as may occur with severe diarrhea may result in hyper­albuminemia as a result of a diminished plasma volume.

Blood trans­fusions and parenteral administration of albumin may increase serum albumin, masking any changes in levels in hospital patients, thus confounding its use.

Interpretive criteria and measure­ment of serum albumin

Reference intervals are a statistically derived range of values that encompass the central 95% of values from a healthy refer­ence population. Values within this interval are assumed to represent a normal range. Bio­marker test results lying outside of the refer­ence interval suggest an abnormal result. For more discussion, see Chapter 1: Introduction; Section 1.6 Evaluation of nutri­tional assess­ment indices. Unfortunately, depending on the bio­marker, refer­ence intervals are not necessarily trans­ferable between labora­tories due to differ­ences in both the refer­ence populations (e.g., ethnic compo­sition, geographic factors, habitual diets, lifestyles) and analytical methods used.

Several countries have established refer­ence intervals for serum albumin. In the U.S, Ritchie et al. (1999) have derived refer­ence intervals for serum albumin stratified by age and gender from a cohort of over 124,000 Caucasians in the northeastern United States. Individuals with evidence of inflam­mation (i.e., CRP ≥10mg/L) were excluded. The refer­ence medians and percentiles by age and sex are presented in Table 16b.1. Measure­ments were stan­dard­ized against the Certified Reference Material (CRM) 470‑RPPHS (Whicher et al., 1994; Baudner et al., 1994). It should be used to check the accuracy of all methods used to determine serum albumin.

Table 16b.1: Reference medians and selected percentiles for serum albumin (g/L), stratified by age and gender. Data from Richie et al (1999).

	Males				Females
Age (y)	2.5th	50th	97.5th		2.5th	50th	97.5th
1.0	35.9	43.1	51.9		36.1	43.5	52.2
4.0	36.6	44.0	52.9		36.9	44.3	53.3
7.0	36.9	44.3	53.3		37.2	44.7	53.7
10.0	37.1	44.6	53.8		37.3	44.9	54.0
14.0	37.7	45.3	54.5		36.6	44.0	52.9
18.0	38.1	45.9	55.1		36.3	43.7	52.5
20.0	38.2	45.9	55.2		35.0	43.6	52.4
30.0	37.9	45.5	54.8		35.0	43.3	52.0
40.0	37.0	44.6	53.6		35.6	42.8	51.4
50.0	36.1	43.6	52.3		35.1	42.1	50.7
60.0	35.2	42.3	50.9		34.4	41.4	49.7
70.0	34.3	41.2	49.6		33.7	40.6	48.8
80.0	33.4	40.2	48.3		33.0	39.7	47.8

Reference intervals for serum albumin are also available for healthy indi­vid­uals in several other countries; see for example, Koerbin et al. (2015) and Adeli et al. (2015) for some examples. Increasingly, a multicenter study design is being used to establish common refer­ence intervals. Such an initiative is in progress to establish pediatric refer­ence intervals for 70 common chemical bio­markers (including serum albumin) in Germany, Australia, South Africa, Scandinavia, and Canada (Karbasy et al., 2015).

Serum albumin is assayed in most clinical labora­tories using an automated dye‑binding method with bromocresol green (McPherson & Everard, 1972). Other methods include standard electro­phoresis and immune-nephelo­metry. Values for serum albumin depend on the analytical method used. Hence, these differ­ences should be consid­ered when inter­preting serum albumin levels.

16b.1.3 Serum pre­albu­min or transthyretin

Pre‑albumin, historically known as trans­thyretin, is a protein synthesized in the liver which acts as a trans­port protein for thyroid hormones. Prealbumin circulates bound to retinol‑binding protein 4 (RBP 4) and its retinol ligand. Pre‑albumin has a high indis­pens­able to dispensable amino acid ratio (Beck & Rosenthal, 2002; Ranasinghe et al., 2022), a relatively short half‑life (2‑3d), and a smaller body pool size than albumin, and is not as sensitive to fluid status (Beck & Rosenthal, 2002; Bharadwaj et al., 2016). Recent research suggests that fluctuations in plasma pre­albu­min reflect both the size and alterations in lean body mass; for more details see Ingenbleek & Bernstein (2015).

Synthesis of pre­albu­min decreases as the body nitrogen pools decline following a reduc­tion in protein intake, suggesting that pre­albu­min may have potential as bio­marker of protein nutritional status. Certainly in the rat model, pre­albu­min was noted to decrease after 14 days on an intake of sixty percent dietary protein (Le Moullac et al., 1992). The response of serum pre­albu­min to dietary protein intake has also been investigated in human studies. In children with severe protein-energy malnu­trition, for example, pre­albu­min levels increased within 48 hours of protein supple­mentation and normalized within eight days of therapy (Ingenbleek et al., 1975). However, in these early studies, the possible influence of inflam­mation on the pre­albu­min response was not consid­ered. In a later system­atic review and meta-analysis of blood bio­markers asso­ciated with risk of malnu­trition in older adults (Zhang et al., 2017), a sensitivity analysis confirmed that serum pre­albu­min concen­tra­tions were markedly reduced in response to acute inflam­matory stress. In a later study of neonates in the neonatal intensive care unit (NICU), concen­tra­tions,of pre­albu­min were also negatively correlated with CRP (‑0.62; P<0.005) (Tian et al., 2018). Hence, together these findings emphasize that caution must be used when inter­preting pre­albu­min concen­tra­tions in acute health care settings.

The use of pre­albu­min as a nutri­tional bio­marker in non-inflam­matory states has also been investigated. In a system­atic review of subjects with restrictive eating disorders but were otherwise healthy with no evidence of inflam­matory illness, those with severe protein-energy malnu­trition had normal pre­albu­min concen­tra­tions that did not decline until their Body Mass Index (BMI) was <12. Based on these findings, the authors concluded that pre­albu­min concen­tra­tions cannot be used as bio­markers of protein nutri­tional status or to identify the need for nutrition support even in otherwise healthy subjects with no evidence of inflam­mation (Lee et al., 2015).

Never­theless, considerable controversy remains over the usefulness of serum pre­albu­min as a bio­marker of protein nutritional status. Investi­gations on the use of pre­albu­min as a bio­marker for monitoring the efficacy of nutrition support in clinical nutrition practice have been extensive, in part because of its short half‑life (i.e., 2‑3d). Results have been mixed. Bernstein et al. (1989) reported weekly increases in serum pre­albu­min of 40‑50mg/L in response to adequate nutri­tional support. They suggested that a response of less than 20mg/L in a week is indicative of either inad­equate nutritional support or a failure to respond to the treatment. Others have argued that if inflam­mation decreases during nutri­tional support, pre­albu­min will surely increase, so the part played by nutri­tional care will be difficult to distinguish (Casati et al., 1998). Indeed, the recent ASPEN position paper on the use of visceral proteins as nutrition bio­markers concluded that normalization of pre­albu­min during nutrition support could indicate several factors including the resolution of inflam­mation, the reduc­tion of nutrition risk, a trans­ition to anabolism, and/or potentially lower calorie and protein require­ments (Evans et al., 2019). In view of these uncertainties, authors of the ASPEN publi­cation discourage the use of pre­albu­min as a bio­marker of total body protein or total muscle mass. They also conclude that pre­albu­min is not useful for monitoring nutrition support, even for indi­vid­uals with no evidence of inflam­matory illness.

Never­theless, the ASPEN recommendations remain open to debate. Despite its recog­nized limitations, according to Delliere & Cynober (2017), plasma pre­albu­min can be useful to diagnose malnu­trition and its severity in those indi­vid­uals with no inflam­mation syndrome. Moreover, they also advocate pre­albu­min as a good bio­marker for monitoring the efficacy of the response to nutrition inter­ven­tions. In their review, they provide two algorithms, one to interpret pre­albu­min levels in patients with a CRP concen­tration indicative of the absence of inflam­mation (i.e.,<15mg/L), and a second to interpret levels in intensive care unit patients; see Delliere & Cynober (2017), for more details.

Factors affecting serum pre­albu­min levels

Age and sex influence serum pre‑albumin. Concen­trations increase linearly from birth during infant growth, plateauing in adulthood until the sixth decade, after which concen­tra­tions decline. At the onset of puberty, hormonal-induced sexual dimorphisms occur with males having higher concen­tra­tions than females which reach a plateau during full sexual maturity. These sex‑related differ­ences are main­tained during adulthood, with concen­tra­tions of 300‑330mg/L in adult males compared to 250‑270mg/L in adult females. From the sixth decade, concen­tra­tions show a stepwise decrease, with sex‑related differ­ences becoming less marked and no longer evident by the eighth decade (Ingenbleek, 2022). Prealbumin is unaffected by ethnic differ­ences or genetics (Delliere & Cynober, 2017).

Inflammation, infection, stress elicit the acute phase response resulting in a decrease in serum pre­albu­min concen­tra­tions. Several mechanisms have been proposed for the observed decline, including the existence of hepatic reprioritization of protein synthesis, possibly increased tissue catabolism, and/or redistribution of pre­albu­min because of an increase in capillary permeability (Evans et al 2019).

Disease states such as kidney disease degrade pre­albu­min so with any renal dysfunction pre­albu­min levels are increased. Conversely, pre­albu­min levels are decreased in liver disease, and dialysis because the ability of the liver to synthesize proteins, including pre­albu­min, is impaired. Likewise, signif­icant hyper­glycemia can decrease pre­albu­min levels through several mechanisms such as an increase in protein degra­dation, inflam­mation, and endothelial dysfunction. The latter causes increased permeability and loss of proteins like pre­albu­min from the blood stream (Delliere & Cynober, 2017; Bharadwaj et al., 2016). Concen­trations may also be low with zinc deficiency because zinc is essential for the synthesis of proteins such as pre­albu­min (Beck & Rosenthal, 2002).

Hyperthyroidism is asso­ciated with a decrease in serum pre­albu­min concen­tra­tions in part because the increased thyroid hormones reduce the amount of thyroxine (T4) available to bind with pre­albu­min as well as an increase in protein turnover and break­down. In contrast in hypo­thyroid states, thyroid hormones are sometimes low resulting in increased pre­albu­min levels. This is because the reduced availability of thyroxine leads to less binding with prealbumin, potentially causing an accumulation of unbound prealbumin in the blood.

Pregnancy and estrogen-containing preparations (e.g. oral contraceptive agents, estrogen replacement therapy) increase serum pre­albu­min concen­tra­tions due to increased liver protein synthesis in response to the increased demand and hormonal changes (i.e., increased levels of estrogen and progesterone).

Medications such a corticosteroids and non‑steroidal anti‑inflam­matory agents increase pre­albu­min concen­tra­tions

Changes in fluid distribution and hydration changes modify serum pre­albu­min concen­tration, increasing with acute hydration and decreasing with hemodilution (Delliere & Cynober, 2017).

Interpretive criteria and measure­ment of serum pre­albu­min

The refer­ence interval for serum pre­albu­min is reportedly 0.20‑0.40g/L, although varying with age and sex. Hence, these factors must be consid­ered for the interpretation. Table 16b.2 presents the refer­ence intervals for serum pre­albu­min (g/L) by age and gender (Richie et al., 1999).

Table 16b.2: Reference medians and selected percentiles for serum pre‑albumin (g/L), stratified by age and gender. Data from Richie et al. (1999).

	Males				Females
Age (y)	2.5th	50th	97.5th		2.5th	50th	97.5th
1.0	0.12	0.18	0.26		0.12	0.18	0.26
4.0	0.12	0.18	0.27		0.13	0.18	0.27
7.0	0.13	0.18	0.27		0.13	0.19	0.28
10.0	0.15	0.22	0.32		0.15	0.23	0.33
14.0	0.18	0.26	0.37		0.17	0.25	0.36
18.0	0.19	0.28	0.41		0.18	0.26	0.37
20.0	0.20	0.29	0.42		0.18	0.26	0.38
30.0	0.21	0.30	0.44		0.18	0.26	0.38
40.0	0.21	0.31	0.45		0.18	0.26	0.39
50.0	0.21	0.30	0.44		0.18	0.27	0.39
60.0	0.20	0.29	0.43		0.18	0.27	0.39
70.0	0.19	0.28	0.41		0.18	0.27	0.39
80.0	0.19	0.27	0.40		0.18	0.26	0.39

Cut-offs for serum pre­albu­min applied for the diagnosis of protein malnu­trition vary across countries. In France and Poland, concen­tra­tions of 0.10g/L and 0.05g/L have been frequently applied to indicate protein malnu­trition and severe protein malnu­trition, respectively. A recent two-part investigation, based first on a large retro­spec­tive estimation to refine the two cutoffs, followed by a prospective validation study, has redefined cutoff values asso­ciated with malnu­trition based on Receiver Operating Charac­ter­istics (ROC) curve analysis. In this study, serum pre­albu­min concen­tra­tions indicative of malnu­trition and severe malnu­trition were reported as 0.17g/L (sensitivity 0.68; specificity 0.67) and 0.12g/L (sensitivity 0.53; specificity 0.73), respectively (Delliere et al., 2021). However, the Area Under the Curve (AUC) values obtained from the prospective study were both relatively low (0.674 and 0.609) highlighting that pre­albu­min is not an optimal bio­marker to define protein malnu­trition and severe malnu­trition.

