The difference between sarcopenic obesity and malnutrition in obesity

Sarcopenic obesity and malnutrition in obesity share some similarities but also exhibit distinct characteristics. Sarcopenic obesity is defined by the coexistence of obesity (high fat mass) and sarcopenia (low muscle functionality and altered body composition)^{(Reference Donini, Busetto and Bischoff31–34)}. In contrast, malnutrition in obesity refers to undernutrition related to specific energy or nutrient deficiencies (such as PEM) within the context of obesity^{(Reference Saunders and Smith11,34)} .

Despite common factors like high fat mass, reduced muscle mass, systemic inflammation and consequences such as a higher risk of morbidity and mortality^{(Reference Donini, Busetto and Bauer33)}, recognising the differences between the two conditions is important. Malnutrition in obesity encompasses a broader range of nutritional deficiencies, including unintentionally reduced energy and/or nutrient intake leading to weight loss, setting it apart from sarcopenic obesity^{(Reference van der Pols-Vijlbrief, Wijnhoven and Schaap12)}. Sarcopenic obesity manifests in two forms: age-related (primary) and disease-related (secondary). Both involve a decline in muscle function and altered body composition, which may not necessarily be attributed to nutritional deficiencies^{(Reference Cruz-Jentoft, Bahat and Bauer35,Reference Muscaritoli, Anker and Argilés36)} . Figure 1 illustrates the key characteristics of sarcopenic obesity and malnutrition in obesity, effectively highlighting their differences and similarities.

Fig. 1.

Differences and similarities of sarcopenic obesity and malnutrition in obesity.

A notable difference between these conditions lies in their prevalence across different age groups. Although sarcopenic obesity, particularly the secondary form, associated with specific comorbidities such as cancer or rheumatoid arthritis, can occur in individuals under 65 years, the primary form is more prevalent and particularly observed in those aged 65 and older^{(Reference Cruz-Jentoft, Bahat and Bauer35,Reference Carneiro, Mazurak and Prado37,Reference Gupta and Kumar38)} . Sarcopenic obesity in the elderly is mainly attributed to age-related hormonal changes that can lead to a decrease in hormones critical for maintaining muscle mass, such as insulin-like growth factor-1 (IGF-1), testosterone, dehydroepiandrosterone (DHEA) and oestrogen^{(Reference Cruz-Jentoft, Bahat and Bauer35,Reference Cruz-Jentoft, Baeyens and Bauer39)} . In contrast, malnutrition in obesity can occur across the lifespan, affecting individuals of different ages^{(Reference Wells, Sawaya and Wibaek15,17,Reference Bird, Murphy and Ciappio40)} .

Another difference between sarcopenic obesity and malnutrition in obesity is the focus on muscle function. While both conditions commonly exhibit reduced muscle mass and function, the algorithm for defining sarcopenic obesity prioritises muscle function over muscle mass^{(Reference Landi, Camprubi-Robles and Bear10,Reference Gortan Cappellari, Guillet and Poggiogalle32)} . Additionally, impaired physical performance is integral to the definition^{(Reference Donini, Busetto and Bauer33)}. Diagnostic procedures for sarcopenic obesity begin with an assessment of skeletal muscle function, followed by an assessment of body composition. The confirmation of the diagnosis involves identifying the presence of excess adiposity and low skeletal muscle mass or related body compartments^{(Reference Cruz-Jentoft, Baeyens and Bauer39,Reference Cederholm, Barazzoni and Austin41)} .

