Introduction
Undernutrition can be defined as ‘a state resulting from lack of intake or uptake of nutrition that leads to altered body composition (decreased fat free mass) and body cell mass leading to diminished physical and mental function and impaired clinical outcome from disease’(Reference Sobotka1). A commonly used synonym of undernutrition is malnutrition, and these terms are often used interchangeably in the literature. However, where possible, the term undernutrition is preferred, as malnutrition in theory includes both under- and overnutrition. Previously, three main types of undernutrition were defined(Reference Cederholm, Barazzoni and Austin2). These include disease-related undernutrition with inflammation (e.g., cancer cachexia), disease-related undernutrition without inflammation (e.g., non-cachectic undernutrition), and undernutrition without disease (e.g., hunger-related undernutrition).
Since 2019, a global consensus on how to diagnose undernutrition is made available by the Global Leadership Initiative on Malnutrition (GLIM)(Reference Cederholm, Jensen and Correia3,Reference Jensen, Cederholm and Correia4) . A two-step approach was advised. First, screening to identify who is likely to be undernourished (often called ‘at risk’) should take place using a validated screening tool. While a previous study identified 48 screening tools being applied to older adults in the literature, a much smaller number of tools have been developed for and/or are validated in older adults for a specific setting(Reference Power, de van der Schueren and Leij-Halfwerk5). Thus, care should be taken to select the proper tool for screening purposes. Second, in those considered ‘at risk’ a full diagnosis should take place by assessing three phenotypic criteria (weight loss, low BMI, and low muscle mass) and two etiologic criteria (reduced food intake or assimilation, and inflammation). When at least one phenotypic and one etiologic criterion is met, the diagnosis undernutrition can be made. Extensive guidelines on how to assess low muscle mass and inflammation have been published more recently by GLIM to further stimulate implementation(Reference Jensen, Cederholm and Correia4,Reference Barazzoni, Jensen and Correia6,Reference Cederholm, Jensen and Ballesteros-Pomar7) .
Some criteria of undernutrition overlap with the criteria of other age- and nutrition-related concepts such as frailty and sarcopenia (Figure 1). For example, recent weight loss is one of the Fried criteria for frailty(Reference Fried, Tangen and Walston8), and low muscle mass is one of the criteria of the European(Reference Cruz-Jentoft, Bahat and Bauer9) and the global definition of sarcopenia(Reference Kirk, Cawthon and Arai10). Despite the fact that some characteristics are shared between these concepts, and that undernutrition frequently is present in older adults who are frail and/or sarcopenic, it is important to emphasise that undernutrition is a concept very distinct from frailty and sarcopenia.
Figure 1. The overlap between undernutrition and other age-related concepts.
The aim of this paper is to review several key aspects of undernutrition in later life, with a major focus on undernutrition in community-dwelling older adults.
Prevalence of undernutrition
The prevalence of undernutrition in older adults reported in individual studies varies due to differences in the setting (hospital, long-term care or community), sample characteristics (such as age, functioning level, and health status) and method to identify undernutrition (screening tool or diagnosis).
In an extensive review, prevalence data based on 583,972 older adults from 196 studies and 24 European countries were pooled(Reference Leij-Halfwerk, Verwijs and van Houdt11). Studies were only included when a screening tool validated for older adults was used. Overall, the prevalence of being at high risk of undernutrition was 22.6% (95% confidence interval 20.9% to 24.3%). The prevalence was highest in the hospital setting based on 127 studies (Figure 2). Within the hospital setting, the highest prevalence of undernutrition was previously observed in the medical speciality geriatrics(Reference Kruizenga, van Keeken and Weijs12). In the residential care setting the prevalence of being at high risk was 17.5% based on 30 studies, while the lowest prevalence, based on 32 studies, was observed among older adults living at home (8.5%)(Reference Leij-Halfwerk, Verwijs and van Houdt11).
Figure 2. European prevalence of high malnutrition risk in adults aged 65 years and older based on validated screening tools, according to setting. Prevalence is reported as percentage with 95% confidence intervals. Based on Leij-Halfwerk S et al. (2019) Maturitas 126, 80–89.
