Nutritional status study of inpatients in hospitals of Galicia

Article metrics



To know the prevalence of malnutrition and to validate a nutritional screening protocol (SP) in patients hospitalised in Hospitals representative of inpatients admitted for acute illnesses in Galicia.


Cross-sectional study of 376 randomised patients (189 female, 210 65 y old) from 12 public hospitals admitted to hospital for acute medical, surgical or trauma illnesses. The patients elicited were evaluated by a simple SP, including variables relative to recent weight changes, serum albumin, lymphocytes, food ingestion and diagnosis at admission (Cardona's Protocol), and with a diagnostic protocol (DP, Subjective Global Assessment). Both SP and DP were performed by personnel trained in nutritional evaluation. Results of SP and DP were compared; principal factors related to malnutrition were also analysed; statistical significance was considered at P<0.05.


From patients studied, according to DP 169/360 (46.94%), patients presented malnutrition (134 B category and 35C category). SP rate was significantly related to severity of malnutrition detected by DP (P<0.001). The principal factors related to the presence of malnutrition were older age and degree of metabolic stress.


In adult patients admitted for acute illnesses, the prevalence of protein-energy malnutrition is high. The risk was related to age and to metabolic stress. The risk of malnutrition in a hospital setting is evaluated appropriately by a simple screening procedure that may contribute to detecting and correcting malnutrition risk.


Malnutrition is a frequent cause of morbidity and mortality, which is present in several risk populations. Among the more affected groups are inpatients hospitalised for a variety of causes.

There is a high prevalence of malnutrition in hospitalised patients, as has been recognised in several studies in Spain and other similar countries (McWhirter & Pennington, 1994; Thorsdottir et al, 1999; González Sánchez et al, 2000; The malnutrition prevalence group, 2000; Aznarte Padial et al, 2001).

Among the causes of malnutrition in inpatients, are mentioned the following:

  • Previous malnutrition because of subjacent illness: anorexia, feeding difficulties and increase of nutritional requirements.

  • ‘Ex novo’ malnutrition during hospital stay: deficiencies in hospital diets, fasting periods because of diagnostic or therapeutic procedures and metabolic stress associated with illness.

It is possible that the existence of other problems in a hospital setting favours the appearance or non-detection of the problem: lack of knowledge among sanitary workers, no screening protocols, nonfulfilment of protocols including nutritional parameters, lack of Nutritional Units or those that are frequently understaffed and lack of registered dieticians in public hospitals.

All the causes that generate malnutrition in 30–35% of hospitalised patients, do not seem to have changed in the last few decades (Roldán et al, 1985; de Ulíbarri Pérez et al, 2002).

The consequences of malnutrition on morbimortality of inpatients are important, because several systems are affected: immunity, respiratory, cardiocirculatory, metabolism, tissue and wound healing, and others. There are multiple studies indicating the presence of malnutrition as a marker of a worse prognosis, increasing postsurgery complications, mortality rates, hospital stay and even influencing readmission rates. It has been calculated that this implies an increase of hospitalisation costs of upto 60% (Roldán et al, 1985).

Several methods have been proposed for detecting the presence or risk of malnutrition in the hospital setting.Those that have demonstrated their utility are the Subjective Global Assessment (SGA) (Detsky et al, 1987), Cardona's method (Cardona, 1998), and in several populations, the Mini-Nutritional Assessment method (MNA) (Guigoz et al, 1996) or Cardona's method, as previously applied by our group (Martínez Olmos et al, 2002).

Because of all the outdated methods, and because of the lack of knowledge of the actual situation of this problem in the hospitals of our region, we proposed this multicentric study, with the following objectives:

  1. 1)

    To determine the malnutrition rates in hospitalised patients in public hospitals in Galicia.

  2. 2)

    To find out the predominating types and degrees of malnutrition in this population.

  3. 3)

    To determine the principal risk factors associated with the presence of hospital malnutrition in our setting and its relationship with subjacent illness.

  4. 4)

    To estimate the value of screening methods by their correlation with nutritional diagnosis methods, in order to select those of better sensitivity and specificity.