Furthermore, a difference emerges in the screening and assessment approaches for these two conditions. For sarcopenic obesity, a diagnostic procedure has already been proposed. This starts with screening by evaluating high fat mass through measures such as BMI and waist circumference, along with surrogate markers for sarcopenia. The subsequent diagnosis then includes an analysis of muscle function and body composition^{(Reference Donini, Busetto and Bischoff31)}. In contrast, there are no specific screening or assessment tools available for identifying malnutrition within obesity. Despite the existence of screening and assessment methods for sarcopenic obesity, a notable similarity arises: the absence of specific cut-offs for both conditions. The Sarcopenic Obesity Global Initiative (SOGLI) expert panel has recognised the lack of specific cut-offs for sarcopenic obesity and formulated research questions to further define muscle function and body composition measures for sarcopenic obesity^{(Reference Gortan Cappellari, Guillet and Poggiogalle32,42)} . Meanwhile, the group supports the use of BMI and waist circumference cut-offs while recognising their limitations. Similarly, cut-offs for muscle function and muscle mass have been suggested, with a recommendation to adopt them from previous studies^{(Reference Dodds, Syddall and Cooper43–Reference Batsis, Barre and Mackenzie47)}. This highlights the need for future research to define optimal cut-off points for application in both research and clinical practice for sarcopenic obesity, as well as for screening and assessing malnutrition in overweight and obesity, where defined cut-off points are currently non-existent.

The challenge with diagnosing malnutrition in obesity

Despite the widespread use of the GLIM criteria for diagnosing malnutrition, challenges arise when applying these criteria to patients with obesity. The phenotypic criterion of low BMI is not useful in individuals living with obesity^{(Reference Dickerson, Andromalos and Brown28)}. Moreover, the suggested cut-off values for recent weight loss and reduced muscle mass in the GLIM guidelines may also not be applicable for individuals living with obesity as these cut-off points are established in studies with individuals with a normal BMI^{(Reference Cederholm, Jensen and Correia26,Reference Correia, Tappenden and Malone48)} .

Weight loss can be defined and reported using different measures, including percentage weight loss, changes in weight in kilograms and changes in BMI (kg/m²)^{(Reference Deitel, Gawdat and Melissas49)}. In the GLIM criteria, unintentional weight loss is specifically defined and reported as a percentage of an individual’s usual weight^{(Reference Cederholm, Jensen and Correia26)}. This choice can lead to major differences when interpreting weight loss outcomes, especially in the context of obesity, with a pronounced impact on individuals living with obesity. To illustrate this, let us consider two individuals with different body weights. A weight loss of 3 kg may greatly impact an individual weighing 60 kg, representing a 5% reduction in body weight, which is considered clinically relevant if assumed over 3 months^{(Reference Douketis, Macie and Thabane50)}. However, in an individual weighing 160 kg, the same 3 kg weight loss would represent only a 1·9% reduction in body weight, which may not be as clinically relevant. This discrepancy highlights how using percentages for weight loss in individuals living with obesity may require a considerable number of kilograms lost before reaching the threshold for malnutrition diagnosis, making it challenging to accurately assess malnutrition risk in this population.

In individuals living with obesity, involuntary weight loss impacts all-cause mortality, but the severity is less pronounced than in individuals with normal weight and underweight^{(Reference De Stefani, Pietraroia and Fernandes-Silva51)}. Research indicates that ±5% involuntary weight loss in a population with obesity, unlike that in a population with a normal BMI, is not associated with increased mortality, and higher weight loss percentages might be more indicative of malnutrition risk^{(Reference Aldahhan, Motawei and Al-Hariri52)}. NHANES data reveal that weight loss beyond 15% increases mortality risk in individuals living with obesity, with no significant association for losses between 5% and −15%, except among women who are overweight^{(Reference Ingram and Mussolino53)}. Hospitalised patients’ data (from the nutritionDay data collection) reinforces this, showing that mortality risk increases with weight loss across all BMI categories, but in the BMI category above 30 kg/m², this becomes significant only when the loss exceeds 12·6%^{(Reference Barazzoni, Sulz and Schindler54)}. Additionally, distinguishing between fat and muscle mass loss during weight loss is important, given the increased risk of conditions like sarcopenia associated with muscle mass loss^{(Reference Cava, Yeat and Mittendorfer55)}. Furthermore, malnutrition diagnostic criteria that investigate unintentional weight loss may also face challenges in assessing malnutrition in patients with obesity. Due to the desire for weight loss among many individuals living with obesity, they may engage in unhealthy weight loss practices and not report weight loss as unintentional, possibly leading to an increased malnutrition risk that goes unnoticed^{(Reference Dickerson, Andromalos and Brown28,Reference Barazzoni, Breda and Cuerda56)} .