While the reported prevalence of undernutrition in older adults is generally lowest in the community setting, its prevalence is substantially higher in specific subgroups. A study among 1745 Dutch adults aged 65 years and older (mean age 74 years), showed an overall prevalence of high undernutrition risk of 8.5%(Reference Zügül, van Rossum and Visser13). However, the prevalence was higher in those 85 years and older (19.3%), those reporting a poor appetite (22.4%), those reporting mobility limitations (11.5%), those living alone (13.5%), those receiving home care (15.8%), and those with poorer self-reported health (12.2%). Thus, among community-dwelling older adults, subgroups are present that show prevalence rates approaching those of older adults in the hospital setting.
As stated above, the GLIM advices to first screen for undernutrition. When screened positive, a full diagnosis of undernutrition should be performed. In the previously mentioned Dutch prevalence study, and selecting those older adults who had complete data on all required criteria, the prevalence rate was 5.4% based on the screening tool, 7.1% based on GLIM diagnosis, and only 3.1% when a positive screening was followed by a positive GLIM diagnosis(Reference Zügül, van Rossum and Visser13). These results importantly suggest that undernutrition cases are being missed when the diagnosis of undernutrition has to be preceded by a positive screening, and that the two-step GLIM approach thus may lead to an underestimation of the problem. This observation, which was made in other studies as well(Reference Van Dronkelaar, Tieland and Cederholm14), may be dependent on the choice of the malnutrition risk screening tool, and is currently being discussed within the GLIM community(Reference Jensen, Cederholm and Correia4). Overall, as different screening and diagnostic tools are being used to assess undernutrition, it remains very important to consider which tools have been used in order to properly interpret and compare prevalence data.
Determinants of undernutrition
Multiple factors are involved in the aetiology of undernutrition, of which some are directly causing undernutrition, while others may indirectly increase the risk of undernutrition. In a systematic review of observational studies investigating determinants of protein-energy malnutrition in older adults, strong evidence was obtained for poor appetite, and moderate evidence for edentulousness, having no diabetes, hospitalisation and poor self-reported health increasing the risk of undernutrition(Reference Van der Pols-Vijlbrief, Wijnhoven and Schaap15). A meta-analysis of six prospective datasets that were harmonised to identify determinants of incident undernutrition in community-dwelling older adults, older age, not being married, having difficulties walking or climbing stairs, and previous hospitalisation increased the risk of developing undernutrition(Reference Streicher, van Zwienen-Pot and Bardon16). In a qualitative study, the causes of undernutrition as reported by older adults themselves were revealed(Reference Van der Pols-Vijlbrief, Wijnhoven and Visser17). All participants were screened to be at high risk for undernutrition. Of interest, many participants reported multiple causes of their undernutrition. Causes within the mental, social and food & appetite domain, such as mourning, loneliness, inability to do groceries and poor food quality provided by meal services were mentioned. Other reported causes included aging, surgery and hospitalisation, and forgetfulness.
To structure the potential determinants of undernutrition identified in previous studies, a theoretical framework was developed by Volkert et al.(Reference Volkert, Kiesswetter and Cederholm18): the DoMAP (Determinants of Malnutrition in Aged Persons) model. The key aspects of this model are summarised in Figure 3. The model indicates three central mechanisms through which undernutrition may develop: reduced nutrient availability, low intake, and high requirements. Certain risk factors may directly lead to one of these central mechanisms and could be interpreted as direct causes of undernutrition. Examples are vomiting, chewing problems, lack of food, poor appetite and inflammation. These factors would require immediate attention by healthcare professionals in order to prevent the development of undernutrition or further aggravation of nutritional status. Other factors may not necessarily lead to undernutrition, but may indirectly cause undernutrition through one or more of the direct risk factors. An example is stroke, that may likely induce undernutrition when chewing and swallowing are negatively affected, but will not necessarily always lead to undernutrition.
Figure 3. Potential determinants and central mechanisms of the development of undernutrition in older adults. Based on Volkert D et al. (2019) Gerontol Geriatr Med 5, 2333721419858438.