Assay of serum pre­albu­min is inexpensive and easy. Of the two methods available, immune-nephelometry is more accurate than immuno­turbidimetry and was the method used in the cut-off study of Delliere et al (2021). Immuno-nephelometry is consid­ered the refer­ence method. Assay kits are available for both methods which can operate on automated chemistry analyzers.

16b.1.4 Serum retinol‑binding protein

Retinol binding protein (RBP) is the serum protein responsible for the trans­port of retinol throughout the blood stream and is bound to pre­albu­min. Of the visceral proteins, retinol binding protein has the shortest half‑life (approx­imately 12h) with a relatively small body pool. Retinol binding protein is catabolized in the kidneys and is elevated with renal failure. Some of the factors affecting serum RBP are similar to those outlined for pre­albu­min, including the decrease in serum RBP concen­tra­tions during the acute phase response. Unlike pre­albu­min, however, RBP is signif­icantly impacted by both vitamin A and zinc status, low levels of these micronutrients inhibiting mobilization of RBP in the liver. Hence, these are additional factors limiting RBP as a useful protein bio­marker, limiting its use (Bharadwaj et al., 2016); Keller, 2019).

16b.1.5 Serum crea­ti­nine

Serum crea­ti­nine is a derivative of the skeletal muscle protein creatine phosphate. Creatine is a nitrogenous organic acid made up of three amino acids, glycine, arginine, and methionine which is produced mainly in the liver and kidney, and then trans­ported to skeletal and heart muscles. Once in the tissues, creatine is phosphorylated to creatine phosphate, a readily available energy source used for muscle contractions (Kashani et al., 2020).

Creatinine in serum is produced from the spontaneous, non‑enzymatic break­down of creatine in the skeletal muscle cells. This process is done contin­uously at a nearly constant rate in healthy subjects and is irreversible. Conse­quently, the body creatine pool must be replenished daily by the intake of indis­pens­able amino acids. As serum crea­ti­nine originates mainly from the creatine in skeletal muscle cells, the circulating level of crea­ti­nine has been consid­ered a potential bio­marker reflecting muscle mass, provided kidney function is stable and under the steady state. However, many other factors besides muscle mass influence serum crea­ti­nine concen­tra­tions, and the contribution of muscle mass to the variation in serum crea­ti­nine appears small.

Several studies have shown serum crea­ti­nine can serve as a bio­marker of skeletal muscle in healthy adults. For example, Baxmann et al. (2008), in a study of healthy adults classified as sedentary or having mild or moderate / intense physical activity, observed serum crea­ti­nine was signif­icantly correlated with lean body mass measured by bioelectrical impedance analyses (BIA), after adjust­ment for protein / meat intake and physical activity. Several other investigators have reported that serum crea­ti­nine is also a surrogate marker of skeletal muscle mass in dialysis patients, and that low lean body mass in the dialysis population is asso­ciated with increased mortality; see Patel et al. (2012) for a table summarizing these studies.

Table 16b.3: Differences in the lean-body mass, using modified Bland-Altman tests, comparing DXA‑measured LBM and each of the three estimates of the LBM using serum crea­ti­nine, MAMC and HGS in 118 HD patients. *Pearson correlation between difference and DXA values.
DXA Dual energy X‑ray absorp­tiom­etry, MAMC mid‑arm muscle circum­ference, HGS handgrip strength, SCr serum crea­ti­nine. Data from Patel et al. (2012)

LBM (kg) estimated by variables	Limits of agreement	Mean difference 95% CI	Correl- ation* (r)	Correl- ation (p)
SCr	−6.7 to 6.7	0.0 (−0.6,0.6)	0.34	<0.001
MAMC	−7.0 to 7.0	0.0 (−0.7, 0.7)	0.35	<0.001
HGS	−7.0 to 7.0	0.0 (−0.6, 0.6)	0.35	<0.001

These studies have used a variety of methods to measure lean body mass, including BIA, a portable near-infrared technique, and dual‑energy X‑ray absorp­tiom­etry (DXA). For example, in a study of 118 long‑term hemodialysis patients Patel et al. (2012) showed that the 3‑month averaged serum crea­ti­nine correlated well with the DXA‑measured lean body mass, as shown in Table 16b.3.

Recently, Bistrain et al. (2020) have advocated for the application of serum crea­ti­nine to identify patients likely to be at high risk of protein-energy malnu­trition and for whom assess­ment of height and collection of a 24h urine sample for calculating the crea­ti­nine height index (CHI) is impractical. They have compiled a table (see Table 16b.4) of serum crea­ti­nine levels for males and females of various heights that correspond to 60% of the expected CHI. Any patient whose admission values indicated a normal serum urea nitrogen (SUN) (i.e., <20mg/dL) to indicate the presence of reasonable renal function on admission but a serum crea­ti­nine at or below the level indicated for their sex and height (see Table 16b.4) could be presumed to have severe protein-energy malnu­trition. Bistrain et al. (2020) caution that this approach cannot be used for patients with some degree of renal dysfunction (i.e., a SUN >20mg/dL).

Table 16b.4 Serum creatinine levels indicating a creatinine height index of 60% of expected. Modified from Bistrain et al. (2020).

Males				Females
Height (cm)	60% Expec- ted Urine Creatinine (mg)	Serum Creatinine (mg/dL)		Height (cm)	60% Expec- ted Urine Creatinine (mg)	Serum Creatinine (mg/dL)
152.4	773	0.7		147.3	498	0.4
154.9	795	0.7		149.9	511	0.4
157.5	815	0.7		152.4	525	0.5
160.0	832	0.7		154.9	540	0.5
162.5	856	0.7		157.5	555	0.5
165.1	880	0.8		160.0	569	0.5
167.6	910	0.8		162.5	586	0.5
170.2	933	0.8		165.1	604	0.5
172.7	958	0.8		167.6	626	0.5
175.3	985	0.9		170.2	646	0.6
177.8	1015	0.9		172.7	665	0.6
180.3	1043	0.9		175.3	685	0.6
177.8	1071	0.9		177.8	704	0.6
180.3	1099	1.0		180.3	724	0.6
182.9	1135	1.0		182.9	744	0.6

De Rosa et al. (2023), however, highlight the many factors that may influence serum crea­ti­nine concen­tra­tions, limiting its use as a bio­marker for muscle mass; these are summarized below.

Factors affecting serum crea­ti­nine

Age and sex influence serum crea­ti­nine with lower levels in the elderly and in females than males, both asso­ciated with lower muscle mass.

Dietary intake of creatine from meat is converted to crea­ti­nine after cooking, so a large meal of cooked meat may increase serum crea­ti­nine. High doses of creatine supple­ments by indi­vid­uals with risk factors for kidney disease may also lead to elevated serum crea­ti­nine concen­tra­tions. Dietary protein also has a small effect because protein is the main source of the amino acid precursors of creatine.

Ethnicity can influence serum crea­ti­nine due to differ­ences in muscle mass and genetic factors. For example, black African Americans often have a higher muscle mass resulting in higher serum crea­ti­nine levels than Caucasians.

Intravenous fluid administration can also impact levels. High fluid volumes dilute serum crea­ti­nine levels, thus reducing concen­tra­tions, whereas fluid restriction (e.g., dehydration) results in elevated levels (Kashani et al., 2020).

Pregnancy is asso­ciated with low serum crea­ti­nine concen­tra­tions reflecting a physio­log­ically increased glomerular filtration rate.

Physical activity increases muscle mass which is reflected by a higher serum crea­ti­nine concen­tration. Bouts of strenuous exercise can cause trans­ient elevations in crea­ti­nine levels due to muscle break­down.

Chronic renal failure is asso­ciated with an elevated serum crea­ti­nine reflecting a decrease in glomerular filtration rate. Hence, serum crea­ti­nine cannot be used as a nutri­tional bio­marker in those with reduced renal function. In fact, crea­ti­nine is used as an indicator of kidney function with high serum levels representing reduced kidney function.

Other disease states such as advanced liver disease reduce serum crea­ti­nine due to decreased crea­ti­nine production arising from a reduc­tion in hepatic creatine synthesis.

Numerous drugs compete with tubular trans­port mechanisms and inhibit crea­ti­nine secretion, leading to a rise in serum crea­ti­nine. Examples include Probenecid, Cimetidine, Trimethoprim, and possibly fenofibrate (Levy & Inker, 2017).

Sarcopenia index

The sarco­penia index (SI) is a simplified approach that aims to predict skeletal muscle mass by comparing metabolites of different cellular origins (Barretto et al., 2019; Kashani et al., 2017). It is consid­ered more reliable for estimating muscle mass than serum crea­ti­nine alone. The index is calculated as follows:

Sarcopenia Index= [(serum crea­ti­nine/serum cystatin C) × 100]

Both serum crea­ti­nine and Cystatin C are bio­markers that are filtered out by the kidneys and used routinely to estimate glomerular filtration rate (GFR) in clinical settings. Creatinine represents muscle tissue, whereas Cystatin C is a protein derived from all nucleated cells at a constant rate, is distributed in extracellular fluid, and representative of all body tissue. Additionally, Cystatin C is completely reabsorbed and catabolized within the proximal tubule, irrespective of age, sex, inflam­mation, malignancy, infection, diet or muscle mass. Concen­trations of Cystatin C increase in hyper­thyroidism and with steroid use and are lower in elderly persons and women. With normal kidney function, a low SI index indicates that the patient has less skeletal muscle available for serum crea­ti­nine production (relatively low numerator) and thus, relative to Cystatin C, this would indicate low muscle mass and potentially sarco­penia (i.e., age‑related loss of muscle mass combined with low muscle strength) (Barretto et al., 2019).

The SI has been validated against skeletal muscle mass measured by abdominal computed tomography (CT) as the refer­ence method. In a retro­spec­tive cross‑sectional study, the SI in 81 critically ill adults independently predicted the CT muscle surface area after adjusting for age, sex, severity of illness, and BMI (Barretto et al., 2019). Figure 16b.1 depicts the correlation between SI and skeletal muscle surface area in 81 critically ill adults.

For every 10 unit decrease in Sarcopenia Index (SI), the odds of a low skeletal muscle index increased 1.4‑fold. This associ­ation was superior to that of serum crea­ti­nine alone after adjust­ment for the same covariates. SI has also been shown to be an independent marker of CT‑measured femoral muscle in more than 1,000 Japanese com­mu­nity residents (Kusunoki et al., 2022) and to predict muscle mass, strength, and physical per­for­mance in Chinese elderly with low bone mineral density (Ge et al., 2022). In the Chinese study, both muscle mass and bone density were measured by DXA, muscle strength by hand grip strength (see Section 16b.5), and physical per­for­mance by the five-chair stand test. A recent system­atic review and meta-analysis has confirmed that serum crea­ti­nine and Cystatin‑C-based index show moderate diagnostic accuracy for sarco­penia (Lin et al., 2024). In this review, low muscle mass accompanied by low muscle strength and low physical per­for­mance were the criteria used as the refer­ence standard for sarco­penia. However, the optimal cutoff point for sarco­penia detection based on the SI could not be established, so more research is warranted.

Studies have also reported associ­ations between a low SI and a high risk of adverse clinical outcomes. For example, in the study of Barreto et al. (2019), a lower SI (indicative of lower muscle mass) independently predicted frailty and worse short-term clinical outcomes such as fewer ventilator days, longer discharge length from the ICU and hospital, and a greater 90d mortality. Lower values for SI have also been observed in patients with type 2  diabetes (Osaka et al., 2018) and gastric cancer (Sun et al., 2022). In these patients, skeletal muscle mass was measured using BIA and CT, respectively. Others have shown associ­ations between a lower SI and higher risk of respiratory failure and severe pneumonia in older adults with chronic conditions such as chronic obstruc­tive pulmonary disease (COPD) (Zhao et al., 2023). Based on these findings, the SI has been proposed as an inexpensive and readily accessible component of a malnu­trition screening tool in critically ill adults (Kashani et al., 2017).

