Introduction

Malnutrition is the most predominant and frequent condition among older people due to physical, physiological and psychological alteration which may all impact nutritional intake and clinical outcome (1-3). Results of population studies have reported that 23-62% of older hospitalized patients suffer from malnutrition (4, 5). Since malnutrition in older persons is generally multifactorial in origin, the causes of malnutrition are extremely varied including dental problems, chewing and swallowing disorders, medication side effects, general diseases, depression and dementia (6, 7).

With advancing age, there is also loss of appetite which reduces food intake and leads to unintentional weight loss and malnutrition development (8). Some clinical observations have demonstrated that an acute inflammatory process may play a role in inhibiting appetite and food intake by interacting in several central nervous system pathways particularly in the hypothalamus (9, 10). Indeed, acute disease links to tissue inflammation resulting in inflammatory responses which may alter the feeding behavior and negatively affect the health related quality of life in older patients (6, 11, 12).

C-reactive protein (CRP) is the most widely used inflammatory marker which is elevated with both acute and chronic inflammation (13, 14). In addition, since higher CRP level may be associated with reduction of food intake and dietary behavior (15), it seems to be good candidate as an appetite indicator. Apart from a few animal-experimental studies which have focused on food intake, not appetite, there is a lack of knowledge on the association of inflammation with appetite and food intake in humans.

Although, the presence of chronic or acute inflammation in renal failure patients (16, 17) and in cancer patients (18) have been associated with reduction of food intake and poor appetite, such association in older hospitalized persons has not been investigated so far. Indeed, there are no data available to clarify to what degree a reduction of food intake is a consequence of inflammation. Therefore, the purpose of the present study was to investigate the association of CRP, as an inflammatory marker, with food intake in acutely ill older hospitalized patients.

Subjects and methods

The study population consisted of 177 older participants (116 females), aged 64-100 y with a body mass index (BMI) range of 14.7–43.6 kg/m2 who consecutively hospitalized between October 2015 and February 2016 to a geriatric acute care ward at St. Marien-Hospital, Borken, Germany. Exclusion criteria were age < 60 years and missing or withdrawn consent of the patients. The study protocol had been approved by the ethical committee of Friedrich-Alexander-University, Erlangen-Nürnberg.

Table 1
figure 1

Characteristics of the study population stratified according to food intake during previous week

Full size image

Assessment of nutritional status and nutritional intake

Nutritional Risk Screening (NRS-2002) (19) and Mini Nutritional Assessment Short Form (MNA®-SF) (20) were performed within two days after hospital admission. Screening of the NRS-2002 comprises of two criteria: impaired nutritional status based on weight loss, BMI and food intake during last week (1-3 points) as well as severity of disease (1-3 points). Accordingly, patients were grouped as no risk (<3 points) or at risk (3 points). In addition, for older individuals ≥70 years, one point to the total score was added. Based on the food intake item, patients were grouped into two categories as intake ≥75% and <75% of requirements.

Screening of MNA-SF was based on reduction in food intake, weight loss during previous 3 months, mobility, psychological stress and acute diseases, neuropsychological problems (i.e. dementia and depression) and BMI. Accordingly, older patients were classified as having normal nutritional status (12-14 points), at risk of malnutrition (8-11 points) and malnourished (0-7 points). In addition, based on reduction of food intake in the past 3 months due to loss of appetite, subjects were classified into three groups as no, mild and severe reduction in food intake.

Anthropometric measurements

Weight was assessed in light clothing with an accuracy of 0.1 kg, and height was measured to the nearest 0.5 cm with a stadiometer in first day after admission to hospital. The magnitude of unintentional weight loss was obtained either by interviewing the patients and their relatives, if competent, or asking their proxy, where necessary.

Measurement of C-reactive protein

CRP was analyzed according to standard procedures directly at hospital admission or the day after. The levels between 0.0–0.5 (mg/dl) indicated no inflammation (= normal values) whereas levels between 0.5–3.0 (mg/dl) and >3.0 (mg/dl) considered as mild inflammation and moderate to severe inflammation, respectively.

Data analysis

The statistical analysis was completed with SPSS statistical software (SPSS Statistics for Windows, IBM Corp, Version 23.0, Armonk, NY, USA). Continuous variables are expressed by their means and standard deviations (SDs) for normally distributed variables and median values with interquartile ranges (IQR) for non-normally distributed data. Categorical variables are expressed as absolute numbers and relative frequencies (%). Differences between patient’s variables in intake ≥75% and <75% requirements were analyzed by using an unpaired t-test in normally distributed data, the Mann-Whitney U test for continuous variables with non-normal distribution and Chi-square test for categorical variables. Comparison of CRP levels between no, mild and severe reduction of food intake in the past 3 months was performed by using one-way ANOVA post hoc test. A logistic regression analysis was performed to examine the impact of potential influencing factors (i.e. CRP levels, disease severity, mobility, neuropsychological problems, BMI, gender and age as independent variables) on low food intake during previous week (as dependent variable). A P value <0.05 was accepted as the limit of significance.

Results

Baseline characteristics and CRP level of study participants stratified according to food intake during previous week are summarized in Table 1. The recruited population consisted of 177 patients with a mean age of 83.1 ± 6.5 y, predominantly females (66.0%). Of total population, 67 (38.0%) had moderate to severe inflammation (CRP>3.0 mg/dl). According to MNA-SF, 46.0% of patients were malnourished, 13.0% at risk of malnutrition and 52.0% had no mobility. In addition, 28.2% and 17.5% of the study participants have displayed mild and severe neuropsychological deficits, respectively. According to NRS-2002, 20.0% and 2.0% of patients reported moderate and severe disease severity.