Population and methods

Study population

The study population consisted of adult patients (over 18 y, excluding pregnant or childbearing women, and children) admitted in all public hospitals of the Autonomic Community of Galicia, which is a north-western region of Spain and has 2.7 million inhabitants. Eligible hospitals and bed distribution are shown in Table 1. In order to obtain a representative sample, we made a stratification and randomisation of patients to be evaluated in each hospital from data on number of beds, number of patients under 65 y of age and those over or equal to 65 y of age (because previous studies show higher malnutrition rates in the latter group; Guigoz et al, 1996). The sample size needed to detect significant differences in malnutrition prevalence was estimated in 160 patients under 65 y of age and 190 patients over or equal to 65 y of age. In each hospital, staff involved in the Clinical Nutrition and Dietetics area were contacted in order to apply the questionnaire and evaluate the patients. We obtained a positive response in 12 hospitals, representative of 88.13% hospital beds (6310 out of 7160 beds) and 80.50% of inhabitants attended by the Public Health System in Galicia (2 131 369 of 2 647 746 people); the remaining hospitals declined to participate because of lack of staff. A specific±2 days (3rd to 7th February 2003) were selected for data recovery, applying a table of random number to eligible patients at each centre. Data were recorded through a common questionnaire for all participants (Appendix A1), and applied by personnel with experience in clinical nutrition and dietetics in each hospital.

Table 1 Participating hospitals, characteristics and randomisation of patients

Nutritional evaluation

Each evaluated patient underwent screening by the following nutritional scales:

  1. a)

    Screening method: Cardona's Method.

  2. b)

    Diagnosis method: Subjective Global Assessment.

Cardona's screening method for malnutrition of inpatients is based on Nagel's method (Nagel, 1991) and evaluates through a numerical scale some aspects relative to recent weight changes, serum albumin, lymphocyte count, food ingestion and diagnosis at admission (Cardona, 1998), a score of 14 or greater being considered to represent being at nutritional risk with (Appendix A2).

SGA is a nutritional status diagnosis method universally validated and reproducible when applied by skilled personnel, evaluating some aspects relative to weight changes, oral ingestion changes, gastrointestinal symptoms, functional capacity, degree of stress caused by comorbidity and a physical examination attending to subcutaneous fat, muscular mass and presence of oedema or ascities (Detsky et al, 1987).

Anthropometric parameters were recorded as previously described and compared to 50th percentiles of healthy individuals of similar age, sex and height according to standards published for the Spanish population (Alastrué et al, 1988; Esquius et al, 1993); % of each parameter with respect to reference values was calculated.

We registered functional capacity with a scale, taking into account the activity level of the subject prior to admittance and at the time of evaluation. We also registered presumed degree of metabolic stress related to diagnosis at admittance (see Appendix A1).

In addition, we registered the number of medicaments used by each patient at the time of evaluation and the presence of medicaments known to influence body weight and protein metabolism (corticosteroids, adrenergic beta-blocking agents).

Statistical analysis

For statistical analysis, the results of the screening method and diagnosis method were compared. For independent means comparison, we used the Student's t-test for unpaired data. Frequency distribution analysis was made through Pearson's χ2 test, using the Yates correction for small frequencies. In order to search for associations between different parameters, we used Spearman's linear correlation coefficient. We considered a value of P<0.05 as statistically significant. All analyses were made with SPSS v 10.0 for Windows.


General evaluation of patients

From 410 randomised patients in participating hospitals, we could finally evaluate 376 (91.70%); the remaining 34 were unavailable because of clinical instability, diagnostic procedures or discharge, representing 8.30%. In Table 2, we can see the number of evaluated patients at each hospital, 189 (50.13%) women and 210 (55.85%) 65 y old. The mean age was 63.67±18.41 y (range 14–94 y). With respect to distribution of patients in medical services, 197 (52.39%) patients were admitted to medical services and 157 (41.76%) patients were hospitalised in surgical or intensive-care services. All of them were admitted because of acute diseases or chronic diseases with acute exacerbation. The distribution by diagnostic groups is presented in Table 3; at the time of evaluation, they were admitted for a mean of 15.43±24.48 days. Diabetes mellitus was present in 64 patients (17.0%). Of the 376 patients studied, 21 (5.58%) were institutionalised.