Interpretive criteria and measure­ment of serum crea­ti­nine and Cystatin C

Reference intervals for serum crea­ti­nine concen­tra­tions vary with age, gender, muscle mass, and overall health. Examples of refer­ence intervals for adults are 0.6‑1.2mg/dL (53‑106µmol/L) for healthy adult men and 0.5‑1.1mg/dL (44‑99µmol/L) for healthy women. These refer­ence intervals will vary slightly depending on the crea­ti­nine assay method and level of muscle mass. For children, for example, the refer­ence interval is typically lower due to their reduced muscle mass compared to adults, whereas for elderly adults, lower values are due to age-related reduc­tions in both muscle mass and kidney function.

Two colorimetric methods are used to measure serum crea­ti­nine: enzymatic methods and the Jaffe method. Of these, the enzymatic method is more sensitive and thus more suitable for detecting lower serum crea­ti­nine concen­tra­tions anticipated in pediatric populations and patients with lower muscle mass. Point-of-care handheld devices have been developed to assay crea­ti­nine in whole blood by enzymatic methods, providing results in 2‑15mins. Such methods could be useful in the field in low-income countries; see Kashani et al. (2020) for more details of these assays. A standard refer­ence material is widely available (SEM‑967). The refer­ence assay is isotope-dilution mass spectrometry.

Several immuno­assays are available for serum Cystatin C, although their precision varies. The refer­ence assays include a particle-enhanced nephelometric immuno­assay, a particle-enhanced turbimetric immuno­assay, and enzyme-amplified single radial immuno­diffusion (Levy & Inker, 2017). A refer­ence material for standardization of Cystatin C has been developed, although international standardization remains in progress (Levy & Inker, 2017).

16b.1.6 Serum Insulin-like growth factor 1

Insulin-like growth factor (IGF‑1) (formerly called Somatomedin C) is a peptide hormone produced mainly in the liver in response to pituitary growth hormone. Insulin-like growth factor circulates in the serum largely bound to binding proteins (mainly IGF‑BP3) with minimum diurnal variation and a short half-life (approx­imately 24h). Serum IGF‑1 concen­tra­tions are dependent on age, with the lowest levels at birth, peaking during puberty, and gradually declining in adulthood. There are also physio­logical changes in IGF‑1 levels during normal pregnancy, with levels decreasing during the first and second trimesters, followed by an increase by the 36th week of pregnancy (Huang et al., 2024).

IGF‑1 has an important role in up-regulating protein synthesis and main­taining muscle mass, with low serum levels possibly reflecting nutri­tional insufficiency and tissue catabolism (Livingstone, 2013). During periods of fasting, for example, serum IGF‑1 levels decrease (by nearly fourfold), increasing with nutri­tional repletion (Keller, 2019). Some investigators suggest that serum IGF‑1 is more sensitive and specific bio­marker of protein-energy nutri­tional status than pre­albu­min, trans­ferrin and retinol binding protein (Livingstone, 2013).

Never­theless, many factors other than nutri­tional status affect serum IGF‑1 and reduce its specificity. Of these summarized in Table 16b.5, the most notable is the acute phase response. Conse­quently, in patients who have an active systemic inflam­matory response, serum IGF‑1 has limited use. Table 16b.5 summarizes the other factors influencing serum IGF‑1 concen­tra­tions; see Livingstone (2013) for more details.

Table 16b.5 Pathophysiological influences on total serum IGF-1 concen­tra­tions. From Livingstone (2013). This refer­ence also provides literature citations detailing the observed changes for each factor.

Factor	Effect on serum IGF-1
Genes	Unchanged or Increase or Decrease
Age	Increases until 20y then decreases until 50y.
Pregnancy	Increases as pregnancy progresses
Malnutrition	Decreases
Obesity	Unchanged, increases or decreases
Zinc deficiency	Decreases
Chronic wasting conditions	Decreases
Catabolic conditions	Decreases
Severe liver disease	Decreases
Chronic kidney disease	Decreases or Unchanged
Hypothyroidism	Decreases
GH deficiency	Decreases
Acromegaly	Increases

In anorexia nervosa patients, low serum IGF‑1 levels are asso­ciated with nutritionally acquired growth hormone resistance. With nutritional rehabilitation, IGF‑1 levels seem to correlate with improved BMI z‑score. As such, IGF‑1 has been proposed as a measure of leaness and could be used to monitor both short-term weight changes and nutritional rehabilitation in anorexia nervosa patients (Swenne et al., 2007). In the future, free IGF‑1 may be used as a nutri­tional bio­marker (Livingstone, 2013).

Interpretive criteria and measure­ment of serum IGF‑1 concen­tra­tions

Normal values for serum IGF‑1 vary considerably between indi­vid­uals, ranging from 27.0 to 45.4% making it difficult to establish lower limits across populations. Moreover, IGF‑1 concen­tra­tions vary with the assay used so assay-specific refer­ence intervals are required. Currently, there is an urgent need to develop refer­ence intervals for serum IGF‑1; for more discussion of the issues involved see Huang et al. (2024). Conse­quently, currently, relative changes in serum IGF‑1 may be more useful than absolute values when monitoring nutritional rehabilitation of chronically malnourished patients, provided the same assay is used (Livingstone, 2013).

Currently assay of serum IGF‑1 is expensive and limited to a few specialist laboratories. The predominant methods for assaying IGF‑1 concen­tra­tions in clinical labora­tories are immuno­assays, based on either chemi­luminescence or electro­chemiluminescence. Prior to the assay, interfering binding proteins must be removed from the sample. The most widely adopted method for removing binding proteins involves acid / ethanol precip­itation, now modified to include the addition of IGF‑2 to dissociate IGF‑1 from the IGF binding proteins. This additional step ensures complete elimination of binding protein interference and is now incor­porated into automated systems in commercial laboratories.

Modern mass spectrometric methods such as liquid chro­matog­raphy-tandem mass spectrometry (LC‑MS/MS) and liquid chro­matog­raphy ‑ high-resolution accurate mass spectrometry (LC‑HRAMS) have also been developed which detect IGF‑1 based on its molecular weight and minimize interference caused by IGF‑1 variants. Both methods deliver robust, reliable and accurate results; see Huang et al. (2024) for more details. A WHO 02/254 refer­ence standard is now available, and its use encouraged to minimize inter-assay differ­ences between the assay methods used.

16b.1.7 Plasma amino acids

Plasma amino acids concen­tra­tions reflect a complex pathway influenced by many factors including protein intake, amino acid oxidation, tissue break­down, and protein synthesis. As a result, there is signif­icant variability in plasma amino acid concen­tra­tions between individuals (Schanler & Garza, 1987; Wu et al., 1986), making them an insensitive marker of protein status (Rigo & Senterre, 1987).

In early studies, Ozalp et al. (1972) investigated changes in plasma amino acid concen­tra­tions in healthy adults fed a series of diets deficient in essential amino acids. They noted a 30% decrease in fasting plasma concen­tra­tions for some amino acids (phenylalanine, lysine, isoleucine, and leucine) over the 12d study, whereas other amino acid levels (threonine and valine) decreased by about half within four days of the study initiation (Ozalp et al., 1972). Initially, the group hypo­thesized that the plasma levels of the deficient amino acids would reflect the concen­tration found in skeletal muscle. Even though their initial hypo­thesis may have been a contributing factor, their results suggested other processes may also be involved. For example, the observed differ­ences in plasma amino acid concen­tra­tions between the fasting and fed state may reflect the role of carbo­hydrates on amino acid trans­port into the liver for protein synthesis (Ozalp et al., 1972) and the role of insulin in reducing proteolysis in adults (Souba & Pacitti, 1992).

Plasma amino acid concen­tra­tions in pre-term and term infants also demonstrate large variations in response to feeding inter­ven­tions involving different protein sources. For example, in infants fed formula with protein concen­trations at 30g per litre, Rassin et al. (1977) noted that plasma and urine concen­trations of phenylalanine and tyrosine were elevated. Plasma levels of these amino acids were even higher when the infants were fed a predominately casein-based formula (Rassin et al., 1977). In another study of very low birthweight infants, plasma amino acid concen­tra­tions on contin­uous enteral infusions of either breast milk or a whey-based formula (Schanler & Garza, 1987) varied and were thought to be related to amino acid profiles and bioavailability of the various feeding regimens (Schanler & Garza, 1987).

Similarly, using a neonatal piglet model to represent the physiology of a human neonate, Bertolo et al. (1998) and Wykes et al. (1993) followed plasma amino acid concen­tra­tions throughout a 12h fast. They observed plasma amino acid concen­tra­tions decreased during the initial fast, but as the fast became prolonged, amino acid concen­tra­tions continued to fluctuate. Authors suggest this occurred as a result of decreased protein synthesis and increased catabolism (Bertolo et al., 1998).

16b.1.8 Blood urea nitrogen/urea

Both urea and blood urea nitrogen (BUN) are endpoints of the meta­bolism of dietary protein and tissue turnover. Measured urea levels reflect the whole urea molecule whereas BUN only reports the nitrogen content of urea. As nitrogen comprises about 50% of the urea molecule, BUN levels are roughly half that of measured urea. Nitrogen utilization occurs very rapidly following enteral or parenteral nitrogen intake whereas the overall urea pool is slow to change (7‑10 turnover time). This means urea/BUN can be used as bio­markers of protein intake in clinical practice although they must be interpreted cautiously with short term inter­ven­tions . Additionally, serum urea/BUN levels may be impacted by fluid status. For example, levels may be artificially elevated in a dehydrated state or artificially lowered with high intravenous fluid administration and/or hyper­volemia (i.e., fluid overload).

Urea is commonly used as a bio­marker in neonates admitted to neonatal intensive care units (NICU). A positive correlation between protein intake and BUN has been observed in several studies of preterm infants (Mathes et al., 2018; Sanchez-Holgado et al., 2024). In clinically stable enterally fed preterm infants, a BUN value lower than 1.6mmol/L (4.48 mg/dL) indicates insufficient protein intake and is used to guide the use of human milk fortifiers. Periodic determinations of BUN are recommended by the European Milk Bank Associ­ation to guide indi­vid­ualized fortification of human breast milk in very low birth weight infants (Arslanoglu et al., 2019). One fortification strategy endorsed is called Adjustable Fortification (AF) and recommends adjusting protein intake based on measured BUN. For example, if BUN levels are <10 mg/dL then protein intake should be increased, whereas when BUN levels are >10mg/dL and inad­equate growth is observed, then non-protein calories should be increased with no change to protein intake. In contrast, with BUN >16mg/dL and adequate growth, a decrease in protein intake should be consid­ered (Arslanoglu et al., 2006; Arslanoglu et al., 2019). Sanchez-Holgado et al. (2024) recently examined the effect of this fortification strategy compared to targeted breast milk fortification. The latter is based on adjusting fortification/intake based on analysis of breast milk compo­sition instead of BUN measure­ments. Protein intake was overall higher with Adjustable Fortification but growth param­eters did not differ between the two strategies.

16b.1.9 Immunological biomarkers

Consistent changes in immuno­logical responses have been observed during protein-energy malnu­trition which have been reversed with nutri­tional repletion (Lesourd & Mazari, 1997; Roebothan & Chandra, 1994). As a result, measure­ments of immuno­competence may be useful markers of functional protein-energy status. However, many non-nutri­tional factors (infec­tions, illnesses, major burns, medications, surgery, emotional and physical stress) influence the immune response in the absence of protein-energy malnu­trition . There­fore, to correctly interpret the results of immuno­logical measure­ments, the results should always be interpreted in conjunction with other bio­markers of protein status.

Although no single test can measure the adequacy of the immune response, the most frequently used tests are discussed below:

Lymphocyte count. Lymphocytes are small cells that circulate between blood and lymphoid tissues and are derived from hematopoietic stem cells in the bone marrow. Lymphocytes are the primary cells of the acquired immune system and comprise 20‑40% of the total white blood cells (WBCs). A decrease in total lym­pho­cytes, often to less than 1500/mm³ (refer­ence range 2000‑3500 cells/mm³) can be supporting evidence of severe protein-energy malnu­trition in the absence of any other hematological abnormalities (Shenkin et al., 1996; Keller, 2019). Even in mildly under­nourished elderly, the total lymphocyte count is often reduced compared with data for both healthy elderly and young adults (Lesourd, 1995). Never­theless, lymphocyte count is a very insensitive bio­marker, responding slowly to correction of nutri­tional status (Shenkin et al., 1996).

In clinical labora­tories the lymphocyte count (and total white cell count) are deter­mined automatically using an electronic particle counter (e.g., Coulter Counter). These counters discriminate cells based on their induced resistance as they pass through an electrical potential.