Assessing muscle mass is vital in nutritional assessment for individuals living with obesity but presents specific challenges. Studies suggest that a higher BMI is associated with a higher muscle mass due to the additional weight, comprising both fat and muscle^{(Reference Cava, Yeat and Mittendorfer55,Reference Tallis, James and Seebacher57)} . This association can lead to an overestimation of muscle mass when using methods like upper-arm or calf circumference, as excess fat can distort measurements, potentially underdiagnosing malnutrition^{(Reference Sizoo, de Heide and Emous58)}. To address this, there have been proposals for age and BMI-specific cut-off points for calf circumference^{(Reference Gonzalez, Mehrnezhad and Razaviarab59)}. Furthermore, techniques such as dual-energy X-ray absorptiometry (DEXA) and computed tomography (CT) scans have a weight limitation, that is, the maximum body weight that the equipment is designed to accommodate while still providing accurate results. This limitation can pose difficulties in obtaining precise measurements in individuals living with obesity. Bio-electric impedance (BIA) measurements are less reliable in this population due to the overestimation of fat-free mass (FFM) by BIA equations, probably explained by a higher hydration of FFM in adipose tissue^{(Reference Jensen, Braun and Geisler60)}. Conversely, ultrasonography shows potential as a non-invasive and accessible technique for accurately evaluating muscle mass and its characteristics in individuals living with obesity. However, challenges such as a lack of standardisation need to be addressed for its full integration into clinical practice^{(Reference Simó-Servat, Ibarra and Libran61)}.

Additionally, these techniques can also misclassify reduced muscle mass within obesity, where the difference between muscle mass in absolute terms (kg) and relative terms (%) is high^{(Reference Sizoo, de Heide and Emous58,Reference Kyle, Bosaeus and De Lorenzo62–Reference Brownbill and Ilich64)} . Since cut-off values are based on absolute terms, individuals living with obesity may have a relatively low muscle–fat ratio but still possess a relatively high muscle amount in kg, especially with high adipose tissue. Thus, many individuals could exceed the cut-off for reduced muscle mass, potentially escaping malnutrition diagnosis^{(Reference Kruizenga, Seidell and De Vet25)}.

Applying the GLIM aetiologic criterion of low food intake (a 50% reduced food intake cut-off) in individuals living with obesity presents challenges^{(Reference Cederholm, Jensen and Correia26)}. Bariatric surgery, especially gastric sleeve surgery, has gained popularity as an alternative treatment for obesity, particularly in those with morbid obesity (BMI >40 kg/m²)^{(Reference Gulinac, Miteva and Peshevska-Sekulovska65)}. However, this surgery has implications for the dietary intake of individuals with obesity, resulting in a decrease in nutrient intake and absorption, potentially compromising accurate screening of the GLIM criterion related to food intake^{(Reference Mohapatra, Gangadharan and Pitchumoni66)}. Moreover, this criterion may not accurately reflect inadequate nutrient intake in individuals living with obesity due to their higher nutritional requirements based on their higher body weight^{(Reference Kaidar-Person, Person and Szomstein18)}. Based on literature, the most reliable energy intake data indicate high levels, exceeding 4000 kcal/d (16,736 kJ/d) for individuals maintaining weight stability at the highest levels of morbid obesity^{(Reference Harbury, Verbruggen and Callister67)}. Thus, a 50% intake reduction in this population remains a substantial intake. Furthermore, determining the appropriate cut-off values for reduced food intake in individuals living with obesity accustomed to overeating is challenging. This is due to their different eating patterns and nutritional requirements compared with those with healthy body weights^{(Reference James68)}.