Table 2 Evaluated patients in each participating hospital
Table 3 Diagnostic groups considered

The mean number of medicaments used was 9.46±5.80 (range 1–33) at evaluation; corticosteroid and beta-blocking agents were used in 113 (30.05%) and 14 (3.72%) evaluated patients, respectively.

Metabolic stress related to underlying disease was considered to be present to some degree in 195/360 (54.17%) patients, being moderate or high in 69/360 (19.17%).

Values of different anthropometrical and biochemical parameters evaluated are presented in Table 4. It is of note that we were able to register weight loss in the last 2 weeks in 111/358 (30.73%) patients.

Table 4 Anthropometric and biochemical data of patients evaluated

With respect to food ingestion of the evaluated patients, in 130/357 (36.41%) there was a decrease of oral ingestion within 2 weeks prior to evaluation, with only 208/358 (58.10%) patients having complete solid ingestion at the time of evaluation, 17/358 (4.75%) with enteral nutrition as the exclusive and total feeding source and 10/358 (2.80%) with total parenteral nutrition; 117/350 (32.50%) had no food ingestion for 3 days and 57/360 (16.29%) went for at least 5 days with an incomplete liquid diet. Gastrointestinal symptoms in this sample were absent in 225/359 (62.67%) patients, but 41/359 (11.42%) presented anorexia, 38/359 (10.58%) had vomiting, 28/359 (7.80%) nausea and 27/359 (7.52%) had diarrhoea at the evaluation.

Also, functional capacity has been considered very important in general evaluation and in nutritional risk; in Figure 1, we can observe that there is a significant decrease in functional capacity during the stay, with respect to previous functional capacity.

Figure 1

Functional capacity of inpatients evaluated (%).

Nutritional risk screening scale

From Cardona's scale, 177/359 (49.30%) patients were found to be nutritionally at risk.

Nutritional status diagnosis scale

According to SGA, 169/360 (46.94%) patients presented malnutrition (134 B category and 35°C category).

Validation of screening scale and identification of risk factors

Screening scale had a positive significant correlation with nutritional classification by SGA (P<0.0001).

With respect to malnutrition risk factors evaluated, we found a statistically significant association between presence of malnutrition and degree of metabolic stress caused by underlying disease (P=0.003) and also between age group (under 65 y of age or over) and presence of malnutrition by SGA (P=0.027), the older ones having higher rates of malnutrition to a mild degree. Malnourished patients had a significantly greater number of medicaments than nonmalnourished patients (11.90±6.53 vs 7.96±4.51, P=0.005). We could not find any association with the rest of the studied variables (sex, service, diagnosis, diabetes mellitus, days from admission, institutionalisation, risk medications, etc).

The distribution of diagnosis and malnutrition by SGA relative to age group is shown in Figure 2.

Figure 2

% Distribution of malnutrition by SGA (a) and diagnosis relative to age group (b) (under and over or equal to 65 y old).


Hospital malnutrition has been recognised as one of the more important factors implicated in morbimortality in affected patients. The presence or appearance of malnutrition can perpetuate the vicious circle morbidity–malnutrition–morbimortality and increase suffering and economical costs caused by acute diseases.

Nutritional diagnosis can be hindered by interferences with nutritional parameters caused by the underlying disease and other conditions (Alcok, 1999), and that is because there exist several methods of nutritional assessment based on history, anthropometry, physical examination, and values of biochemical and immune parameters. Various methods have demonstrated their validity and reproducibility in several populations.

Given the complexity of nutritional status diagnosis and the importance of its advancement, many studies have attempted to identify screening scales that simply allow one to diagnose in advance and operate on patients who are at risk, and are also easily applied and reproducible (Powewll-Tuck & Hennessy, 2003).