The DoMAP model clearly highlights the large range of potential causes and mechanisms that underlie the development of undernutrition. The model contains relevant and frequent problems that are often modifiable with appropriate interventions. Thus, the model could support the initiation of preventive measures in clinical practice in order to reduce the risk of developing undernutrition.
Data from a randomised clinical trial conducted in older adults living at home, who were all screened being at high risk of undernutrition, showed that the prevalence of certain risk factors for undernutrition is high. For example, 74–81% reported to have a poor appetite, 69–74% needed support for obtaining groceries, and 73–77% reported an oral health problem(Reference Van der Pols-Vijlbrief, Wijnhoven and Bosmans19). Active screening for undernutrition risk factors and subsequent treatment of those factors that are modifiable may help prevent the aggravation of nutritional status in older adults and the actual development of undernutrition. The potential screening for risk factors of undernutrition (instead of, or in addition to, screening for the presence of undernutrition) as part of the GLIM approach is currently being discussed within the GLIM community(Reference Jensen, Cederholm and Correia4).
Consequences of undernutrition
Most of the research investigating the consequences of undernutrition has focused on mortality risk. Recent studies report an increased mortality risk across all settings for those diagnosed with undernutrition using GLIM(Reference Jensen, Cederholm and Correia4). A scoping review indicated, based on five cohort studies conducted in community-dwelling older adults, that a positive GLIM diagnosis increased mortality risk 1.6 to 4.4-fold during a follow-up of 2 to14 years(Reference Sánchez-Rodríguez, De Meester and Minon20). The independent association of two phenotypic GLIM criteria (i.e., low BMI and involuntary weight loss) with mortality in older adults has since long been established(Reference Di Angelantonio and Bhupathiraju21,Reference Wijnhoven, van Zon and Twisk22) . However, the independent association between low muscle mass in older community-dwelling adults and mortality seems less clear(Reference Newman, Kupelian and Visser23–Reference Wang, Mu and Wang25). Prospective studies conducted in community-dwelling older adults, clearly show that a positive screening for undernutrition is associated with other negative clinical outcomes, such as falls(Reference Eckert, Gell and Wingood26), rehospitalisation(Reference Wahyudi, Ronoatmodjo and Setiati27), frailty(Reference Rodríguez-Mañas, Rodríguez-Sánchez and Carnicero28), and functional decline(Reference Kiesswetter, Pohlhausen and Uhlig29). Moreover, malnutrition is also associated with poorer functional recovery and higher mortality risk after a hip fracture(Reference Inoue, Misu and Tanaka30,Reference Gumieiro, Rafacho and Gonçalves31) . Importantly, several studies suggest higher healthcare costs for community-dwelling older adults with undernutrition compared to those without undernutrition, mainly due to a higher use of healthcare resources(Reference Abizanda, Sinclair and Barcons32,Reference Rodríguez-Sánchez and Sulo33) .
It should be acknowledged that most of the knowledge on the consequences of undernutrition is based on observational studies, which do not allow drawing any conclusions regarding causality. The presence of undernutrition in older adults is often co-occurs with other health problems such as chronic disease, functional limitations, depression, and cognitive problems, which by themselves are strongly associated with poor clinical outcomes. While intuitively a causal association between undernutrition and clinical outcomes can be expected, as the body needs proper nutrition to remain well-functioning and healthy, experimental studies are necessary to causally link undernutrition to relevant clinical outcomes by testing the impact of nutritional interventions in those who are undernourished.
The EFFORT trial is a landmark study, testing an in-depth, stepwise protocol to treat undernutrition in 2088 patients (mean age 72 y), and showing a greater increase in energy intake and protein intake compared to the usual care group(Reference Schuetz, Fehr and Baechli34). In the intervention group, 79% reached their energy requirement, while this was only 54% in the control group. The nutritional intervention reduced the risk of any adverse clinical outcome by 21%, which included all-cause mortality, admission to intensive care, non-elective hospital readmission, major complications, and decline in functional status at 30 days. Of interest, only the individual outcomes all-cause mortality and decline in functional status at 30 days were statistically significant reduced. Although this study was conducted in the hospital setting, it clearly highlights the causal impact of reducing or eliminating undernutrition using optimised nutritional care on hard clinical endpoints. Furthermore, a secondary analysis showed that the in-hospital nutritional support was highly cost-effective, mainly due to the reduction in ICU admissions and hospital-associated complications(Reference Schuetz, Sulo and Walzer35).