Thymus-dependent lym­pho­cytes. Approximately 75‑80% of the circulating lym­pho­cytes are T‑cells which affect destruction of virus-infected cells, bacteria, and some malignant cells. During severe protein-energy malnu­trition, both the proportion and absolute number of T‑cells in the peripheral blood may be reduced (Keusch, 1990). This low proliferation of T‑cells has been attributed in part to alterations in monokine meta­bolism, particularly decreased activity of interleukin‑1 (Bhaskaram & Sivakumar, 1986; Hoffman-Goetz & Kluger, 1979a; 1979b). In most cases, these changes can be rapidly reversed with nutri­tional therapy. Hence, sequential measure­ments of T‑cell numbers can provide an index of the effect of a nutrition inter­ven­tion and monitor nutri­tional recovery in malnourished patients.

Measure­ment of T‑cells is performed by flow cytometry. Isolated immune cells are stained in suspension with fluorescently tagged antibodies to identify cells of interest prior to analysis through a flow cytometer; for more details, see Manhas & Blattman (2023).

Delayed cutaneous hyper­sensitivity (DCH). This is used as a direct measure of T‑cell-mediated immunity in response to an antigen that is observed on the skin. When healthy persons are re-exposed to recall antigens intradermally, the T‑cells respond by proliferation and then by the release of soluble mediators of inflam­mation. This produces an induration (hardening) and erythema (redness). These skin reactions are often reduced in malnourished persons, although the response is non-specific and occurs with marasmus or kwash­iorkor and with micronutrient deficiencies such as vitamin A, zinc, iron, and pyridoxine. The reactions are reversed, however, after appropriate nutri­tional rehabilitation. Several other diseases and drugs also influence DCH, so it is a poor measure of protein malnu­trition, espe­cially in sick patients. Box 16b.3 summarizes the non-nutritional factors that affect the DCH response, and thus reduce the specificity of the test.

For this test, generally a panel of 4‑5 antigens are injected intradermally, often into the forearm. The area of induration is measured after 48h, with a positive result being taken as an area of induration of at least 0.5cm. Lack of response to all antigens is called anergy. A cell mediated immunity test kit for measuring DCH is available which is preloaded with stan­dard­ized doses of seven antigens (Tuberculin, Tetanus toxoid, Diphtheria toxoid, Streptococus, Candida, Tricophyton and Proteus) in glycerol solution and a glycerin control. Use of seven antigens reduces the number of false negatives and increases the sensitivity of the test. Use of this kit eliminates some of the methodological problems asso­ciated with the traditional skin testing method.

16b.2    Assessment of muscle mass based on technology

Muscle is a major component of the fat-free mass and the primary site for glucose uptake and storage as well as a reservoir of amino acids stored as protein, as noted earlier. assess­ment of muscle mass can there­fore provide an index of the protein reserves of the body, which can be metabolized during prolonged periods of negative nitrogen balance, leading to loss of muscle. Indeed, loss of muscle mass is now an important diagnostic criterion for the definition of protein malnu­trition by the Academy of Nutrition and Dietetics (AND)/American Society for Parenteral and Enteral Nutrition (ASPEN) (Cederholm et al., 2019).

Muscle wasting characterizes the marasmic form of protein-energy malnu­trition. In children with the marasmic form of protein-energy malnu­trition, depleted muscle mass increases the risk of mortality during infec­tions (Briend et al., 2015). In the elderly, loss of muscle mass is a defining characteristic of sarco­penia (defined as loss of muscle mass as well as strength). Sarcopenia often occurs in the face of chronic and acute illness and may be asso­ciated with several negative outcomes, including falls, fractures, and mobility disorders, cognitive impairments, and mortality (Cruz-Jentoft et al., 2019).

The consensus report by the Global Leadership Initiative on Malnutrition (GLIM) endorsed methods such as dual-energy absorp­tiom­etry (DXA), bioelectrical impedance (BIA), ultra­sound, computed tomography (CT), or magnetic resonance imaging as validated measures of body compo­sition. However, the advan­tages and limitations of each technique vary, and must be consid­ered prior to selecting the appro­priate method. For more details of each of these methods, see Chapter 14: Body compo­sition: Laboratory Methods

Recommendations to support the use of these technological methods for the assess­ment of muscle mass in clinical patients are inconsistent across expert groups. For example, DXA has been endorsed for body compo­sition assess­ment by the GLIM Body compo­sition Working Group, and European and Asian sarco­penia working groups (Prado et al.,2022), even though DXA does not measure muscle mass directly. Instead, DXA measures lean soft tissue mass, which includes muscle mass and other tissues and organs. In contrast, the ASPEN does not support the use of DXA for the assess­ment of lean soft tissue mass in patients with a variety of disease state. Sheean et al. (2020) in their system­atic review, found no reports exploring the validity of DXA for lean mass assess­ment in clinical populations. Furthermore, consistent associ­ations between DXA measures of lean soft tissue and adverse outcomes in older adults have not been reported (Bhasin et al., 2020).

Both ultra­sound and BIA are available as portable measure­ment devices so their use in both com­mu­nity and clinical settings for muscle mass assess­ment is increasing. Currently the diverse technology across the commercially available devices for ultra­sound and BIA makes it difficult to compare results across studies (Lee et al., 2020). Diagnostic ultra­sound, also called sonography, is an imaging technique used to measure various tissue thick­nesses, including muscle, bone, and subcutaneous and visceral adipose tissue. The device is low cost, and capable of making fast and noninvasive regional estimates of body compo­sition with no exposure to ionizing radiation. In the ASPEN system­atic review, only three of the seven ultra­sound studies included compared muscle thick­ness and cross-sectional area against lean mass by DXA so no recommendation could be made to support the use of ultra­sound in the clinical setting for muscle mass assess­ment (Sheean et al., 2020). More research is needed to develop a stan­dard­ized protocol for ultra­sound measure­ments, and to generate population-specific refer­ence data as well as cut-off values to identify low muscle mass, before ultra­sound gains widespread use for assessing muscle mass and diagnosing malnu­trition in clinical (Earthman, 2015) and com­mu­nity (Prado et al., 2022) settings. Challenges also exist in using ultra­sound to measure muscle param­eters in indi­vid­uals with obesity or edema (Stock & Thompson, 2021).

Bioelectrical impedance analysis measures the electrical conductivity of a weak alternating current at one or more frequencies passed between two electrodes attached to the body in various locations. The current flows at various rates depending on the compo­sition of the body. The current is well conducted by tissues with a high water and electrolyte content such as lean tissues but is poorly conducted by adipose tissue and bone which have a low water and electrolyte content. The drop in voltage as the current passes through the body (i.e., impedance) is detected through the current sensing surface electrodes, and the impedance data (i.e., resistance, R; reactance, X; impedance, Z; and phase angle, PA) are recorded by the BIA device. From the data generated, estimates of fat-free mass and total body water can be made. BIA has gained widespread use in both com­mu­nity and clinical settings because it is a low cost, portable, and noninvasive tool with minimal risks. These features render BIA an ideal method for follow-up studies. However, the accuracy of the estimates of fat-free mass is often compromised in clinical populations who may have alterations in body water or abnormalities in body geometry. See Kyle et al. (2004) for more details of the theory of BIA.

Each BIA device works with an inbuilt algorithm specific to the device and for the population for which the algorithm was developed so it is not possible to compare studies unless the same combination of device / equation / population is used (Ward, 2019; Sheean, 2020). Furthermore, the refer­ence population on which the BIA algorithm was based must be appro­priate for the target subject being measured (Lemos & Gallagher, 2017). The BIA algorithms applied are based on or validated against other body compo­sition refer­ence methods (e.g., multiple dilution, DXA, and CT). However, none of these refer­ence methods are totally accurate and error-free.

Many factors may influence the precision and accuracy of BIA techniques. They include factors asso­ciated both with the indi­vid­ual (e.g., degree of adiposity, fluid and electrolyte status, skin temper­ature) and with the environment (ambient temper­ature, proximity to metal surfaces and electronic devices), the assumptions under­lying prediction or modeling equations, instrumentation, and variations in the protocols used for the BIA measure­ments. For more discussion of these factors, see Section 14.8 in Chapter 14.

Interestingly, the ASPEN does not support the use of BIA for the assess­ment of body compo­sition in the clinical setting, based on their system­atic review of 23 BIA studies. Their main objections included the scarcity of data on the validity of BIA in specific clinical populations, difficulties comparing studies using different BIA devices, variability in the body compartments estimated, and the proprietary nature of manufacture-specific BIA regres­sion models to procure body compo­sition data. For a summary and discussion of the studies reviewed which led to this conclusion, see Sheean et al. (2020).

The questionable validity of BIA approaches for the assess­ment of whole body compo­sition estimates in clinical populations, espe­cially those with abnormal body geometry or altered fluid homeostasis, has led to the use of raw BIA measure­ments per se for bedside assess­ment of nutri­tional status and/or clinical outcomes. Raw BIA measure­ments do not depend on equations or models, thus eliminating bias and are available from any single frequency BIA device. From the raw BIA measure­ments generated, phase angle can be estimated, now known to be a surrogate of both the quantity and quality of lean muscle mass (Sheean et al., 2020).

16b2.1 Bioelectrical impedance analysis (BIA) - derived phase angle

The derived phase angle (PA) is a specific non-invasive measure­ment obtained directly from a phase-sensitive BIA device. The phase angle concept is based on changes in resistance and reactance as low-level alternating currents pass through evaluated tissues, providing information on hydration status and body cell mass. As such, phase angle is being used as a marker of abnormal body compo­sition and low muscle strength. The phase angle is also sensitive to changes over time and hence could be useful in the evaluation of different treatments. Phase angle is obtained directly from the raw param­eters, resistance, R and reactance, X and is calculated using the following equation (Mulasi et al., 2015). \[\small \mbox { PA (degrees) = arctan (Reactance/Resistance) × (180 degrees/π )}\] Where arctan =arctangent; Reactance, X represents the health and integrity of cells and reflects how much of the current is stored in cell membranes; Resistance R indicates how easily the electric current flows through body fluids and is deter­mined primarily by the amount of water and electrolytes in the body.

Bellido et al. (2023) provide an extensive review of the clinical applications of phase angle assess­ment. For example, a lower phase angle can indicate compromised cellular health and may be asso­ciated with malnu­trition and sarco­penia in the elderly. A system­atic review has confirmed that phase angle is decreased in sarcopenic subjects, with a higher prevalence of sarco­penia reported in those with low phase angle (Di Vincenzo et al., 2021). In several other studies, the accuracy of phase angle to detect sarco­penia has also been good, with lower phase angle scores asso­ciated with reduced muscle mass and poor muscle function measured by BIA and handgrip strength, respectively (Gonzalez et al., 2018; Akamatsu et al., 2022). These obser­vations have led the EWGSOP 2019 consensus on sarco­penia to suggect that phase angle could be regarded as an index of overall muscle quality (Cruz-Jentoft et al., 2019). Phase angle has also been used to predict protein-energy wasting and all-cause mortality in older patients under­going hemodialysis (Kojima e al., 2024). Typically, low phase angle values are related to more-severe illnesses and worse overall health outcomes.

Factors affecting phase angle

Age-related changes in phase angle are relatively marked. Phase angle values peak during early adulthood and then gradually decline as indi­vid­uals age. The basis of this age-related decline is uncertain. Several factors may be involved, including loss of muscle mass and strength and physio­logical changes in the ratio of extracellular water (ECW) and intracellular water (ICW) with aging.

Sex differ­ences in phase angle exist, with women having lower values due to a lower muscle mass.

Body mass index (BMI) influences phase angle. In adults, values increase proportionally with BMI depending on the BMI range. See Bellido et al. (2023) for more details.

Ethnicity/race affects phase angle values. In the US NHANES (1999‑2004) survey data, Hispanic and African Americans had larger phase angle values than Caucasians. Hence, ethnicity/race should be consid­ered when evaluating phase angle values (Kuchnia et al., 2017).

Alterations in fluid distribution affect phase angle values. In disease-related malnu­trition there is an early shift of fluids from intracellular water (ICW) to extracellular water (ECW) space with increased ECW/ICW and a concomitant decrease in body cell mass. Both factors contribute to the lower phase angle values (Gonzalez et al., 2016 ).

Interpretive criteria

Several countries have generated normative refer­ence values for phase angle measured by BIA based on population samples; details are given below. Notable differ­ences in phase angle normative refer­ence values across countries have been observed, some of which may be related to ethnicity-specific differ­ences in relative leg length, frame size, and body build. Such differ­ences may also be contributed in part by the devices used and/or lack of adequate standardization of the phase angle measure­ment technique. Hence, stan­dard­ized population-specific refer­ence data may be necessary. In general, values for phase angle in a healthy population are reported to range from 5‑8°, with a decrease with increasing age, older adults having values between 4‑6°.

