Comparison of Body Mass Index Range Criteria and Their Association with Nutrition-Related Risk: A Cross-Sectional Study in Polish Older Adults

Introduction

Malnutrition is a common issue among older adults and significantly affects health systems, social services, and aged care. This population is particularly susceptible to malnutrition due to age-related physiological changes, limited access to nutritious food, and the presence of multiple health conditions. Moreover, reduced appetite, use of multiple medications, dementia, frailty, dental problems, difficulty in swallowing, social isolation, and poverty increase the risk of malnutrition. Clinical guidelines emphasize the importance of routinely screening all older adults for malnutrition through nutritional assessments. They recommend providing nutritional intervention support for those identified as being at risk [1–4].

The findings of a systematic review indicate that up to 23% of older adults in Europe face a high risk of malnutrition, while an additional 48.4% are at some level of malnutrition risk, encompassing both moderate and high risk categories combined. In residential care settings outside the European Union, the reported prevalence of high malnutrition risk varies widely – from 31% to 70% – depending on the screening tool used and the level of care provided [3]. The prevalence of high malnutrition risk also differs across various settings among older adults in Europe, with rates reaching 28.0% in hospitals, 17.5% in residential care facilities, and 8.5% in community settings [1].

Among the Polish older population, based on the PolSenior 2 study, 27.8% of participants were found to have malnutrition according to the Mini Nutritional Assessment-Short Form. Simultaneously, using the Global Leadership Initiative on Malnutrition (GLIM) criteria, malnutrition was diagnosed in 13.4% of participants [5].

The prevalence of high malnutrition risk among European older adults, as assessed by the Geriatric Nutrition Risk Index (GNRI), was 19%. When both the moderate and high risk categories are included according to GNRI, the prevalence increases to 39.7% [3].

It is recommended to conduct screening as the initial step before malnutrition diagnosis, with the aim of identifying individuals at risk. The screening process should be standardized, suitable for large groups, quick, simple, and practical, with high accuracy, and the screening criteria should be easily accessible [3]. Among all the malnutrition screening tools available worldwide, only 34 have been validated for use with older adults, and of these, only 22 have shown acceptable validity for this age group in the particular setting where they were tested [3]. Acceptable validity screening tools for older adults include, for example, Nutritional Risk Screening 2002, Malnutrition Universal Screening Tool, Mini Nutritional Assessment, Mini Nutritional Assessment-Short Form, Malnutrition Screening Tool, and Subjective Global Assessment [1,3]. The GNRI has also been validated as a screening tool for older adults in settings such as the community, hospital, and residential care [3]. Anthropometric measures recommended for older adults in all settings include body mass index (BMI), weight loss, calf circumference, and mid-arm circumference [1].

The GNRI is a nutrition-related risk index that helps classify elderly patients at risk of malnutrition-related morbidity and mortality. It is a more reliable prognostic tool for older adults than are indices that rely solely on albumin levels or BMI [6]. To calculate the GNRI, serum albumin levels, actual body weight, and ideal body weight (determined using the Lorentz formula) are used [2,6]. The GNRI is a validated malnutrition screening tool for this age group [3].

BMI is a widely used measure for assessing an individual’s body weight in relation to their height and is commonly used to evaluate the nutritional status of populations. The simplicity of its calculation and interpretation makes it a popular tool for guiding public policies and interventions. For over 30 years, the World Health Organization (WHO) has provided BMI cut-off points for adults aged 25 years and older, which have been used in most studies involving adults. According to the WHO classification, adults with a BMI of less than 18.5 kg/m2 are considered as having underweight, those with a BMI between 18.5 kg/m2 and less than 25 kg/m2 are classified as having normal weight, those with a BMI between 25 kg/m2 and less than 30 kg/m2 are classified as having overweight, and those with a BMI of 30 kg/m2 or above are classified as having obesity [7–11]. However, despite its widespread use as a standard health predictor, the WHO BMI thresholds have limitations when applied to adults aged 60 and above [7,8]. This is caused by several factors due to aging, as when body composition changes occur, including an increase and redistribution of fat, along with a reduction in total and lean body mass. Over time, muscle mass diminishes and is gradually replaced by fat. Additionally, the distribution of fat shifts, with more fat accumulating around the abdomen, resulting in significant metabolic issues. Height loss caused by thinning vertebrae, compression of the spinal discs, development of kyphosis, or the effects of osteoporosis is also a natural consequence of aging. These changes in body composition can result in an increase in BMI by 1.5 to 2.5 kg/m2 in men and women, even if overall body weight remains the same [7,8,12].