In our study, we found a prevalence of hospital malnutrition of about 50%. Taking into account that this is an adult population admitted because of acute diseases and an aged reference population, it is in keeping with what is expected. It is of note that the vast majority of patients considered to be malnourished have a mild affectation of visceral protein, probably reflecting the impact of underlying disease on protein-synthesis capacity, and also clearly reflecting the risk of nutritional complications and the need for a nutritional intervention in advance. Moreover, an important proportion of patients admitted because of acute illnesses or acute exacerbation of chronic diseases have gastrointestinal symptoms and several other circumstances that interfere with normal feeding, which maintained over time and associated with the disease itself, increases malnutrition risks, similar to those found in previous studies in several populations of hospitalised patients with acute illnesses (Kondrup et al, 2002, 2003; Kruizenga et al, 2003). The study of Kruizenga et al (2003) found about 25% Dutch patients to be malnourished, especially those with malignancies, greater unintentional weight loss and older age. In our study, in general, we found a greater prevalence of malnutrition than in previous studies, but as we stated before, the majority of our cases corresponded to a visceral protein affection, probably reflecting metabolic stress caused by acute disease. In accordance with previous studies (Kondrup et al, 2002; Kruizenga et al, 2003), these patients must be considered as nutritional risk subjects that must receive nutritional support if we want to avoid the consequences of malnutrition in hospitals.

On the other hand, our study is representative of the totality of adult patients admitted in the hospitals of our region; that permits us to ensure that the results of nutritional status are quite close to our reality.

Also, the results of screening scale evaluated in our study have an excellent correlation with nutritional diagnosis scales, especially Cardona's method, that includes clinical and analytical variables and takes into account underlying disease that causes admission, being applicable to the vast majority of studied patients. Elmore's equation, as others based in numerical data, is more limited, as an important percentage of patients is lacking some equation variable; thus its application is less suitable, and also its correlation with nutritional diagnosis is poorer than that of Cardona's method. With respect to nutritional assessment methods, both SGA and Chang's method present a recognized diagnostic precision when applied by skilled personnel, as in our case.

Of the risk factors identified, as in previous studies, the older age of patients may be of great importance, as is the degree of metabolic stress caused by underlying disease. The greater number of medications used in malnourished patients probably reflects the greater complexity of underlying disease; nevertheless, neither the diagnostic group nor the presence of ‘risk’ medications was related to malnutrition. The low number of institutionalised patients in our study is due to the conditions of our hospitals (acute care hospitals) and we could not find it to be related to malnutrition. We did not observe other relationships, because in this study we were not evaluating other aspects like admission costs, complexity or stays generated.

In conclusion, in this multicentric study representative of the Autonomic Community of Galicia (Spain), we found malnutrition prevalence to be about 50% in the adult population hospitalised because of acute diseases, the majority with mild protein malnutrition, the principal risk factors being older age and degree of metabolic stress caused by underlying disease. We applied a nutritional risk screening scale, the results of which correlated with diagnosis method, and that can be useful in detecting and in treating in advance patients at risk of malnutrition. The introduction and establishment of protocol for these advance diagnosis mechanisms and the development of multidisciplinary Clinical Nutrition and Dietetics Units in hospitals will allow us to overcome this important challenge for Health Systems in our setting in the 21st Century.


  1. Alastrué A, Rull M, Campos I, Ginesta C, Mellus MR & Salvá JA (1988): Nuevas normas y consejos en la valoración de los parámetros antropométricos en nuestra población: índice adiposo-muscular, índices ponderales y tablas de percentiles de los datos antropométricos útiles en una valoración nutricional. Med. Clin. (Barc.) 91, 223–236.

  2. Alcok NW (1999): Laboratory tests for assessing nutritional status. In Modern Nutrition in Health and Disease 9th Edition, eds Shils ME, Olson JA, Shike M, Ross AC, pp 923–936. Baltimore: Williams and Wilkins.

  3. Aznarte Padial P, Pareja Rodríguez de Vera A, de la Rubia Nieto A, López Soriano F & Martínez de Guzmán M (2001): Influence of hospitalization on patients following nutritional assessment on admission. Nutr. Hosp. 16, 14–18.