U.S. refer­ence data for phase angle by race and sex were compiled by Kuchnia et al. (2017) for non-pregnant indi­vid­uals aged 8‑49y. Data were generated by Bioelectrical Impedance Spectroscopy (BIS) from participants in the U.S NHANES 1999‑2004 surveys (Table 16b.6.)

Table 16b.6 Mean Phase Angle (PA) at 50 kHz by Race and Sex. Numbers are mean ± SD and were generated from Hydra Model 4200 (Xitron Technologies, San Diego, CA). From Kuchnia et al. (2017).

	Males		Females
Race- Ethnicity	PA (50 kHz)		PA (50 kHz)
White	7.36 ± 0.71		6.30 ± 0.67
Black	7.55 ± 0.78		6.61 ± 0.73
Hispanic	7.58 ± 0.68		6.54 ± 0.70
Other	7.37 ± 0.76		6.30 ± 0.59

Phase angle refer­ence values generated by BIA and standardized for sex, age, and BMI have also been developed based on a large German population of children and adults (Bosy-Westphal et al., 2015) and in a Swiss study of older adults aged >65y (Genton et al., 2016).

Currently, the major challenge of using phase angle for clinical assess­ment is the lack of consensus on the choice of cut-off values to identify malnu­trition (or poor clinical outcomes) (Mulasi et al., 2015). Such cutoff values are population and device-specific so fixed cutoff values for specific diseases should not be used. Instead, population-specific risk cutoffs based on specific refer­ence phase angle values should be used.

Table 16b.7 Cut-Points (<5th percentile) defining low phase angle at 50kHz. From National Health and Nutrition Examination Survey Reference Data. Phase angle values generated by the Hydra Model 4200 bioimpedance spectroscopy device (Xitron Technologies, San Diego, CA) BMI: Body mass index (Mean ± SD). Data from Kuchnia et al. (2017).

Ethnicity	18–19y	20–29y	30–39y	40–49y
Hispanic/black	6.42	6.47	6.68	6.13
Men (n)	441	477	404	465
BMI	24.4 ± 5.1	26.5 ± 5.2	27.8 ± 4.7	28.1 ± 4.8
White/other	6.35	6.30	6.32	6.08
Men (n)	213	423	405	407
BMI	24.6 ± 4.6	25.8 ± 4.7	27.0 ± 4.6	27.9 ± 4.6
Hispanic/black	5.43	5.52	5.60	5.37
Women (n)	412	404	370	485
BMI	26.0 ± 6.6	28.1 ± 6.9	29.7 ± 6.8	30.9 ± 6.4
White/other	5.29	5.47	5.38	5.11
Women (n)	172	350	403	406
BMI	24.9 ± 6.2	26.0 ± 6.5	27.0 ± 6.9	27.4 ± 6.5

Kuchnia et al. (2017) provided phase angle cut-points (defined as <5th percentile) stratified by age decade, ethnicity, and sex based on the US NHANES 1999‑2004 data; these are presented in Table 16b.7. For the phase angle cut-points, fat free mass (FFM) index, generated by dividing DXA‑derived FFM by height squared (FFM[kg]/height [meters]²) was employed as the refer­ence technique for assessing body compo­sition. Phase angle was positively asso­ciated with FFMI. These phase angle cut-points have the potential to serve as markers of lean tissue and/or nutri­tional status in studies of the U.S population.

Phase angle cut-off values for predicting sarco­penia have also been developed by Akamatsu et al. (2022), based on a sample of Japanese University students and com­mu­nity-dwelling elderly. In this study phase angle and body compo­sition were measured by BIA and muscle quality by dividing handgrip strength by upper limbs muscle mass (Table 16b.8).

Table 16b.8 Predictive ability of Phase Angle and cut-off values (degrees) for sarco­penia.
AUC, area under the receiver operating characteristic curve; CI, confidence interval; PhA, phase angle. Data from Akamatsu et al. (2022).

Sex Age	AUC	95% CI	Cut-off (°)	Sensitivity (%)	Specificity (%)
Male Young	0.882	0.796–0.967	5.95	100.0	71.8
Male Elderly	0.838	0.516–1.160	5.04	100.0	67.6
Female Young	0.865	0.804–0.926	5.02	100.0	79.4
Female Elderly	0.850	0.674–1.026	4.20	80.0	87.0

More studies are needed to derive diagnostic cut-off values for phase angle in view of the differ­ences observed between populations, clinical disease states, devices used, and lack of adequate standardization of the phase angle measure­ment techniques (Di Vincenzo et al., 2021).

Bioelectrical impedance vector analysis (BIVA) is being explored to enhance the value of phase angle measure­ments, espe­cially in indi­vid­uals with alterations in body water and for whom the correct calculation of body compo­sition is challenging (Bellido et al., 2023). BIVA is a graphical procedure which uses the plot of resistance and reactance stan­dard­ized for height to create a vector that can be compared with gender‑ and race-specific refer­ence values from healthy population samples. In a recent study of children with severe-acute under­nutrition in Ethiopia, BIVA measure­ments successfully differentiated between those children who were dehydrated and those with edema (Girma et al., 2021). During treatment, edematous children lost fluid whereas non-edematous children gained small amounts of fat-free tissue. Moreover, BIVA param­eters correlated with bio­markers of nutri­tional status (Girma et al., 2021).

16b.2.2 Fat-free mass index

Use of a fat-free mass (FFM) index has been explored in place of fat-free mass alone to overcome the confounding effect of variation in stature on fat-free mass. Stature reportedly explains 45% of the variance in FFM and 2% of the variance in fat mass (Van Itallie et al., 1990). Hence, the fat-free mass index (FFMI) assesses the amount of muscle mass relative to a person's height. It is calculated by dividing fat-free mass (FFM) in kilograms by height in meters squared.

Fat-free mass index is used to evaluate nutri­tional status among patients with chronic illnesses or those under­going treatments that may affect muscle mass, such as cancer or chronic obstruc­tive pulmonary disease (COPD). For example, Zhang et al. (2021) reported low FFMI was asso­ciated with a worse prognosis even among cancer patients with a normal/high BMI, emphasizing the prognostic value of FFMI in these patients. Fat-free mass index is also used to diagnose sarco­penia in the elderly, and to evaluate the efficacy of treatment such as exercise programs or medications aimed at improving muscle mass (Cruz-Jentoft et al., 2019). Based on these findings, the GLIM have recommended the inclusion of FFMI in their diagnostic criteria of malnu­trition for use for all patients and in all clinical settings (Cederholm et al., 2019).

The use of FFMI has also been explored in studies of under­nutrition in children in low-income countries (Fabiansen et al., 2017; Wells, 2019). To date, however, challenges remain when applying FFMI to assess nutri­tional status in these children. Under­nutrition in children often coexists with edema, so depending on the method used to measure body compo­sition, some of the assumptions applied to convert raw data into body compo­sition values may have been violated due to altered hydration (see Chapter 14. Section 14.8 for these assumptions). Increasingly in the future, BIVA may be employed to differentiate between children who are dehydrated and those with edema, as discussed earlier. An additional challenge limiting the use of FFMI in children has been the absence of normative refer­ence values and cut-off points that can be used to identify malnu­trition. Current data indicate that universal refer­ence data for FFMI may not be feasible, and instead population specific refer­ence data may be required. Furthermore, any comparison with the refer­ence values should be made using the same body compo­sition measure­ment technique as that used to assess body compo­sition in the child under study (Wells, 2014).

Interpretive criteria for fat-free mass index

Table 16b.9 FFMI, BFMI, and %BF values for various BMI values in healthy Caucasian adults. FFMI, BFMI, and BF were deter­mined from regres­sion equations.
BMI, body mass index; FFMI, fat-free mass index; BFMI, body fat mass index; BF (%), percentage of body fat.

BMI (kg/m2)	FFMI (kg/m2)	BFMI (kg/m2)	BF (%)
Men (n = 2982)
30.0	21.7	8.3	28.8
27.8	20.9	6.9	25.8
25.0	19.8	5.2	21.7
20.0	17.5	2.5	13.4
18.5	16.7	1.8	10.8
Women (n = 2647)
30.0	18.2	11.8	40.0
27.3	17.5	9.8	36.5
25.0	16.8	8.2	33.2
20.0	15.1	4.9	24.6
18.5	14.6	3.9	21.7

Generation of normative refer­ence values of FFMI have potential in prospective studies to evaluate changes in protein-rich FFM in patients pre‑ and post-inter­ven­tion. Such data have been compiled by several investigators. Schutz et al (2002) assessed FFM by BIA in healthy Caucasian Swiss adults aged 24 to 98y. For adults with BMIs within the normal range (i.e., 18.5 to 25kg/m²), normal ranges for FFMI were 17.5 to 19.7kg/m² for men and 15.1 to 16.6kg/m² for women. In a later study, Kyle et al. (2003) published additional data for the FFMI and fat mass index (FMI) ranges from this Swiss study for selected BMI classifications based on this Swiss study; these are presented in Table 16b.9.

Normative values of FFM and FFMI have also been developed for the U.S. heathy weight population based on BIA data from NHANES III for males and females aged from age 12 to over 90y old. Selected percentiles were derived by age, sex (independent of ethnicity) after eliminating the obese (i.e. BMI >30) and under­weight (i.e., BMI <18.5) subjects (Kudsk et al., 201); these are depicted for female subjects in Figures 16b.2.

In addition, tables of indi­vid­ual data for FFMI, restricted by weight for each decile by gender (but not ethnicity) were also compiled. These data are limited to patients 25 to 69y of age. For details of these tables and their potential clinical applications for indi­vid­ual patients, see Kudsk et al. (2017).

Cutoffs for FFMI and FMI consid­ered low, normal, and high based on the Swiss data were derived by Kyle et al. (2004) and Pichard et al. 2004). These are: for "low" FFMI <17.4 (men) and <15.0 (women); for normal FFMI range from 15.5‑19.7 (men) and 15.1‑16.6 for (women); for "high" FFMI >19.8 (men) and 16.7 (women). (Kyle et al., 2004). Corresponding FMI values were for "low" FMI <2.4 (men) and <4.8 (women); for normal FMI ranging from 2.5‑5.1 (men) and 4.9‑8.2 for (women); for "high" FMI 5.2‑8.1(men) and 8.3‑11.7 (women) and "very high" >8.2 (men) and >11.8 (women). The same cut-off values for low FFMI were adopted by the GLIM as a component of the diagnostic criteria for use for all patients and in all clinical patients (Cederholm et al., 2019).

However, caution is needed when applying these FFMI cutoffs across populations. Ethnic-specific cut-off values for FFMI may be needed. South Asians are known to have relatively less lean body mass and more fat at any given BMI value, and hence are likely to have lower FFMI values than non-Asians (Wells, 2014). In a large multi-centre Chinese study, lower BIA derived cutoffs for FFMI (i.e., 14.5kg/m² for females and 16.3kg/m² for males) predicted survival than those adopted by GLIM (Zhang et al., 2021).

A major challenge limiting the use of FFMI in children has been the absence of both normative refer­ence values and cut-off points that can be used to identify malnu­trition in children in low-income countries. Current data indicate that universal refer­ence data for FFMI may not be feasible, and instead population specific refer­ence data may be required. Furthermore, any comparison with the refer­ence values should be made using the same body compo­sition measure­ment technique as that used to assess body compo­sition in the child under study (Wells, 2014).

In an early study normative refer­ence data for BIA-derived FFMI (and FMI) were compiled by Nakao & Komiya (2003) on healthy Japanese children aged 3‑11y. More recent refer­ence data for FFMI has been derived by Wells et al. (2020) for UK infants to age 5y. In this newer study, FFM was calculated as TBW/hydration by applying published age and sex-specific hydration values instead of the hydration constant for fat-free mass of 0.732 (see Chapter 14, Section 14.3.1 for age-and sex-specific hydration factors with discussion). Figure 16b.3 presents the percentiles for fat-free mass index in boys (a) and girls (b). These refer­ence data can be used to monitor FFM/FM accretion in clinical practice in infants and young children (Wells et al., 2020).

More research is needed to define population specific refer­ence data by sex and ethnicity for FFMI in both adults and children.

16b.3    Assessment of muscle mass using anthropometry

The technology-based methods described above are not always available in resource-poor settings. Conse­quently, the GLIM have acknowledged the use of selected non-invasive anthro­pometric measure­ments as alternative proxy measures of muscle mass in both com­mu­nity and clinical settings, complemented, where possible, with hand grip strength as a proxy measure of muscle strength (see Section 16b.5) (Cederholm et al., 2019). The anthro­pometric measure­ments selected include mid-upper arm circum­ference (MUAC), either alone (Hu et al., 2021), or in combination with triceps skinfold thick­ness (i.e., arm-muscle circum­ference and arm-muscle area), together with calf circum­ference. All of these measure­ments have been shown to correlate with muscle mass assessed by in vivo laboratory-based methods such as BIA (Hu et al., 2021), DXA, or CT (Heymsfield et al., 1982; Carnevale et al., 2018). As a result, they are used to predict changes in protein status in resource-poor settings, provided the changes are not small; they are discussed below.