Inflammation triggers increased protein turnover, which leads to the loss of muscle mass, strength and function^{(Reference Costamagna, Costelli and Sampaolesi69)}. For this reason, inflammation has been incorporated into the aetiologic criteria of the GLIM definition^{(Reference Cederholm, Jensen and Correia26)}. Determining suitable thresholds for disease burden and inflammation in individuals living with obesity remains uncertain. Obesity often involves a mild, chronic inflammation driven by adipose tissue releasing pro-inflammatory cytokines^{(Reference Calder, Ahluwalia and Brouns70)}, consequently increasing C-reactive protein (CRP) levels^{(Reference Monteiro and Azevedo71)}. An individual meets the GLIM aetiologic criterion of inflammation if they show repeated CRP levels above 3·0 mg/L^{(Reference Cederholm, Jensen and Ballesteros-Pomar72)}. In individuals with metabolic syndrome, obesity primarily drives CRP elevation, typically exceeding 1·0 mg/L, indicating inflammation^{(Reference Aronson, Bartha and Zinder73)}. Thus, many individuals living with obesity would meet the GLIM criteria’s inflammation aetiologic criterion, potentially resulting in false positives due to the prevalent low-grade inflammation in obesity^{(Reference Kruizenga, de Vet and Van Marissing74)}. The applicability of the provided guidance for the GLIM criteria’s inflammation aetiologic criterion may vary within the population with obesity due to the variability of obesity-related inflammation, influenced by factors such as obesity duration, genetic predisposition and underlying health conditions^{(Reference Wijnhoven, Schilp and van Bokhorst-de van der Schueren13)}.

Moreover, the coronavirus disease 2019 (COVID-19) pandemic highlighted major differences in disease outcomes as a result of systemic inflammation stemming from pre-existing conditions in patients such as diabetes and obesity^{(Reference Buicu, Cernea and Benedek75)}. The guidance for the GLIM criteria’s inflammation aetiologic criterion outlines a categorised list of example diseases based on inflammation levels ranging from mild to severe^{(Reference Cederholm, Jensen and Ballesteros-Pomar72)}. The risks of these diseases might be different between individuals living with obesity and those who do not live with obesity, with individuals living with obesity carrying higher risks of severe consequences^{(Reference Aghili, Ebrahimpur and Arjmand76)}.

Obesity can alter disease experiences, potentially worsening disease severity due to factors such as compromised immune function, increased disease risk, additional stress on the cardiovascular and respiratory systems and limited treatment effectiveness^{(Reference Kalligeros, Shehadeh and Mylona77,Reference Hruby and Hu78)} . This can result from altered pharmacokinetics and potential side effects of disease treatments, such as chemotherapy^{(Reference Purcell, Kok and Ketterl79,Reference Smit, De Hoogd and Brüggemann80)} . Consequently, managing chronic conditions becomes more challenging as surgical risks increase and certain diseases such as congestive heart failure, diabetes and rheumatoid arthritis may progress faster in the presence of obesity^{(Reference Barazzoni, Bischoff and Boirie81)}.

In summary, both the phenotypical and aetiological aspects of the widely accepted GLIM criteria for diagnosing malnutrition may not fully address the complexities of diagnosing malnutrition in the context of individuals living with obesity as the cut-off and threshold values used are not specifically tailored for the population with obesity. Further research and adaptations of the criteria should aim at integrating adapted cut-off values within the GLIM criteria to also suit the specific nutritional challenges faced by individuals living with obesity, similar to the approach taken with ethnicity-specific cut-offs. The current reference values in the GLIM criteria may not be entirely applicable to this population, warranting further investigation and modification to accurately identify and diagnose malnutrition in individuals living with obesity.

The challenge with current malnutrition screening tools

The foundation of malnutrition diagnosis according to the GLIM criteria relies on the use of a validated screening tool^{(Reference Cederholm, Jensen and Correia26)}. Various screening tools have been developed and validated to assess malnutrition in different populations and healthcare settings^{(Reference Leij-Halfwerk, Verwijs and Houdt82–Reference Schueren, Guaitoli and Jansma84)}. Table 1 provides an overview of frequently used and validated malnutrition screening tools^{(Reference Schueren and Jager-Wittenaar85)}. These tools encompass a range of criteria, incorporating phenotypic aspects such as unintentional weight loss, low BMI and reduced muscle mass, alongside aetiologic indicators such as reduced food/fluid intake, disease burden/inflammation and risk factors for malnutrition^{(Reference Cederholm, Jensen and Correia26,Reference Schueren and Jager-Wittenaar85,Reference Borkent, Schuurman and Beelen86)} .

Table 1.

Overview of criteria in a selection of frequently used malnutrition screening tools (from de van der Schueren, 2022)

¹ DETERMINE Your Nutritional Health Nutrition Screening Initiative.