Food intake during previous week

Of the 177 patients, 109 (62.0%) had intake <75% of requirements, in which, 50.0% demonstrated moderate to severe inflammation whereas 31.0% and 19.0% had mild and no inflammation, respectively. Compared to the patients with intake ≥75% of requirements, subjects with intake <75% of requirements were more malnourished and at risk of malnutrition (P=0.003) according to MNA-SF, displayed higher weight loss in the past 6 months, mobility limitation and moderate to severe disease as well as a lower actual body weight and BMI. In addition, there were significant differences in CRP levels between both groups with higher value in patients as categorized with intake <75% of requirements (P<0.001; Table 1).

Food intake in the past 3 months

Of the 177 patients, 92 (52.0%) had no reduction of food intake whereas 59 (33.0%) and 26 (15.0%) reported mild and severe reduction of food intake. In addition, CRP levels were not statistically significant between no and mild (P=0.165), between no and severe (P=0.978) and between mild and severe (P=0.513) reduction of food intake during the past three months (data not shown).

To test the association of CRP levels with nutritional intake during previous week (dependent variable), we performed a series of logistic regression analyses (Table 2). In a first analysis, we included only CRP levels as independent variable. Significant associations between inflammation and food intake were observed and CRP levels explained 11.3% of the variance in food intake during previous week. When age, gender, BMI, mobility, disease severity and neuropsychological problems were also added in a second analysis, an additional increase in the proportion of the explained variance was observed (21.3%). Every one mg/dl increase of CRP-level is associated with an increase in the odds of food intake <75% of demands by 15%. CRP levels, disease severity, mobility and BMI were the major independent predictors for food intake. Although, the association between CRP and food intake remained significant following second analysis, the effect of disease severity on food intake during previous week was markedly higher compared to increased CRP levels (Table 2).

Table 2
figure 2

Results of logistics regression analysis with low food intake (<75% of requirements) during previous week as dependent variable in total study population (n=177)

Full size image

Discussion

An important finding of this study is a close association between food intake and inflammation in older hospitalized patients. CRP is the most widely measured inflammatory marker which is associated with reduction of food intake and dietary behavior (13). A diminished appetite and food intake which were associated with increased serum levels of CRP and other inflammatory markers have been previously reported in hemodialysis patients (21, 22). However, in a randomized, double-blind nutritional intervention study, Gariballa et al. (23) examined the impact of CRP concentration, as a marker of acute-phase response, on nutritional status and clinical outcome in hospitalized older patients. In accordance with our findings, this study demonstrated that those patients with elevated CRP levels had significantly lower energy intake and lower body weight whereas in a multivariate analysis, CRP was a significant and independent predictor on nutritional status and clinical outcome (23). However, in the present study, we examined very old and acutely unwell patients who admitted to a geriatric acute care ward whereas Gariballa et al. analyzed a mixed hospital population and did not control for severity of the disease which could have biased the effect of inflammation.

The current study shows that serum concentration of CRP was higher in patients who had intake <75% of requirements during previous week. However, we found that food intake in the past 3 months was not statistically associated with inflammation, presumably because we predominantly analyzed acutely ill patients. In addition, in agreement with previous studies, our findings demonstrate that the reduction of food intake is accompanied with deterioration in nutritional status (21, 23) and poorer anthropometric values (21, 24). Some evidence has indicated the association between inflammatory makers i.e. CRP and muscle catabolism and protein synthesis reduction (16, 25). Therefore, we suggest that in our older patients as categorized with intake <75% of requirements, inflammation is linked to higher weight loss and a worse nutritional status. These findings may be explained by the effectiveness of pro-inflammatory cytokines on satiety center inducing reduction of appetite and development of muscle catabolism (16, 22, 25) which have been previously confirmed in both renal failure patients (21, 22, 26) and animal studies (27-29). Nevertheless, it is impossible to prove causality with a cross sectional study, but this causality has been proven in longitudinal animal studies (28).

Another important finding is that serum CRP concentration and disease severity were the most important independent factors associated with acute reduction of food intake whereas the association with disease severity was markedly higher compared to increased CRP levels. This is not surprising since geriatric patients mostly suffer from multiple comorbidities (6, 30) that reduce appetite, consequently contributing in reduction of food intake and progression of malnutrition in these patients (23).

Some limitations of the current study should be mentioned. First, food intakes during previous week and in the past 3 months were estimated according to the respective NRS-2002 and MNA-SF items, which may be imprecise. Second, we used CRP alone to measure inflammation and we did not consider other inflammatory markers which may demonstrate different or even stronger impact on food intake. Third, we did not analyze the origin of inflammation, which would be interesting to differentiate. Fourth, food intake is dependent on many variables which were only partially accounted in the current study, since we were particularly interested on the association of food intake and inflammation. Therefore, residual, uncontrolled confounding cannot be excluded. Finally, self-reported weight loss, the relatively small sample size and the cross-sectional design may have limited the value of our results. Further studies should differentiate acute and chronic inflammation as well as the origin of inflammation and have longitudinal design to confirm causality.

Conclusion

Our findings confirm a close association between inflammation and food intake during previous week in older hospitalized patients. In addition, CRP and disease severity together were the most important independent predictors associated with food intake during previous week among these patients.

Conflicts of interest: The authors declare no conflict of interest.

Financial disclosure: The study received no financial support.

Ethical standard: The authors declare that the study procedures comply with current ethical standards for research involving human participants in Germany. The study protocol had been approved by the ethical committee of Friedrich-Alexander-University, Erlangen-Nürnberg.

Statement of authorship: The study was designed by all authors. Data were obtained by RW. Statistical analysis was performed by MP. MP, SB, GJ, LS and RW prepared the manuscript. All authors read and approved the final manuscript.