16b.3.1 Mid-upper arm circumference

Mid-upper arm circum­ference (MUAC) has been used in low-income countries to identify protein-energy malnu­trition and to monitor progress during nutri­tional therapy. In low-income countries the amount of subcutaneous fat in the upper arm is often small, so changes in MUAC tend to parallel changes in muscle mass. As changes in MUAC are easy to detect and require a minimal amount of time and equipment, they are often used for screening for protein-energy malnu­trition in emergencies such as famines and refugee crises. In these situations, measure­ment of weight and height may not be feasible, and ages of the children are often uncertain (de Onis et al., 1997). In addition, use of a weight-based nutri­tional assess­ment (e.g., WLZ) can be misleading in children with severe acute protein-energy malnu­trition (SAM) who frequently have diarrheal disease accompanied by dehydration, which lowers the weight of a child. For example, Modi et al. (2015) showed that MUAC outperformed weight-for-length-score <−3 to identify SAM in children from Bangladesh aged 6‑60mo with diarrhea. In such cases, a fixed cutoff point (e.g., 11.5cm) has been used to distinguish normal and children with protein-energy malnu­trition and diarrhea.

The use of fixed cutoffs to distinguish normal and malnourished children assumes that MUAC is relatively independent of age. However, the age independence of MUAC has been questioned (Hall et al., 1993; Bern & Nathanail, 1995; WHO, 1995) and has led to the use of MUAC Z‑scores, that adjust for age and sex differ­ences (Houssain et al., 2017). However, some investigators have concluded that the additional estimation of age, required when using MUAC-for-age Z‑score was not justified, espe­cially in humanitarian settings. They urged the need for further research on morbidity and mortality of children with low MUAC-for-age Z‑scores (Leidman et al. 2019).

A major application of MUAC in older adults is as a surrogate for appendicular skeletal muscle mass, and the subsequent detection of sarco­penia (Pinheiro et al., 2020; Hu et al., 2021) as well as a reliable predictor of protein-energy malnu­trition in hospitalized frail older adults (Sanson et al., 2018). MUAC is recommended because its measure­ment is less affected by fluid retention compared to the lower extremities, a condition that often occurs in older adults. Sarcopenia is characterized by decreases in muscle mass, as well as strength and function, all of which have multiple adverse health consequences.

Interpretive criteria and measure­ment of mid-upper arm circumference

WHO (2009) recommends a MUAC cutoff of <115mm to diagnose children aged 6‑60mo with severe acute malnu­trition (SAM), together with the presence of bilateral pitting edema, where possible. This MUAC cutoff was chosen because children with a MUAC less than 115mm were observed to have a highly elevated risk of death compared to those with a MUAC above 115mm (Myatt et al., 2006). Following treatment, the WHO discharge criteria recommended for children with SAM are a MUAC cutoff of >125mm and no edema for at least 2wks (WHO, 2009). To avoid any relapse, children with SAM who have been discharged from treatment should be monitored periodically. Provision to mothers of MUAC tapes with a cut-off value of 115mm have been used in an effort to detect malnu­trition early before the onset of complications, and thus reduce the need for inpatient treatment (Blackwell et al., 2015).

The U.S Academy of Nutrition and Dietetics (AND) and the American Society for Parenteral and Enteral Nutrition (ASPEN) also recommends MUAC cutoffs to classify bedbound children aged 6‑60mos with under­nutrition when measure­ments of weight and length or height are not feasible. For children classified with severe protein-energy malnu­trition, a MUAC cutoff <115mm is also recommended, for moderately malnourished children, a MUAC cutoff of 115‑124mm, and for children at risk of protein-energy malnu­trition, a MUAC cutoff of 125‑134mm (Becker et al., 2014).

Currently, there is no consensus on the MUAC cutoff to predict low muscle mass and diagnose under­nutrition in adults. WHO has proposed universal cutoff values for MUAC of 22cm for females and 23cm for males to screen for under­nutrition (WHO, 1995). However, reginal and ethnic differ­ences in MUAC exist, and a higher MUAC cutoff (23.5cm) has been used to screen for under­weight in women of reproductive age in northern Vietnam. If the lower WHO cutoff (22cm) had been used, many women at risk for under­nutrition would have been misclassified as healthy.

Cutoffs for MUAC to predict low muscle mass and identify sarco­penia in adults have also been investigated. In a study of com­mu­nity-dwelling Chinese adults >50y, Hu et al. (2021) recommended MUAC cutoffs of <28.6cm for men and <27.5cm for women for predicting low muscle mass, and <27cm for both sexes to identify sarco­penia. This cutoff was developed based on low muscle mass diagnosed using the European Working Group on Sarcopenia in Older People 2 (EWGSOP2) criteria. Whether these MUAC cutoffs for adults are applicable for other race-ethnic groups warrants investigation.

WHO has developed MUAC-for-age refer­ence data (Z‑scores and percentiles by sex) for children aged 3‑60mo for international use. The curves show both age-specific and sex-specific differ­ences for boys and girls aged <24mos. Numerical Z‑score tables and charts for boys are also available.

Details on the measure­ment of MUAC can be found in Section 11.2.1 of Chapter 11.

16b.3.2 Mid-upper arm muscle circumference

Mid-upper arm muscle circum­ference (MUAMC) has been used in the past as a proxy for total body muscle mass and to diagnose protein-energy malnu­trition in com­mu­nity surveys in resource-poor setting (Jelliffe, 1966). Strong correlations between calculated values for MUAMC and fat-free mass estimates based on refer­ence in vivo methods, such DXA Carnevale et al., 2018) and computed tomography (Lambell et al., 2021), have also led to the use of MUAMC as a proxy for muscle mass in the elderly (Akin et al., 2015; Landi et al., 2010) when using these more direct in vivo methods are not feasible (Gort-van Dijk et al., 2021).

Never­theless, it is important to realize that the MUAMC is a one-dimensional measure­ment, whereas MUAMA is two-dimensional, and mid-upper-arm-muscle volume is three dimensional. Conse­quently, if the volume of the mid-upper-arm muscle declines during protein-energy malnu­trition or enlarges following a program of nutri­tional support, the MUAMC change will be proportionally smaller than the change in the mid-upper-arm muscle area (Heymsfield et al., 1982). Hence, MUAMC is insensitive to small changes of muscle mass that might occur, for example, during a brief illness.

Interpretive criteria and measure­ment of mid-upper-arm-muscle circumference

Mid-arm-muscle circum­ference is the calculated circum­ference of the inner circle of muscle surrounding a small central core of bone (Gurney & Jelliffe, 1973). The muscle circum­ference of the mid-upper arm is derived from measure­ments of both the MUAC and triceps skinfold thick­ness. For details on the measure­ment of triceps skinfold thick­ness, see Section 11.1.2 of Chapter 11. The equation for the calculation of MUAMC is based on the same assumptions as those described for mid-upper-arm fat area (see Chapter 11, Section 11.1.4).

As variations in skinfold compressibility are ignored and as the triceps skinfold of females is generally more compressible than that of males, female MUAMC may be under­estimated. As a further complication, the MUAMC equation does not take into account between subject variation in the diameter of the humerus relative to MUAC (Frisancho, 1981). If MUAC = mid-upper-arm circumference, TSK = triceps skinfold,
    d1 = arm diameter, and d2 = muscle diameter. Then:
Where MUAMC = mid-upper-arm-muscle circumference, MUAC = mid-upper-arm circumference, and TSK = triceps skinfold thick­ness. Note that this equation is only valid when all measure­ments are in the same units (preferably mm).

There are no refer­ence ranges for MUAMC for children based on the WHO Multicentre Child Growth Study or for the U.S. children used to compile the CDC 2000 BMI growth charts (Kuczmarski et al., 2000a). Some population-specific refer­ence data based on calculated MUAMC are avail­able for Argentinian children aged 4‑14y (Oyhenart et al., 2019a).

Kuczmarski et al. (2000b) compiled MUAMC refer­ence data for adults from the U.S. NHANES III survey (1988‑1994), but only for those adults >50y. Mean (SE) and selected percentile values of males and females for four age groups are available. During this time, values for MUAMC increased up to age 65y in women and up to middle age in men and then steadily decreased. However, secular-related changes in MUAC and triceps skinfold thick­ness have been reported in U.S. males and females, so caution must be used when comparing more recent MUAMC data with these earlier MUAMC refer­ence data for U.S. adults (Frisancho, 1990).

16b.3.3 Mid-upper-arm-muscle area

Mid-upper-arm-muscle area (AMA) is said to be preferable to mid-upper-arm-muscle circumfer-ence as an index of total body muscle mass because it more adequately reflects the true magnitude of muscle tissue changes (Frisancho, 1981). Several studies have examined the validity of mid-upper-arm-muscle area by comparison with in vivo body compo­sition refer­ence methods such as magnetic resonance imaging (MRI), computer tomography (CT), bioelectrical impedance (BIA), and DXA. Unfortunately, the validity of the calculated mid-upper-arm-muscle area (AMA) as a proxy for actual arm-muscle mass is dependent on the characteristics of the study population and the in vivo refer­ence method used. For example, the traditional equation appears to overestimate AMA in obese patients and may not be appro­priate for under­nourished children (Heymsfield et al., 1982; Rolland-Cachera et al., 1997).

Despite these limitations, calculated AMA as a proxy for arm-muscle mass has been used by several investigators. Pinto et al. (2021), reported that AMA, like MUAMC, was linked with length of hospital stay. In Caucasian adult patients with AMA values lower than the 5th percentile (i.e., indicative of depletion), the probability of being discharged from the hospital was lower. However, this finding has not been consistent in all studies (Luis et al., 2013). AMA has also served as an early indicator of deteriorating nutri­tional status in a longitudinal study of pediatric patients with cystic fibrosis (Ellemunter et al., 2021).

Interpretive criteria and measure­ment of mi-upper-arm-muscle area

The follow­ing equation may be used to estimate mid-upper-arm-muscle area (AMA): \[ \small \mbox{Arm muscle area} = \mbox{(MUAC − (π × TSK))}^{2}\mbox{/4π}\] where MUAC = mid-upper-arm circumference and TSK = triceps skinfold thick­ness (Frisancho, 1981). Consistent units, preferably mm, should be used throughout.

Age and sex-specific smoothed percentiles for AMA based on the same population of healthy U.S. children aged 1‑20y used to construct the CDC 2000 BMI charts (Kuczmarski et al., 2000a) and the 2010 skinfold thick­ness percentiles of Addo and Himes (2010) are shown in Figure 16b.4 (Addo et al., 2017). The authors also provide the necessary least mean squares coefficients to calculate Z‑scores; see Addo et al. (2017) for more details.

Note the percentile curves in Figure 16b.4 represent the entire race-ethnicity groups that were represented in the survey, so that interpretation based on these percentile curves should take into account any potential effects of specific race-ethnicity.

Some local region-specific percentiles curves for AMA are available, including those for boys and girls aged 4‑14y in Argentina (Oyhenart et al., 2019a). The mean AMA values for 3rd, 50th, and 97th percentiles are lower for the Argentinean children than their U.S. counterparts (Oyhenart et al., 2019a).

Prediction equations adjusted for the age-dependent effects of height on AMA have also been developed by Addo et al. (2017). These prediction equations may be espe­cially helpful in populations with a high prevalence of stunted children or those in which even well-nourished children are shorter than the average height.

Reference values for AMA for U.S. adults (18‑74y) or an older subset aged >50y have not been compiled from the U.S. NHANES III data. Earlier data on AMA percentiles (5th through 95th) are available based on the merged NHANES I and II data (Frisancho, 1990). These data are presented by age, sex, and race-ethnicity for persons 1‑74y; by height for boys and girls 2‑17y; and by age, sex, and frame size for adults aged 18‑74y. However, secular changes in AMA in U.S. males and females have been reported, so caution must be used when inter­preting data based on these older refer­ence data.

16b.3.4 Calf circumference

Calf circum­ference is often used as a surrogate marker of skeletal muscle mass when body compo­sition techniques are not available. The measure­ment is simple and practical and has been reported as the most used measure­ment in clinical practice to assess muscle mass, based on an international survey in 55 countries (Bruyere et al.,2016). Calf circum­ference has the added advan­tage of having a lower fat mass compared to other body sites. As a result, the impact of fat mass on the measure­ments will be less (Bahat, 2021).

