Frequency of malnutrition and associated risk factors in geriatric patients in a palliative care unit

INTRODUCTION

Technological advances and the development of new treatment methods result in increased life expectancy. As the geriatric population grows, there is also an increase in chronic diseases, associated symptoms, and the need for palliative care for older adults. Palliative care aims to provide symptom relief and improve quality of life through the collaboration of a multidisciplinary team. Adequate and balanced nutrition is one of the parameters that increase the quality of life [1]. Inadequate or unbalanced nutrition leads to malnutrition. The prevalence of malnutrition is high in the older population due to cognitive impairment, comorbid diseases, polypharmacy, depression, and senile anorexia [2]. Malnutrition increases morbidity and mortality [3], disrupts physical, mental, and social function, and leads to increased health expenditures [4, 5]. In addition, the presence of malnutrition is associated with other geriatric syndromes such as frailty and sarcopenia [6]. Its frequency among older inpatients varies between 11 and 45% [7–9].
All patients hospitalized in our country are evaluated using a nutritional screening instrument according to the European Society of Clinical Nutrition and Metabolism guidelines within the first 24 hours of admission [10]. Although many instruments are used in practice, there is controversy over which should be considered the gold standard [11]. The mini nutritional assessment (MNA) has low specificity but a high diagnostic rate for malnutrition [12]. However, disadvantages of the MNA are that the test is long and cannot be applied in people with dementia or communication problems [9]. The geriatric nutritional risk index (GNRI) was developed especially for older adults and is used to diagnose malnutrition and predict malnutrition-related complications [13]. It has also been demonstrated to be reliable in patients with dementia, aphasia, and apraxia [4].
In light of this information, our study aimed to determine the frequency of malnutrition and associated risk factors in older patients receiving palliative care.

MATERIAL AND METHODS

This prospective cross-sectional study was performed with older patients hospitalized and treated in the palliative care unit of our university between December 1, 2022 and June 1, 2023.
Inclusion criteria were being 65 years of age or older, being hospitalized in the palliative care unit for 24 hours or more, and providing written informed consent to participate in the study. Exclusion criteria were age less than 65 years, hospitalization in the palliative care unit for less than 24 hours, repeated admission to the palliative care unit, and not providing informed consent.
The demographic characteristics of the patients (age, gender, chronic diseases, and drugs used), reasons for palliative care admission, and the referring unit were recorded.
The patients’ nutritional status was assessed using the full MNA and the GNRI. The full MNA test consists of 18 items:

• anthropometric measurements (4 questions about height, weight, body mass index, upper arm circumference, calf diameter),
• dietary habits (6 questions about number of meals, food intake, fluid intake, whether eating assistance is needed),
• general assessment (6 questions about level of independence, drugs used, mobility, mental status, skin changes, and acute stress in the last 3 months),
• subjective assessment of oral intake (2 questions about self-perceptions about health and nutrition).

The patients’ nutritional status was determined according to their score on the full MNA. The total score ranges from 0 to 30, with scores of 24 or higher interpreted as normal nutritional status (well-nourished), scores of 17–23.5 as risk of malnutrition, and scores lower than 17 as malnutrition [14]. The validity and reliability of the Turkish version of the instrument were analyzed by Sarıkaya et al. in 2015 [15].
The geriatric nutritional risk index was first used in 2005 to evaluate the nutritional status of older immobile patients. This scoring system utilizes the patient’s ideal body weight, calculated using the Lorentz formula:
For men: height – 100 (height – 150)/4; for women: height – 100 (height – 150/2.5).
The geriatric nutritional risk index is calculated using this value and the patient’s height, weight, and serum albumin values as follows:
GNRI = 1.489 × albumin (g/l) + 41.7 × (actual weight/ideal body weight)
The geriatric nutritional risk index < 82 is interpreted as severe malnutrition, 82–92 as moderate malnutrition, 92–98 as malnutrition, and GNRI > 98 as no risk of malnutrition [13].
Selected laboratory parameters were recorded at admission to the palliative care unit: blood urea nitrogen (BUN), creatinine, sodium (Na), potassium (K), calcium (Ca), phosphorus (P), chlorine (Cl), magnesium (Mg), uric acid, white blood cells, neutrophils, lymphocyte and platelet counts, mean platelet volume, hemoglobin (Hb), hematocrit, sedimentation, procalcitonin, C-reactive protein (CRP), albumin, prealbumin, and glucose.