In 1989, experts from the United States National Research Council Committee on Diet and Health (CDH) proposed the use of age-related BMI criteria. For people over 65 years, BMI below the norm is defined as a BMI <24 kg/m2, normal weight as 24 to 29 kg/m2, and BMI above the norm is defined as a BMI >29 kg/m2 [8,12,13].

Since the BMI remains a fundamental indicator for monitoring public health and screening, and considering its limitations in the elderly population, it is crucial to conduct research promoting the use of appropriate BMI interpretation criteria for older adults. Currently, most studies involving the elderly focus on the WHO criteria for BMI. Only a few address BMI classifications specifically designed for seniors, such as the CDH criteria. There are no studies assessing malnutrition risk based on GNRI while considering both the WHO and CDH BMI criteria. This represents a gap in the research.

The aim of the present study was to compare the distribution of BMI according to the ranges proposed by the WHO and CDH, and to examine their correlation with nutrition-related risk as assessed by the GNRI in older adults.

Material and Methods

ETHICS APPROVAL:

Before participating, all respondents were fully informed about the study’s purpose and procedures, and the voluntary nature of their involvement. Participants were given the freedom to withdraw from the study at any point, without the need to provide a reason or face any negative consequences. Each study participant gave their consent to participate in the study. The study sheet, containing data from interviews, anthropometric measurements, and biochemical tests, was coded to ensure the anonymity of the participants.

The research was carried out following the principles outlined in the Declaration of Helsinki. Approval for the study was granted by the Bioethics Committee at the Medical University of Silesia in Katowice, Poland (KNW/0022/KB1/53/14).

STUDY PARTICIPANTS:

This cross-sectional study involved 185 patients over 60 years of age hospitalized in the geriatric ward. The inclusion criteria were as follows: informed consent for participation in the study, absence of conditions leading to immobilization (eg, limb fractures, patients unable to get out of bed, patients using a wheelchair). The exclusion criteria were as follows: severe hepatic conditions (eg, liver cirrhosis), chronic kidney disease stage 4 or higher (glomerular filtration rate [eGFR] <30 mL/min/1.73 m²), active cancer (under treatment or recently diagnosed), clinical signs of fluid retention at the time of examination, mobility impairments, and lack of informed consent. All patients meeting the inclusion criteria and hospitalized during the study period were included in the study.

LABORATORY MEASURES:

An interview was conducted with each patient, during which information was collected, including sex, age, medical history, dietary supplements taken, and diet followed. At this stage, anthropometric measurements were also taken. Biochemical tests were performed after an overnight fast on the first day of hospitalization. Laboratory measurements were for C-reactive protein, eGFR (mL/min/1.73 m²), serum creatinine (μmol/L), aspartate aminotransferase (U/L), alanine aminotransferase (U/L), gamma-glutamyl transferase (U/L), albumin (g/L), and thyroid-stimulating hormone (uIU/mL).

ANTHROPOMETRIC MEASURES:

Weight and height measurements were conducted in the morning on the first day of admission for all patients, who were fasting, without shoes, and dressed in light outer clothing. Standardized methods and a validated scale were used for these evaluations. Body weight was measured with the Tanita BC 420 S MA scale (certified under MDD 93/42 EEC), with a precision of 0.01 kg. Height was assessed using the Tanita HR 100 stadiometer, with a precision of 0.05 cm.

BMI CALCULATION:

BMI was calculated using the following formula: weight [kg]/(height [m])2 [11].

NUTRITION-RELATED RISK INDEX CALCULATION:

The GNRI [6] was calculated to assess the risk of mortality and morbidity related to malnutrition.

The GNRI is calculated using the following formula: GNRI=[1.489×albumin (g/L)]+[41.7×(weight (kg)/ideal weight (kg)].