The use of calf circum­ference as a proxy for muscle mass to identify older adults with or at risk for sarco­penia is supported by both the Asian Working Group for Sarcopenia (AWGS) (Chen et al., 2020) and the European Working Group on Sarcopenia in Older People (EWGSOP) (Cruz-Jentoft et al., 2019), espe­cially for use in settings where no other muscle mass diagnostic methods are available. Several investigators have compared the per­for­mance of calf circum­ference as an indirect anthro­pometric marker of appendicular skeletal muscle mass against in vivo refer­ence methods. Most of these studies have been on the elderly and have reported moderate to good correlations of calf circum­ference against direct measure­ments of skeletal muscle mass using refer­ence methods such as DXA (Kawakami et al., 2015) and BIA (Gonzalez-Correa et al., 2020). However, whether changes in calf circum­ference can be used as a valid proxy for changes in muscle mass for hospitalized patients remains uncertain (Hansen et al., 2024).

Factors affecting calf circumference

Age has little effect on calf circum­ference in adult males and females until age 60y, after which calf circum­ference decreases slowly with age, irrespective of ethnicity or race. These age-related trends in calf circum­ference values appear to be confounded by BMI.

Sex impacts on calf circum­ference with males having higher values than females.

Ethnicity or race differ­ences exist due to alterations in muscularity. Such ethnic differ­ences are confounded by BMI.

Body mass index influences calf muscle values. Persons with overweight/obesity have higher calf circum­ference values, whereas for those with BMI<18.5, values are smaller irrespective of age, ethnicity or race. Such differ­ences are attributed to the larger amount of adipose and inter­muscular adipose tissue in the calves, limiting the usefulness of calf circum­ference diagnostic per­for­mance as a marker of muscle mass. However, BMI adjust­ment factors are available; see below (Gonzalez et al., 2021).

Edema can affect the measure­ment and its interpretation. Ishida et al. (2019) reported that edema in the lower limb increased calf circum­ference by approx­imately 2cm. Adjustment factors for low extremity edema are available (Ishida et al. 2019).

Interpretive criteria and measure­ment of calf circumference

Normative refer­ence values for calf circum­ference by sex, ethnicity and race for U.S. adults 18‑39y with a normal BMI (18.5‑24.9kg/m²) are presented in Table 16b.10 (Gonzalez et al., 2021).

Table 16b.10 Reference and cutoff values for calf cir­cum­fer­ence according to sex, ethnicity, and race, from participants with normal BMI^3
1 Reference values defined as mean values from participants aged 18‑39y.
² Values defined as −1 SD below the mean values.
³ BMI: 18.5‑24.9 kg/m²; for other BMI groups, use the adjusting factors for correction of calf cir­cum­fer­ence (Table 16b.11). Data from González-Arellanes et al. (2021).

	Males			Females
	n	Reference1 mean ± SD	Cutoff2 −1 SD	n	Reference1 mean ± SD	Cutoff2 −1 SD
White Non- Hispanic	633	36.6 ± 2.2	34.4	656	35.6 ± 2.2	33.4
Black Non- Hispanic	429	36.4 ± 2.2	34.2	279	35.3 ± 2.2	33.1
Mexican American	428	34.9 ± 2.1	32.8	378	33.9 ± 2.3	31.6
Other races & ethnicities	149	36.0 ± 2.1	33.9	152	34.6 ± 2.2	32.4

Cutoff values indicative of a moderately low calf circum­ference by sex and ethnicity and race, and defined as ‑1SD below the corresponding mean value, are also shown in Table 16b.10. These cutoff values are appro­priate to detect low muscle mass in adults >65y (validated by DXA measure­ments) for sarco­penia diagnosis / screening.

Gonzlez-Arellanes et al. (2021) have compiled BMI adjust­ment factors to apply for males with BMIs (kg/m²) outside the normal range (i.e., BMI 18.5‑24.9) by sex and by ethnicity and race. An example of the adjust­ment factors for calf circum­ference for males is shown in Table 16b.11.

Table 16b.11 Adjustment factors for calf cir­cum­fer­ence by BMI for males outside the 18.5‑24.9 BMI range, calculated from linear regres­sion of calf cir­cum­fer­ence, adjusted by age. Data from González-Arellanes et al. (2021), who also present comparable data for females.

BMI Group (kg/m2)	Total All subjects	White Non- Hispanic	Black Non- Hispanic	Mexican- American	Other races & ethnicities
< 18.5	+4.3	+4.7	+4.2	+4.0	+3.4
25‑29.9	−3.4	−3.4	−3.4	−3.1	−3.5
30‑39.9	−6.8	−6.7	−7.2	−6.4	−6.9
≥ 40	−12.0	−11.9	−12.0	−12.1	−12.2

To apply the adjust­ment factors, add 4cm to the calf circum­ference measure for BMI <18.5 or subtract 3, 7, or 12cm from the calf circum­ference measure for BMI 25‑29, BMI 30‑39, and BMI = 40, respectively. Without these adjustments, raw calf circum­ference measure­ments would result in an under­estimate of the prevalence of low calf circum­ference in those with BMI = 25 and an overestimate in those with a BMI <18.5.

Note that the BMI-adjust­ment factor should not be applied to indi­vid­uals with a BMI <18.5kg/m² who have weight or muscle loss, espe­cially in aging and clinical populations because the adjust­ment factors for people with BMI were derived from "healthy" and young indi­vid­uals. For these indi­vid­uals, no adjust­ment value should be applied.

However, ethnic difference in optimal cutoff values for calf circum­ference to predict skeletal muscle mass exist. The Asian Working Group for Sarcopenia (AWGS) 2019 consensus recommend screening cutoff values of <34 for men and <33 for women in Older People (Chen et al.,2020).

The measure­ments are taken using a steel measuring tape while the subject is in a seated position. The tape is placed around the calf and moved up and down to locate the maximum circum­ference as shown in Figure 16b.5. The tape must be held snugly but not tight and the measure­ment taken to the nearest 0.1cm. In the U.S. NHANES, measure­ments are performed on the right calf (CDC 2000).

16b.4    Assessment of muscle mass using urinary metabolites

16b.4.1 Urinary crea­ti­nine

Urinary crea­ti­nine is a break­down product of creatine phosphate, a metabolite distributed primarily in skeletal muscle, as noted earlier; (see serum crea­ti­nine Section 16b.1.5). Under the steady-state and a stable kidney function, crea­ti­nine is usually produced by the body at a relatively constant rate depending on the absolute amount of muscle mass. Several studies have investigated the relation between 24h urinary crea­ti­nine and skeletal muscle mass using refer­ence methods to measure muscle mass. These studies led to the devel­op­ment of prediction equations to estimate total body muscle mass based on 24h urinary crea­ti­nine excretion (Forbes & Brunning, 1976; Wang et al., 1996). Based on regres­sion analysis, each gram (9mmol) of crea­ti­nine in urine is derived from about 17kg skeletal muscle (Forbes & Brunning, 1976) This prediction equation was developed with total body potassium-measured lean body mass as the refer­ence for skeletal muscle (see Chapter 14, Section 14.2). In a more recent study, the total body creatine pool was measured in subjects who ingested a gel capsule filled with the stable isotope deuterated creatine (DCR). Skeletal muscle mass was deter­mined in these same subjects using magnetic resonance imaging (MRI) as the refer­ence and reported to be well correlated with skeletal muscle mass measured from the stable isotope deuterated creatine using mass spectroscopy (Clark & Manini, 2010).

However, several factors affect daily crea­ti­nine excretion; for more details see Box 16b.4.

Traditionally, the crea­ti­nine height index (CHI) as a percentage has been used to evaluate the degree of depletion of muscle mass in children with the marasmic form of protein-energy malnu­trition. In these subjects, there will be a decrease in CHI because of loss of lean body mass to main­tain serum protein levels. Knowledge of the exact age of the child is not required, which can be an advan­tage in low-income-countries. The CHI can also be used to monitor the effects of long-term nutritional inter­ven­tions on repletion of lean body mass in hospital patients. Changes in adipose tissue and fluid retention in these patients have no influence on CHI. However, the CHI is not sensitive to weekly changes in lean body mass and must be used over longer periods.

Interpretive criteria and measure­ment of urinary crea­ti­nine

The creatinine height index (CHI as%) is calculated as \[\small \frac{\mbox{measured 24h urinary crea­ti­nine × 100% }} {\mbox{ideal 24h urinary crea­ti­nine for height}} \]

A CHI of <60% of the standard is said to represent a severe deficit in body muscle mass (Bistrain et al., 2020). Table 16b.4 presents the urine crea­ti­nine values that represent 60% of the expected urine crea­ti­nine excretion by height and sex. Bistrain and co-workers advocate the use of CHI to identify severe protein-energy malnu­trition in hospital patients, followed by more advanced body compo­sition techniques to confirm or refine the initial diagnosis; see Chapter 14 for details of more advanced body compo­sition methods.

For details of the analytical methods for crea­ti­nine, see Section 16b1.5 on serum crea­ti­nine.

16b.4.2 Urinary 3‑hydroxyproline

3‑hydroxy­proline is abundant in collagen protein and is a metabolite of collagen break­down (Alcock et al., 2019; Saito et al., 2010). Collagen is essential to main­tain the normal structure and strength of connective tissue such as bone, skin, cartilage, and blood vessels. The measure­ment of urinary 3‑hydroxy­proline reflects both dietary collagen and collagen from tissue break­down, and as such has been suggested as a method for evaluating tissue break­down (Alcock et al., 2019).

Urinary 3‑hydroxy­proline has also been used in earlier studies in low-income countries on malnourished children with impaired growth in whom excretion levels were signif­icantly lower than thoseF in age-matched, well-nourished children, irrespective of whether the malnourished children had kwash­iorkor, marasmus, or marasmic-kwash­iorkor. However, in clinical settings, urinary 3‑hydroxy­proline is commonly used for cancer screening (and not as a nutri­tional bio­marker). In cancer screening, levels are elevated because cancerous cells destroy the high collagen content of the cellular basement membrane (Saito et al., 2010).

A variety of extraneous factors are known to influence urinary 3‑hydroxy­proline levels and are shown in Box 16b.5.

Interpretive criteria and measure­ment of 3‑hydroxy­proline

Recognition of the influence of age on 3‑hyroxyproline excretion led to the devel­op­ment of the hydroxy­proline index. This is independent of age in children and attempts to take body weight into account and is shown below: \[\small \frac{\mbox{mg hydroxy­proline per ml urine}} {\mbox{mg creatinine per ml urine}} × \mbox{kg body weight}\] In well-nourished children between 1 and 6y, the hydroxy­proline index is relatively constant and is approx­imately 3.0 whereas in malnourished children the index is low, irrespective of the type of malnu­trition and statistically related to the weight deficit (Whitehead, 1965). Interpretive criteria for the hydroxy­proline index are shown in Table 16b.12.

Table 16b.12 Guidelines for the interpretation of urinary hydroxy­proline index, crea­ti­nine height index, and urinary urea nitrogen:crea­ti­nine ratios. From Sauberlich (1999).

	Deficient (high risk)	Low (medium risk)	Acceptable (low risk)
Hydroxy­proline index (3mo to 10y of age)	<1.0	1.0–2.0	>2.0
Creatinine height index (3mo to 17y of age)	<0.5	0.5–0.9	>0.9
Urea nitrogen : crea­ti­nine ratio	< 6.0	6.0–12.0	>12.0

The use of 24h urine samples is preferred but it is not practical for children. Instead, early morning fasting samples can be collected, thereby minimizing the effects of the dietary ingestion of hydroxy­proline. Both colorimetric and high-per­for­mance liquid chro­matog­raphy (HPLC) can be used to analyze urinary hydroxy­proline. The colorimetric method is available as a commercial kit. For the method, the urine samples are first hydrolyzed, then decolorized and neutralized prior to oxidation to pyrrole. Pyrrole is estimated colorimet­rically after coupling with p‑dimethylaminobenzaldehyde (Prockop & Underfriend, 1960). The HPLC method of Bianchi & Mazza (1995) is shorter compared with other methods and involves two derivatizations. The first derivatization is with o‑phthalaldehyde to eliminate interferences from some primary amino acids eluting with retention times similar to those of hydroxy­proline, and the second with dabsyl chloride.