² Geriatric Nutrition Risk Index.

³ Mini Nutritional Assessment Short Form.

⁴ Malnutrition Screening Tool.

⁵ Malnutrition Universal Screening Tool.

⁶ Nutritional Risk Index.

⁷ Nutrition Risk Score.

⁸ Nutritional Risk Screening 2002.

⁹ Nutritional Form for the Elderly.

¹⁰ Patient-Generated Subjective Global Assessment Short Form.

¹¹ Seniors in the Community: Risk Evaluation for Eating and Nutrition, version II.

¹² Short Nutritional Assessment Questionnaire.

¹³ Short Nutritional Assessment Questionnaire for the Residential Care.

¹⁴ Short Nutritional Assessment Questionnaire 65+.

¹⁵ Simplified Nutritional Appetite Questionnaire.

Most malnutrition screening tools, following the 2002 European Society for Clinical Nutrition and Metabolism (ESPEN) guidelines, primarily rely on low BMI and recent (unintentional) weight loss to identify malnutrition risk^{(Reference Correia, Tappenden and Malone48,Reference Kondrup, Allison and Elia87)} . As previously mentioned, a low BMI (<18·5 kg/m²) is not useful, and traditional cut-off values for unintentional weight loss might not apply to individuals living with obesity, despite inclusion in these screening tools. In addition, because of the emphasis on weight loss, healthcare providers might not consider screening individuals living with obesity for malnutrition risk^{(Reference Dickerson, Andromalos and Brown28)}. In widely used malnutrition screening tools such as the MNA-short form (MNA-SF) and MUST, there are distinct differences in BMI cut-off points^{(Reference Neelemaat, Meijers and Kruizenga88)}. The MNA-SF sets a cut-off at 23 kg/m^{2(Reference Kaiser, Bauer and Uter89)}, whereas the MUST adopts a different approach with a cut-off of 20 kg/m²⁽⁹⁰⁾. Nevertheless, both tools may not be suitable for individuals living with obesity, and their underlying scoring system might therefore not apply to this specific population. Thus, a pertinent question arises of whether the traditional malnutrition screening tools are still appropriate^{(Reference Schueren and Jager-Wittenaar85)}.

Moreover, it is crucial to note that most malnutrition screening tools were validated against ‘gold standards’^{(Reference Schueren and Jager-Wittenaar85,Reference Kondrup, Allison and Elia87)} , primarily focused on phenotypical criteria such as unintentional weight loss, low BMI and reduced muscle mass. Thus, many screening tools lack aetiological criteria. In addition, the use of validated tools specific to the setting and population is of utmost importance^{(Reference Phillips, Foley and Barnard91)}. Within the context of obesity, only one screening tool has been proposed for assessing malnutrition: the JaNuS tool^{(Reference Donini, Ricciardi and Neri92)}. This tool screens for over- and undernutrition in two separate sections, enabling a patient to score positively for one or both conditions. However, the JaNuS tool was validated in a pre-geriatric population and uses low albumin and low lymphocyte count as part of its criteria to assess nutrition status, which poses limitations to its universal applicability^{(Reference Donini, Ricciardi and Neri92,Reference Bharadwaj, Ginoya and Tandon93)} . This further emphasises the urgent need to develop and validate specific tools designed to accurately identify malnutrition in this context.

In many malnutrition screening tools, assessing reduced muscle mass is often overlooked, except in the MNA-SF and SNAQ⁶⁵, which use calf and arm circumferences^{(Reference Schueren and Jager-Wittenaar85,Reference Borkent, Schuurman and Beelen86)} . Again, the traditional cut-off values for these measurements may not apply to individuals living with obesity^{(Reference van Baar, Winkels and Brouwer94)}. Due to higher fat mass presence, muscle mass and predominantly fat mass will be included in the measurement, rendering it inaccurate^{(Reference Sizoo, de Heide and Emous58)}. Thus, to effectively address this challenge, questions need to be adapted to include appropriate cut-offs specifically tailored to the context of obesity.