To determine the ideal weight, the Lorenz equations were used: For men: ideal weight (kg)=height (cm)-100-[height (cm)-150/4] For women: ideal weight (kg)=height (cm)-100-[height (cm)-150/2.5].

The study population was divided into 2 groups. The first group, named the “no nutrition-related risk group”, consisted of patients with a GNRI index greater than 98.0. The second group, named the “nutrition-related risk group”, consisted of patients with a GNRI index of 98.0 or less. More details are provided in Table 1.

STATISTICAL ANALYSIS:

Data analysis was performed using SAS software, version 9.4 (SAS Institute Inc., Cary, NC, USA). To assess the normality of the data distribution, the Shapiro-Wilk test was used. Data are presented as mean±SD for normally distributed variables and as median with interquartile range for non-parametric data. For comparisons of normally distributed data, the t test was used, whereas the Mann-Whitney U test was used for non-parametric data. The chi-square test was used for categorical variables, and Fisher exact test was used for categories with small expected frequencies. Additionally, the gamma and the Kendall Tau-b correlation coefficient were calculated to evaluate the strength and direction of association between ordinal variables. There were no missing data during the analyses; therefore, no imputation or special handling of missing values was necessary. A significance level of P<0.05 was deemed statistically significant.

Results

CHARACTERISTICS OF STUDY POPULATION:

The final group included 185 participants, 129 (69.7%) women and 56 (30.3%) men. In the no nutrition-related risk group there were 95 women (69.3%), and in the nutrition-related risk group there were 34 women (70.8%). The difference in the proportion of sex between these groups was not statistically significant (P=0.8467). The analysis of age distribution among the study groups revealed significant differences (P=0.0031). The median age for the total group was 79 years (84–92). In the no nutrition-related risk group, the median age was 79 years (74–83), while the nutrition-related risk group had a median age of 82 years (77.5–85).

The results show that only 2 participants (1.1%) were classified as having a major nutrition-related risk (GNRI <82.0). A moderate risk (GNRI 82.0 to <92.0) was identified in 18 participants (9.7%). Additionally, 28 participants (15.1%) fell into the low-risk category (GNRI 92.0 to ≤98.0). Notably, the majority of participants, 137 individuals or 74.1%, were categorized as having no nutrition-related risk (GNRI > 98.0).

The median BMI for the total group was 27.2 kg/m², with an interquartile range of 30.6 to 46.5 kg/m². In the no nutrition-related risk group, the median BMI was 28.5 kg/m² (26.4–31.9 kg/m²). In contrast, the nutrition-related risk group had a significantly (P<0.0001) lower median BMI of 22.8 kg/m² (20.7–24 kg/m²).

Table 2 presents the BMI distribution among the study population, categorized by both the WHO and CDH criteria. Based on CDH criteria in the total group, 43 participants (23.2%) were classified with underweight, 79 (42.7%) with normal weight, and 63 (34.1%) with excess body weight. When using WHO criteria, 4 participants (2.2%) were classified with underweight, 54 (29.2%) with normal weight, and 127 (68.6%) with excess body weight. The distribution of BMI classifications based on WHO and CDH criteria revealed notable differences.

All individuals classified with underweight according to the WHO criteria were also classified with underweight by CDH criteria. The normal weight group according to WHO had 39 participants (72.2%) classified with underweight and 15 participants (27.8%) classified with normal weight by CDH criteria. The excess body weight group according to WHO included 64 participants (85.3%) classified with normal weight and 63 participants (49.6%) classified with excess body weight by CDH criteria.

According to the CDH criteria, participants identified with underweight were classified with underweight (n=4; 9.3%) and with normal weight by WHO criteria (n=39, 90.7%). Participants classified with normal weight by CDH included 15 individuals (18.9%) who were classified with normal weight by WHO, and 64 participants (81.1%) who were classified with excess body weight by WHO. Among those classified with excess body weight by CDH, 63 individuals (100%) were also classified with excess body weight by WHO (gamma coefficients =1; P<0.0001).

The results demonstrated significant differences in BMI classifications between the 2 criteria, with substantial variation in how individuals were categorized across different weight status groups.