16b.4.3 Urinary 3‑methylhistidine

Urinary 3‑methylhistidine (3‑MH) arises from the catabolism of the two skeletal muscle proteins actin and myosin. 3‑methylhistidine cannot be re-utilized for protein synthesis, but instead is excreted quantitatively into the urine without further meta­bolism. There­fore, in indi­vid­uals in a steady state of body compo­sition, 3‑MH urinary excretion has been proposed as a predictor of fat-free mass and muscle break­down (Ballard & Tomas, 1983), although with relatively poor sensitivity to monitor changes in body protein stores (Keller, 2019). Several empirical prediction equations have been published based on urinary 3‑MH to estimate total skeletal muscle mass (Heymsfield et al., 2915). Never­theless, there are several limitations to the use of 3‑MH (Box 16b.6) for assessing the size and turnover of the skeletal protein mass.

Of the confounding factors listed in Box 16b.6, in clinical settings catabolic conditions are of special concern because muscle protein turnover is sensitive to stressful conditions, resulting in an increase in 3‑MH excretion even in patients with depleted skeletal protein mass (Young et al., 1990). The dependence of 3‑MH on appro­priate renal function is also a limiting factor in clinical settings.

Interpretive criteria and measure­ment of 3‑methylhistidine

There is a paucity of data on the excretion of 3‑MH in healthy populations so that interpretive criteria have not been established.

For the measure­ment, multiple 24h urine collections, preferably on a meat-free diet, are required, which limit the use of urinary 3‑MH as a bio­marker of skeletal protein mass (Keller, 2019). Further, the analytical methods available are cumbersome, some involving ion-exchange chro­matog­raphy, and others high-per­for­mance liquid chro­matog­raphy (Minkler et al., 1987).

16b.5    Assessment of muscle function

Malnutrition-related muscle wasting is asso­ciated with signif­icant changes in muscle function. Such changes in muscle function may precede body compo­sition changes, and as a result are used to detect functional impairment, and a key component for diagnosis of sarco­penia (Cederholm et al., 2019). Studies using magnetic resonance imaging have shown that muscle strength decreases by more than 50% in in the elderly (70‑82y) compared to that in young men (18‑30y). Moreover, only half the decrease in muscle strength that occurs with aging is accounted for by a decrease in the volume of muscle (Morse et al., 2005).

Handgrip strength is frequently used to evaluate skeletal muscle function in both clinical and com­mu­nity-based studies and is recommended by the Global Leadership Initiative on Malnutrition (GLIM) consensus for use in situations where muscle mass cannot be readily assessed (Cederholm et al., 2019).

Handgrip strength (HGS) is a simple, inexpensive and non-invasive tool that can be used as a surrogate of skeletal muscle strength (Bohannon, 2019; Sandhu & Lee, 2023). For example, HGS is the recommended technique for assessing muscle strength by the European Working Party on Sarcopenia in Older People (EWGSOP) (Cruz-Jentoft et al., 2019) and the Asian Working Group for Sarcopenia (AWGS) (Chen et al., 2020). It has been introduced by AWGS to identify "possible sarco­penia" with or without reduced physical per­for­mance in both com­mu­nity health care and prevention settings (Chen et al., 2020). A decline in muscle strength generally exceeds changes in muscle size.

For the measure­ment, subjects squeeze a handgrip dynamometer for a few seconds, the resulting value reflecting static force (measured in kilograms) which provides an objective measure­ment of strength (Vaishya et al., 2024). Dynamometers are relatively inexpensive and can be used in both com­mu­nity and hospital-based settings. When used while seated, measured handgrip strength reflects primarily arm strength, whereas if the subject is standing, lower body and core strength may also be reflected.

Handgrip strength, when compared to subjective global assess­ment (SGA), has been shown to be an independent predictor of malnu­trition (Flood et al., 2014; Kaegi-Braun et al., 2021; Norman et al., 2011) and to improve with nutri­tional inter­ven­tion (Norman et al., 2011; Sandhu & Lee, 2023). In a recent randomized control trial of adults admitted to hospital, handgrip strength was negatively asso­ciated with 30d all-cause mortality. Additionally, in this trial, the men not the women with the lowest handgrip strength at admission, experienced the greatest benefit of nutrition support (Kaegi-Braun et al., 2021).

Bohaanon (2017) however, has cautioned that relatively large percentage changes in grip strength may be necessary to conclude with confidence that a real change has occurred in muscle strength over time in response to treatment or an inter­ven­tion. The EWSOP2 group have emphasized that more studies are needed to establish whether the current recommended lower limits (i.e., cutoffs) improve the prediction of outcomes most sensitive to response to treatment for sarco­penia (Cruz-Jentoft et al., 2019).

Several functional health outcomes have been used to define optimum HGS in the elderly. They include slow measured walking speed, self-reported difficulty in walking (Lauretani et al., 2003;. Sallinen et al., 2011), limitations in Instrumental Activities of Daily Living, and altered physical per­for­mance. The latter may be evaluated through the Timed Up and Go test, 5‑time chair stand test (>12s), 6‑meter walk (<1.0m/s) or short physical per­for­mance battery (<9) (Lera et al., 2018; Chen et al., 2020). Most of these relationships have been based on cross-sectional data. However, there have also been some prospective cohort studies in which lower HGS at discharge from acute care hospitals has been asso­ciated with 30d readmission (Allard et al., 2016). In others, associ­ations of HGS with all-cause mortality and cardiovascular mortality have been reported (Lera et al., 2018; Leong et al., 2015; Wu et al., 2017; Rantanen et al., 2003; Steiber, 2016).

Factors affecting hand grip strength

Age influences HGS with increases over the life course. Values peak in early adult life at a strength that is main­tained through to midlife, after which it declines, with the loss accelerating through old age (Roberts et al., 2011).

Sex affects HGS, with males typically having higher HGS than females. Age and sex are the strongest predictor of HGS in healthy aging indi­vid­uals (Norman et al., 2011) (see Figure 16b.6).

Ethnicity also impacts HGS, with South Asians having lower and Africans greater HGS compared with other ethnic groups (Vaishya et al., 2024). These findings have prompted studies to develop cutoffs for weak muscle strength in several geographic regions (Lera et al., 2018; Steiber, 2016; Auyeung et al., 2020).

Other factors such as socio-economic status and physical fitness may also alter hand grip strength. For more details see Vaishya et al. (2024).

Interpretive criteria and measure­ment of hand-grip strength

Measure­ments generated from dynamometers differ depending on the device used. However, currently dynamometer-specific cutoff values are not recommended because of insufficient comparative data (Chen et al., 2020).

Several sets of region-specific normative refer­ence values for hand grip strength are available, although they have not yet been standardized. In the United Kingdom for example, percentiles (10th, 25th, 50th, 75th, 90th) by sex for grip strength based on cross-sectional obser­vations in persons aged 4‑90y from twelve U.K. hand grip studies have been compiled (Dodds et al., 2014). A subset of these results are presented in Table 16b.13.

Table 16b.13 Normative values for grip strength for UK Males. Data from Dodds et al. (2014).

		Centiles (kg)
Age(y)	n	10th	50th	90th	Mean (SD)
10	3222	12	17	22	17.2 (4.1)
20	354	30	40	52	41.5 (7.3)
30	984	38	51	64	51.6 (9.6)
40	880	38	50	63	50.3 (10.3)
50	820	35	48	60	47.6 (10.1)
60	2683	33	45	56	44.6 (9.2)
70	3286	29	39	49	39.1 (8.1)
80	1115	23	32	42	32.2 (7.3)
90	431	16	25	33	24.7 (6.8)

In the United States, cross-sectional data (mean, SD) from the 2011 US National Institutes of Health Toolbox project are available for height, weight, handgrip strength (kg), together with the 5th, 10th, 50th, 25th and 50th percentiles by side of hand (dominant and non-dominant), sex, and age (3‑17y) (Bohannon et al., 2017). The stan­dard­ized method for measuring HGS recommended by Roberts et al. (2011) using a Jamar dynamometer was used in this U.S. study.

Later, Wang et al. (2018) used the same 2011 dataset to compile refer­ence values for US adults 18‑85y. Again, data for height, weight, hand grip strength (kg) and percentiles (10th, 25th, 50th, 75th and 90th) by sex, side of hand (dominant and non-dominant) and 13 age groups (limited to 5y spans, except the strata 18‑24y) are presented. No cutoffs for HGS based on the data for either children or adults were derived.

In Germany, the normative refer­ence values are based on a nationally representative sample of healthy participants aged 17‑90y. Mean values for HGS across seven height groups (within each age group: 17‑19y, 20‑24y, 25‑29y, 30‑34y, 35‑39y, 40‑44y, 45‑49y) were calculated and presented as sex-specific refer­ence values, stratified by age and body height. For example, the refer­ence value for 40‑44y women with a height of 165‑169cm is 34.8kg; this value increases by about 1kg for every 5cm of additional height as shown in Table 16b.14.

Table 16b.14 Normative refer­ence values for handgrip strength for German women 40‑44y. Abstracted from more comprehensive data for men and women aged 17‑90y. Steiber (2016)
¹ Mean −1 age-group-specific SD

Height (cm)	Mean HGS (kg)	Threshhold1 (kg)
150‑154	31.5	25.3
155‑159	32.7	26.4
160‑164	33.7	27.4
165‑169	34.8	28.6
170‑174	35.8	29.6
175‑179	37.1	30.8
180‑184	38.0	31.8

In some regions, refer­ence values of handgrip strength have been restricted to certain age groups. For example, in com­mu­nity-based Asian cohorts (Auyeung et al., 2020) only population data for those age >60y were included, with the lowest quintile (kg) and mean (SD) kg by sex and age group presented, as shown in Table 16b.15.

Table 16b.15 Sex- and age-specific lowest quintile and means of handgrip strength (kg) in subjects age > 60y from com­mu­nity-based Asian cohorts. Data abstracted from Auyeung et al. (2020).

Age (y)	n	Lowest quintile (kg)	Mean (SD) (kg)
Men
60‑69.9	5319	32.7	37.9 (6.5)
70‑79.9	5317	28.0	33.3 (6.3)
≥ 80	1554	23.6	28.4 (6.2)
Women
60‑69.9	6384	20.0	23.6 (4.6)
70‑79.9	6009	17.8	21.1 (4.5)
≥ 80	1761	14.7	18.3 (4.5)

Note that because all these normative refer­ence values are based on cross-sectional data, they are likely to under­estimate indi­vid­ual decline and hence should not be used to monitor the trajectories of indi­vid­uals. Moreover, as with all cross-sectional studies, such a design limits the degree to which causal and age-related inference can be drawn (Perna et al., 2015).

Differences exist in both the methods and functional outcomes used to define cutoff values for dynamometry. So far, the cut-off values have not been validated. Table 16b.16 lists the proposed lower limits for HGS for males and females in several countries (Vaishya et al., 2024).

Table 16b.16 The proposed lower limits for hand grip strength for males and females in several countries (Vaishya et al., 2024).

Location	Guidelines	Mean Age	Males	Females
Europe	EWGSOP2 (Cruz-Jentoft et al., 2019).	77.0y	<27kg	<16 kg
USA	FNIH (Studenski et al., 2014)).	M: 75.2y F: 78.6y	<26kg	<16 kg
Asia	AWGS 2019 Update 2020) (Chen et al., 2020).	Not Available	<28kg	<18kg
India	Sarco-CUBES (Pal et al., 2021).	44.4y	<27.5kg	<18kg

Accurate measure­ment of HGS requires use of a calibrated handheld dynamometer and stan­dard­ized measure­ment procedures (Vaishya et al., 2024). The hydraulic hand dynamometer Jamar measures grip force (kgf) and is accepted as the gold standard instrument. The Jamar dynamometer is small and portable, although relatively heavy (i.e.,1.5 lb), with a dial that reads force in both kilograms and pounds, and allows assess­ment to the nearest 1kg or 2.5 lb.

In Asia, the spring-type dynamometer (Smedley) that detects the amount of spring tension (kgf) is more widely used. Data generated by these two devices are not comparable (Kim & Shinkai, 2017). Never­theless, the AWGS 2019 recommend using either device, provided standard measure­ment protocols for the device are followed (Chen et al., 2020). Box 16b.7 presents a protocol for performing hand grip strength recommended by Vaishya et al. (2024).

Measure­ment of grip strength obtained by dynamometry appears to have good to excellent relative test-retest reliability, even among older adults (2017). Dynamometers should be calibrated every 4‑6mos to main­tain longitudinal validity.

Dominant HGS occurs when the measure­ment is taken in the subject's dominant hand and can provide insight into strength available for daily activities. Some researchers have suggested using a Relative Handgrip strength in which grip strength is stan­dard­ized to BMI to account for differ­ences in body size (Vaishya et al., 2024).

Acknowledgments

The author is very grateful to the late Michael Jory who after initiating the HTML design worked tirelessly to direct the transition to this HTML version from MS-Word drafts. James Spyker’s ongoing HTML support is much appreciated.