  4. Cardona D (1998): La nutrición artificial y la mejora de la calidad asistencial. Rev. Calidad Assist. 13, 120–135.

  5. Detsky AS, McLaghlin JR, Baker JP, Johnston N, Whittaker S, Mendelson RA & Jeejeebhoy KN (1987): What is Subjective Global Assessment of nutritional status? J. Parenter. Enter. Nutr. 11, 8–13.

  6. de Ulíbarri Pérez JI, Picón César MJ, García Benavent E & Mancha Álvarez-Estrada A (2002): Early detection and control of undernourishment in hospitals. Nutr. Hosp. 17, 139–146.

  7. Nagel MR (1991): Nutrition screening: identifying patients at risk for malnutrition. Nutr. Clin. Pract. 6, 151:155.

  8. Esquius M, Schwartz S, López Helín J, Andreu AL & García E (1993): Parámetros antropométricos de referencia en la población anciana. Med. Clin. (Barc.) 100, 692–698.

  9. González Sánchez J, Hernando Cuñado M, Kazemzadeh A, Luna García R, Martínez-Cañavate Montero MJ, Luna García R & Ramos Martínez A (2000): Malnutrición en un servicio de Medicina Interna: influencia de los pacientes procedentes de residencias de ancianos. Ann. Med. Intern. (Madrid) 17, 347–350.

  10. Guigoz Y, Vellas B & Garry PJ (1996): Assessing the nutritional status of the elderly: the Mini Nutritional Assessment as part of the geriatric evaluation. Nutr. Rev. 54, S59–S65.

  11. Kondrup J, Johansen N, Plum LM, Bak L, Hjlund Larsen I, Martinsen A, Andersen JR, Brnthsen H, Bunch E & Lauesen N (2002): Incidence of nutritional risk and causes of inadequate nutritional care in hospitals. Clin. Nutr. 21, 461–468.

  12. Kondrup J, Allison SP, Elia M, Vellas B & Plauth M (2003): ESPEN Guidelines for Nutrition Screening 2002. Clin. Nutr. 22, 415–421.

  13. Kruizenga HM, Wierdsma NJ, Van Bokhorst MAE, Van Der Schueren DE, Hollander HJ, Jonkers-schuitema CF, Van der Heijden E, Melis GC & Van Staveren WA (2003): Screening of nutritional status in The Netherlands. Clin. Nutr. 22, 147–152.

  14. Martínez Olmos MA, Martínez Vázquez MJ, López Sierra A, Morales Gorría MJ, Cal Bouzón S, Castro Núñez I, Del Campo V & Pena González E (2002): Detection of the risk of malnutrition in hospitalized elderly patients. Nutr. Hosp. 17, 22–27.

  15. McWhirter JP & Pennington CR (1994): Incidence and recognition of malnutrition in hospital. BMJ 308, 945–948.

  16. Powewll-Tuck J & Hennessy EM (2003): A comparison of mid upper arm circumference, body mass index and weight loss as indices of undernutrition in acutely hospitalized patients. Clin. Nutr. 22, 307–312.

  17. Roldán JP, Pérez I, Irles JA & Martín R (1985): Malnutrición en pacientes hospitalizados: estudio prospectivo y aleatorio. Nutr. Hosp. 10, 192–198.

  18. The malnutrition prevalence group (2000): Prevalence of malnutrition on admission to four hospitals in England. Clin. Nutr. 19, 191–195.

  19. Thorsdottir I, Eriksen B & Eysteinsdottir S (1999): Nutritional status at submission for dietetic services and screening for malnutrition at admission to hospital. Clin. Nutr. 18, 15–21.

Download references


This study has been conducted by means of a Grant of the Nutrition and Dietetics Society of Galicia (Beca SONUDIGA 2003). We want to acknowledge all participants in the study for their kind efforts in performing this study.

Author information

Correspondence to M A Martínez Olmos.

Additional information

Guarantor: Nutrition and Dietetic Society of Galicia.

Contributors: Nutrition and Dietetic Society of Galicia.


Appendix A1

Appendix A2

Rights and permissions

Reprints and Permissions

About this article


  • nutritional status diagnosis
  • malnutrition
  • risk
  • screening
  • adult patients
  • hospital setting

Further reading