Malnutrition screening tools typically strongly rely on anthropometric measurements and focus only to a lesser extent on nutrient intake^{(Reference Cederholm, Barazzoni and Austin41)}. Most screening tools lack comprehensive questions regarding nutritional intake, let alone their applicability to obesity^{(Reference Neelemaat, Meijers and Kruizenga88)}. As a result, these tools often fail to effectively identify the underlying nutritional issues. Notably, questions related to nutrient intake in malnutrition screening tools tend to focus solely on quantity, without considering the distinction between the quality and quantity of nutrient consumption. This oversight is particularly important given that obesity is influenced by the quality of foods consumed rather than just their quantity^{(Reference Hawkes95,Reference Popkin96)} . Furthermore, screening tools often fail to consider an individual’s specific dietary preferences and restrictions, which could lead to incomplete dietary pattern assessments due to reliance on binary yes-or-no responses^{(Reference Schueren, Guaitoli and Jansma83,Reference Neelemaat, Meijers and Kruizenga88)} . This limitation might undermine the accurate evaluation of nutritional risks.

Besides the phenotypic and aetiological criteria, malnutrition screening tools take various risk factors into account, as shown in Table 1^{(Reference Schueren and Jager-Wittenaar85)}. However, not all these factors apply to obesity. For instance, appetite regulation differs in individuals living with obesity due to disrupted mechanisms of appetite control and body weight maintenance, leading to distinct changes in appetite and dietary intake responses^{(Reference Klok, Jakobsdottir and Drent97)}. Satiety may also not be a suitable malnutrition risk factor for individuals living with obesity given the hormonal imbalances and metabolic dysregulation often present in obesity, potentially resulting in reduced satiety^{(Reference Lean and Malkova98)}. This contrasts with cases of malnutrition where reduced appetite and satiety are more common. Hence, the importance of satiety as a malnutrition risk factor could vary between individuals living with obesity and those with a healthy weight.

Certain screening tools, including the MNA and SCREEN II, extend the standard weight loss question to incorporate additional obesity-related factors such as intentions for weight change and perceptions of body weight^{(Reference Borkent, Schuurman and Beelen86)}. However, relying on self-perception of nutritional status in individuals living with obesity could be compromised by societal norms and personal biases, potentially leading to an underestimation of malnutrition risk^{(Reference Robinson, Haynes and Sutin99)}. This challenge highlights the need for more objective measures in nutritional screening. Furthermore, tools designed to assess self-perception can also be influenced by the desire for weight loss, possibly underestimating nutritional sufficiency and leading to a higher nutritional risk^{(Reference Robinson and Kersbergen100)}. Hence, relying solely on self-reported perceptions, especially in cases of obesity, may not ensure accurate assessments.

There is also a noteworthy gap in screening tools regarding malnutrition risk factors associated with obesity. For example, these screening tools often omit recurrent cycles of weight loss and gain resulting from yo-yo dieting and unhealthy weight loss practices, which can reveal underlying nutritional challenges^{(Reference Montani, Schutz and Dulloo101)}. These cycles should not be ignored, as they pose potential risks including metabolic adaptations, loss of lean muscle mass, psychological consequences, cardiovascular health concerns and the potential for disordered eating^{(Reference Contreras, Schriever and Pfluger102)}. Screening an individual living with obesity’s treatment history is equally crucial but not incorporated into screening tools. Consider the already mentioned bariatric surgery, for instance. It is important due to its association with reduced protein intake resulting from post-surgical dietary restrictions and intolerance to protein-rich foods, thus acting as a risk factor for malnutrition^{(Reference Mohapatra, Gangadharan and Pitchumoni66)}.

In conclusion, current malnutrition screening tools may not accurately identify malnutrition in the population with obesity, a concern heightened by the rising global obesity rates. Our examination of this issue has revealed several knowledge gaps that need to be filled.

Recommendations for future research, within our and other projects

As we look ahead, our research efforts aim to centre around the following recommendations. These suggestions stem from our insights into challenges such as differentiating between sarcopenic obesity and malnutrition in obesity, improving diagnostic accuracy and updating screening tools for identifying malnutrition within obesity. By adopting these recommendations, we can facilitate more precise identification, interventions and care for individuals with both obesity and malnutrition.