Statistical analysis

All analyses were performed using IBM SPSS Statistics version 20.0 (IBM; http://www.spss.com). Categorical variables were presented as frequency and percentage, and numerical variables as mean ± standard error or median and range or interquartile range. The chi-square test was used to compare categorical data, and the Mann-Whitney U test was used to compare continuous variables. A logistic regression model was created for risk analysis. In all tests, p < 0.05 was considered significant.
Approval to conduct the study was obtained from the Ethics Committee of Atatürk University Faculty of Health Sciences (date: 08.11.2022, meeting no: 2022/11/26).

RESULTS

The 371 patients included in the study had a median age of 74 years (range, 65–99 years) and 217 (58.5%) were female. Comparisons of the patients’ basic characteristics according to the presence of malnutrition are shown in Table 1. Malnutrition was detected more frequently in geriatric palliative care patients with cerebrovascular disease (CVD), malignancy, dementia, and congestive heart failure (CHF) (p < 0.05) (Table 1). In addition, patients with malnutrition were significantly older than those without malnutrition, and malnutrition was more common among women (p < 0.001 and p = 0.010, respectively).
The distribution of GNRI values in geriatric patients with and without malnutrition according to the full MNA is presented in Figure 1. Geriatric nutritional risk index values were significantly lower in those with malnutrition (p = 0.003).
Comparisons of laboratory values at admission according to the presence of malnutrition in geriatric palliative care patients is presented in Table 2. Those with malnutrition had significantly lower albumin, prealbumin, Hb, lymphocyte count, uric acid, sodium, calcium, magnesium, chlorine, and potassium values compared to those without malnutrition (p < 0.05). In contrast, BUN, leukocyte, and CRP levels were found to be significantly higher in patients with malnutrition (p < 0.05) (Table 2).
The diagnostic value, specificity, and sensitivity of laboratory parameters at admission to the determined cut-off values are shown in Table 3. Of the evaluated biomarkers, prealbumin and Hb had the highest dia­gnostic value. Prealbumin had the highest specificity (92.3%) at a cut-off value of 0.175 mg/dl.
Comorbidities and admitting laboratory parameters associated with malnutrition risk in geriatric palliative care patients were evaluated by univariate and multivariate logistic regression analysis (Table 4). The results of univariate logistic regression analysis indicated that age, female sex, CVD, malignancy, dementia, CHF, number of drugs used, BUN, K, Ca, Cl, albumin, prealbumin, and Hb were factors statistically significantly associated with malnutrition (Table 3) (p < 0.05). The multivariate logistic regression model created with these variables showed that female sex, Ca, Cl, prealbumin, and Hb were significant independent risk factors for malnutrition (p < 0.05). Female sex was associated with 2.5 times higher odds of malnutrition. In addition, a 1-unit increase in Ca level reduced the odds of malnutrition by 0.515 fold, a 1-unit increase in Cl level by 0.924 fold, and a 1-unit increase in Hb level by 0.657 fold, while a 1-unit increase in prealbumin level reduced the odds of malnutrition (Table 4).