Figures 1 and 2 compare the distribution of nutrition-related risk based on the GNRI in relation to BMI classifications by CDH (Figure 1) and WHO (Figure 2). Among participants classified with underweight by CDH BMI criteria, 75.0% were identified as having a nutrition-related risk, whereas only 6.3% of the study group were categorized similarly under WHO BMI standards. Conversely, 81.3% of patients with normal weight, as diagnosed using the WHO BMI criteria, were classified as having a nutrition-related risk. In contrast, only 22.9% of patients with normal weight, as diagnosed using the CDH BMI criteria, were classified as having a nutrition-related risk. Among the excess body weight group diagnosed using the WHO BMI criteria, 12.5% were classified as having a nutrition-related risk, while only 2.1% of the excess body weight group diagnosed using the CDH BMI criteria were classified as having a nutrition-related risk.

Table 3 presents a comparison of nutrition-related risk based on the GNRI with BMI classifications according to the CDH and WHO criteria. The results highlight distinct differences in how each classification system associates BMI categories with nutrition-related risk levels.

Using the CDH BMI criteria, the study categorized 3.78% of the underweight group as having no risk, 10.81% as having low risk, 7.57% as having moderate risk, and 1.08% as having major risk. In the normal weight category, the majority (36.76%) were classified as having no risk, 3.78% as having low risk, 2.16% as having moderate risk, and none were identified as having major risk. For the excess body weight group, 34.23% were classified as having no risk, 0.54% as having low risk, and there were no cases of moderate or major risk. Kendall Tau-b −0.591 (95% CI: −0.666–0.516) indicated a statistically significant (P<0.001) inverse relationship between BMI status and nutrition-related risk. Higher BMI was associated with lower nutrition-related risk.

Using the WHO BMI criteria, only 0.54% of individuals classified with underweight were identified as having no risk, with each risk category low, moderate, and major also comprising 0.54% of individuals. For those classified as having normal weight, 8.11% were categorized as having no risk, 12.43% were at low risk, 8.11% had moderate risk, and 0.54% were at major risk. Among individuals classified as having excess body weight, a substantial proportion (65.41%) were classified as having no risk, 2.16% were at low risk, 1.08% were at moderate risk, and none were categorized with major risk. Kendall Tau-b of −0.687 (95% CI: −0.790–0.584) indicated a statistically significant inverse relationship (P<0.001), where a higher BMI was associated with lower nutrition-related risk.

The classification better suited for assessing malnutrition is the CDH classification, as it identifies a larger number of individuals with underweight and allows for a stronger correlation with nutrition-related risks based on the GNRI. The CDH classification demonstrates greater sensitivity, detecting more cases of underweight individuals across GNRI risk categories, making it a more effective tool for identifying malnutrition in clinical and population-based assessments.

Discussion

STRENGTHS AND LIMITATIONS:

To the best of our knowledge, this is the first study to compare the prevalence of malnutrition risk assessed by GNRI based on 2 BMI classification systems – WHO and CDH – in an older adult population. The results indicate that the CDH criteria are more sensitive in identifying underweight individuals and those at risk of malnutrition, which has important clinical implications. In interpreting BMI values, we considered the specific characteristics of the older adult population and their distinct needs.

Although we made every effort to conduct a reliable study, we are aware of its limitations. The first limitation was the sample size; however, we achieved the minimum planned number of participants. Another limitation was the lack of patient follow-up after the study, which prevents the assessment of the long-term effects of using different BMI criteria. Additionally, the study did not include a sex-based analysis due to the small size of the group.

Conclusions

This study demonstrates significant discrepancies between the BMI classification systems proposed by the WHO and the CDH when applied to the older population. The WHO criteria may underestimate the number of older adults with underweight and overestimate those with excess body weight. The CDH classification, which is more sensitive in identifying underweight individuals, appears to be a more appropriate tool for assessing nutritional status and related risks in the older adult population.

The CDH classification is better suited for assessing nutrition-related risks. It identifies a higher number of underweight individuals and shows a stronger correlation with nutrition-related risks based on the GNRI. The CDH classification demonstrates greater sensitivity, detecting more cases of underweight across GNRI risk categories, making it a more effective tool for identifying malnutrition in both clinical and population-based assessments.

Therefore, it is essential to develop tailored guidelines for this age group, particularly for interpreting BMI, in order to ensure more accurate health assessments and improve care for older adults.