Unfortunately, a study of similar magnitude conducted among older adults in the community setting has never been performed. Smaller nutritional intervention trials focussed on increasing energy intake and protein intake in undernourished community-dwelling older adults or those at high risk, have shown mixed results. In pooled analyses of individual-participant data from randomised controlled trials conducted in older adults, we earlier observed a clear positive impact of nutrition interventions on energy intake and body weight(Reference Reinders, Volkert and de Groot36), but not on muscle strength and mortality(Reference Schuetz, Fehr and Baechli34). Similar results were observed in systematic reviews of nutritional interventions tested in randomised controlled trials in community-dwelling older adults(Reference De van der Schueren, Wijnhoven and Kruizenga38,Reference Thomson, Rice and Arisa39) .
Unfortunately, many previous trials conducted in the community setting have methodological shortcomings, including a low sample size, low adherence to the intervention (for example, the actual use of provided oral nutrition supplements), study participants not meeting nutritional requirements, inclusion of participants who were not undernourished at baseline, short follow-up, and the inclusion of outcomes considered less responsive to interventions (such a handgrip strength)(Reference De van der Schueren, Wijnhoven and Kruizenga38,Reference Milne, Potter and Vivanti40) . These shortcomings likely contribute to an underestimation of the impact of nutritional interventions on relevant clinical outcomes in undernourished older adults, contributing to the mixed conclusions of these studies. Carefully designed and conducted trials, using a core outcome set of relevant and standardised (patient-reported) outcomes(Reference Mendonça, Avgerinou and Çavdar41), are still necessary to answer the key question: what type of nutritional intervention in what setting is effective for which older adult(Reference Norman, Haß and Pirlich42)? Although the impact of nutritional treatment on clinical outcomes in community-dwelling older adults who are undernourished remains somewhat unclear, this uncertainty should never justify poor nutritional care in older adults, as proper nutrition and hydration are human rights(Reference Beck and Volkert43).
Malnutrition awareness in the community setting
While the reported prevalence of high undernutrition risk in older adults is lower in the community setting compared to all other settings, it should be realised that in Europe about 95% of older adults live at home. Thus, most older adults with undernutrition live at home. In an individual-participant data analysis including 15 samples and 5956 older participants across three settings, two criteria for the presence of undernutrition were evaluated: low BMI and previous weight loss, thereby using a harmonised definition of undernutrition(Reference Wolters, Volkert and Streicher44). In the community setting samples, undernutrition was mostly (overall, about 69%) identified due to recent weight loss, indicating early malnutrition. In contrast, in the nursing home and institutionalised care setting, undernutrition was mostly identified due to low BMI (overall, about 66%), suggesting the presence of chronic undernutrition. In the home care and day hospital setting, as well as in the acute hospital and rehabilitation setting, the frequency of experiencing both recent weight loss as well as low BMI was highest, suggesting very serious nutritional issues that would require immediate attention. These results suggest that undernutrition in older adults often starts at home, which is supported by observations that older adults admitted to the hospital or nursing home are often already undernourished(Reference Russell and Elia45,Reference Chatindiara, Allen and Hettige46) .
Since most older adults live at home and undernutrition frequently starts at home, the community setting is a very relevant setting for initiating preventive actions in order to reduce the actual development of undernutrition, as well as for improving the detection of undernutrition at an early stage in order to prevent further decline of nutritional status. Unfortunately, studies show that the recognition of undernutrition in the community and residential setting is low. Routine screening for undernutrition and proper referral to healthcare professionals is often lacking(Reference Morley47–Reference Castro, Reynolds and Kennelly49). Moreover, older persons themselves are often not aware of their weight loss or are not concerned(Reference Avgerinou, Bhanu and Walters50,Reference Reynolds, Dominguez Castro and Geraghty51) .