Discussion

As in the rest of the world, the need for palliative care centers is increasing in our country. However, the numbers of palliative care centers and studies on this subject are still lower than expected. Therefore, in our study we aimed to identify factors associated with malnutrition in geriatric palliative care patients. Malnutrition was detected in a majority of the geriatric patients followed in our palliative care unit. Female gender, Hb, and especially prealbumin emerged as important determinants of malnutrition.
Malnutrition has physiological, social, and economic consequences [16]. Previous studies have shown that one in four older patients and half of inpatients have malnutrition [2]. A study of older people in Denmark revealed a relationship between malnutrition and the frequency of hospitalization [17]. Other clinical consequences of malnutrition include nosocomial pneumonia [18] and an increase in the number of falls and fractures [19]. At the same time, many diseases are exacerbated by malnutrition [20, 21]. This leads to a vicious cycle in malnutrition. For these reasons, early detection of malnutrition is extremely important. However, studies have shown that malnutrition is underdiagnosed and undertreated [22].
In our study, the prevalence of malnutrition according to the GNRI was found to be 64.8% among 371 geriatric palliative care patients. In a recent study including 319 patients in our country, patients receiving palliative care were screened according to the nutrition risk screening 2002 and malnutrition was detected in 58.6% of the patients [23]. Another study conducted among 136 geriatric palliative care patients in our country showed that 39.7% of the patients had malnutrition and 47.1% were at risk of malnutrition [24]. In studies conducted among patients receiving long-term care, this rate varied widely between 12.5% and 78.9% [25, 26]. Most patients in our study were immobile. Like other geriatric syndromes, malnutrition is multifactorial and overlaps with many factors. Age-related changes in the musculoskeletal and neurological systems lead to sarcopenia, which also increases the frequency of malnutrition [27]. The different results obtained in our study compared to the literature may be related to methodological differences, concomitant diseases, and the inclusion of only older patients.
Serum albumin is a negative acute phase reactant often used as an inflammation marker. In fact, albumin level decreases secondary to acute and chronic inflammation due to reduced production in the liver, leading to the formation of capillary links [28, 29]. Therefore, it is not a good indicator in the assessment of nutritional status, especially in older inpatients [30]. There is little evidence in the literature that serum albumin level is a marker of malnutrition. In a meta-analysis of 63 studies including a total of 2125 healthy individuals, it was found that fasting had no effect on serum albumin levels [31]. Similarly, in a study involving 14 people diagnosed with anorexia nervosa and 15 healthy individuals, patients were followed for one year and no difference in serum albumin levels was observed between the control and patient groups [32]. Another study showed that a serum albumin level of < 3.5 g/dl had low specificity in the detection of malnutrition in functionally dependent older patients [33], meaning that assessment based on this parameter would lead to false positives in these individuals. In addition, it has been reported that change in body posture (supine vs. sitting after moderate exercise) significantly altered serum albumin levels [34].
Like albumin, prealbumin is a negative acute phase reactant synthesized by the liver. As a result, its level is altered in conditions such as infection and liver disease. However, prealbumin has a half-life of 2–3 days, which is much shorter than that of albumin, so it is more frequently used as an indicator of malnutrition. Because prealbumin is excreted by the kidneys, however, its concentration will increase in renal failure. Moreover, as one of the functions of prealbumin is binding to thyroxine, levels are higher in patients with hypothyroidism [35]. In our study, prealbumin was the most sensitive biomarker at a cut-off value of 0.175 mg/dl.
Anemia is prevalent in older adults, and malnutrition and the risk of malnutrition are also associated with a higher frequency of anemia [36]. Consistent with the literature, our study demonstrated that the presence of anemia increased the risk of malnutrition.
In our study, the frequency of malnutrition was 2.5 times higher in women than in men. Other studies have also shown that women more frequently experience anorexia and are more prone to malnutrition [37]. A recent study conducted in Saudi Arabia also showed that malnutrition was more frequent in women [38]. However, other studies have indicated the opposite. In a study conducted by Méndez et al. [39], the rate of malnutrition was found to be higher in men. These differences in results may be attributable to geographic regional and cultural differences.
The strength of our study is that it was carried out prospectively and used validated nutrition assessment tools. However, it also has some limitations. The first of these is that the study was conducted in a single center. Another is that repeated measures were not performed to evaluate changes in the laboratory parameters evaluated at admission. Finally, the study did not examine demographic and social risk factors that can have an impact on malnutrition.

CONCLUSIONS

The prevalence of malnutrition was found to be high among geriatric palliative care patients in our study. Screening for and early detection of malnutrition in geriatric patients receiving palliative care will be beneficial. The geriatric nutritional risk index can be used to diagnose malnutrition in immobile older palliative care patients and may be applied in routine patient follow-up to help improve quality of life.
The authors declare no conflict of interest.

References