Figure 4 highlights several aspects contributing to the poor recognition of undernutrition in the community setting. These aspects are based on the results of mainly qualitative studies conducted among healthcare professionals, including general practitioners, dieticians and home care nurses(Reference Morley47–Reference Castro, Reynolds and Kennelly49,Reference Ziylan, Haveman-Nies and van Dongen52–Reference Geraghty, Browne and Reynolds56) , as well as studies among older adults and their informal carers(Reference Avgerinou, Bhanu and Walters50,Reference Ziylan, Haveman-Nies and van Dongen52,Reference Winter, McNaughton and Nowson57–Reference Castro, Reynolds and Kennelly61) . These results of these studies clearly warrant the need for greater undernutrition awareness in the community setting.
Figure 4. Aspects related to the poor recognition of undernutrition in older adults in the community setting.
Several potential strategies to increase the recognition of undernutrition in primary and community care as well among older adults are summarised in Figure 5. Embedding specific courses on the topic of undernutrition (in older adults) in the medical doctors and nurses’ curricula, schooling of healthcare professionals through (online) courses, and mandatory screening for undernutrition in specific healthcare settings (such as the formal home care setting) are examples of strategies that could contribute to a greater recognition of undernutrition by healthcare professionals(Reference Eglseer, Halfens and Schüssler62–Reference Leistra, van Bokhorst-de van der Schueren and Visser64).
Figure 5. Potential strategies to increase undernutrition recognition in the community.
Interventions among community-dwelling older adults are also needed to increase undernutrition awareness. Recently, we developed the Malnutrition Awareness Scale (MAS) to objectively measure undernutrition awareness in older adults(Reference Visser, Sealy and Leistra65). The MAS is specifically developed for community-dwelling older adults and is based on the Integrated-Change model of behavioural change (I-Change model(Reference De Vries66). The MAS assesses the four domains of awareness: knowledge, cues, risk perception and cognizance. The scale contains a total of 22 items (and one priming item) covering these four domains and provides an awareness score between 0 (very low awareness) and 22 (very high awareness). The MAS was shown to be a feasible and reliable tool with good content validity to objectively assess undernutrition awareness in community-dwelling older adults, and is currently being translated and culturally adapted into several other languages. Since previous qualitative research on undernutrition awareness in older adults is limited to specific countries or regions, application of the translated MAS in more diverse regions is important to understand whether undernutrition awareness is culturally specific.
The MAS can be used to identify older adults with poor awareness, to support the development of effective interventions to increase this awareness, and to objectively evaluate the impact of these interventions. The MAS may contribute to much-needed future efforts to increase undernutrition awareness among older adults living in the community, in order to either prevent the development of undernutrition or to recognise undernutrition at an early stage and obtain proper treatment, thereby reducing the negative health and functional consequences of undernutrition.
Conclusions and recommendations
This overview highlights that undernutrition is frequently present among community-dwelling older adults, especially among vulnerable subgroups. A wide range of (modifiable) risk factors can cause or contribute to its development. Undernutrition in community-dwelling older adults has been linked to serious negative health consequences, higher use of healthcare resources and higher healthcare costs. However, undernutrition recognition by primary care professionals and by older adults themselves is generally low.
Strategies are urgently needed to increase undernutrition awareness among healthcare professionals working in primary and community care, for example by embedding courses on the topic of undernutrition in health professionals’ curricula. Implementation of the GLIM undernutrition criteria for assessing undernutrition should be supported to enhance uniform diagnosis in clinical care, comparison of prevalence data, and the identification of high-risk subgroups. Greater attention to the presence of factors contributing to the risk of undernutrition, and early treatment of these risk factors in primary and community care, may contribute to the prevention of undernutrition and its negative health consequences. Future research should identify the key modifiable risk factors to be assessed in community-dwelling older adults. Furthermore, future trials are needed to better address what type of nutritional intervention in what setting is effective for which older adult. These trials should include a core outcome set of relevant and standardised (patient-reported) outcomes. Lastly, effective interventions to increase malnutrition awareness need to be developed and implemented in order to empower older adults and their informal care givers to act at an early stage of undernutrition development.
Acknowledgements
None.
Author contributions
MV is the only author and fully responsible for the content of this manuscript.
Financial support
None.
Competing interests
None
Funding Statement
Open access funding provided by Vrije Universiteit Amsterdam.
Open access
