Association Between Geriatric Nutritional Risk Index and Mortality Outcomes in Elderly Cancer Survivors in the United States

doi:10.21203/rs.3.rs-4891318/v1

Download PDF

Research Article

Association Between Geriatric Nutritional Risk Index and Mortality Outcomes in Elderly Cancer Survivors in the United States

https://doi.org/10.21203/rs.3.rs-4891318/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Cancer poses a significant global health burden, with increasing incidence and mortality rates, particularly among elderly populations. This study aimed to evaluate the association between the Geriatric Nutritional Risk Index (GNRI) and mortality outcomes (all-cause, cancer, and cardiovascular disease) among elderly cancer survivors in the United States.Data from the National Health and Nutrition Examination Survey (NHANES) were analyzed.

Methods

Participants were categorized into well-nourished, mildly malnourished, and moderately to severely malnourished groups. Weighted multivariable Cox proportional hazards regression models were used to calculate hazard ratios (HR) and 95% confidence intervals (CI) for mortality outcomes.

Results

The analysis included 2,582 elderly cancer survivors. Compared to the well-nourished group, the malnourished groups had higher proportions of older individuals, males, widowed or divorced individuals, current smokers, and deaths. Lower GNRI was associated with a higher risk of all-cause mortality (HR: 2.41, 95% CI: 1.67–3.48), cancer mortality (HR: 2.24, 95% CI: 1.32–3.80), and cardiovascular mortality (HR: 2.72, 95% CI: 1.41–5.25).

Conclusions

Assessing the nutritional status of elderly cancer survivors using GNRI can help determine their prognosis and guide interventions to improve long-term outcomes.

GNRI

NHANES

Elderly cancer survivors

Mortality outcomes

Nutritional status

Cancer, a multifactorial genetic disease, is responsible for approximately 7.5 million deaths annually and poses a significant economic burden worldwide[1]. Recent statistics reveal that the European region recorded 1,261,990 cancer-related deaths in 2023. Moreover, in 2024, the United States witnessed 2,001,140 new cancer cases and 611,720 deaths, drawing considerable attention from various countries[2, 3]. In China, cancer has emerged as the leading cause of death since 2010, with both incidence and mortality rates steadily increasing each year[4]. Consequently, mounting concerns have arisen regarding the long-term prognosis of cancer survivors. In the United States alone, the population of cancer survivors surpasses 15 million, with many enduring a compromised quality of life[5]. This issue becomes particularly pressing as the United Kingdom is projected to reach 4 million cancer survivors by 2030, underscoring the imperative need to investigate the health-related challenges faced by this specific demographic[6]. Therefore, it is of utmost importance to thoroughly examine the intricate relationship between the nutritional status of cancer patients and their long-term prognosis within the context of the United States.

Older individuals are at a higher risk of malnutrition due to age-related physiological changes and increased vulnerability to common diseases like hypertension and diabetes[7]. Elderly cancer patients often experience increased metabolic demands and protein loss due to systemic inflammation, which can further elevate the risk of malnutrition, especially in long-term survivors[8]. Hospitalized elderly cancer patients undergoing treatments are particularly at risk of poor prognosis and mortality[9]. The GNRI is a useful tool for assessing the nutritional status of older individuals, with straightforward statistical indicators that accurately predict adverse events during hospitalization[10]. While the GNRI is commonly used to assess the prognosis of elderly patients with conditions like hypertension, diabetes, and chronic obstructive pulmonary disease[11–13]., there is limited research on its application in cancer patients Therefore, further investigation is needed to understand how the GNRI can help determine the prognosis of elderly cancer survivors.

Utilizing a sizable long-term follow-up cohort from the NHANES, we examined the correlation between GNRI and mortality in cancer survivors. Our hypothesis posits that individuals with a lower GNRI among cancer survivors will exhibit a higher mortality rate in comparison to those with a higher GNRI.

2.1. Study population

NHANES is a series of surveys conducted by the National Center for Health Statistics (NCHS) to evaluate the health and nutritional status of the non-institutionalized population in the United States. It utilizes complex, multi-stage probability sampling in each survey cycle to ensure the national representativeness of the samples[14]. The survey consists of household interviews and physical examinations carried out at Mobile Examination Centers (MEC). This cohort study utilized data from elderly participants (aged ≥ 60) in the NHANES cycles from 1999 to 2018, following the guidelines of the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE). A total of 3791 cancer survivors aged 60 years or older from 1999 to 2018 were included. After excluding participants with missing GNRI data, loss to follow-up, and missing covariate data, a final cohort of 2582 participants was established. Figure 1 provides detailed information on participant inclusion.

2.2. Diagnosis of cancer

The assessment of cancer diagnosis in this study was based on self-reporting. Participants were asked the question, "Have you ever been informed by a doctor or other healthcare professional that you have had cancer or any type of malignancy?" Those who responded positively were classified as cancer survivors[15].

2.3. The Geriatric Nutrition Risk Index

The GNRI calculation involves objective factors such as height, weight, and serum albumin, as per the formula: GNRI= (1.489* serum albumin (g/L)) + (41.7* weight (kg)/ideal weight (kg)) [10, 16]. Ideal weight is calculated using the formula: 22* height (m) squared[17]. In cases where weight exceeds ideal weight, the weight-to-ideal-weight ratio is set to 1. Participants are categorized based on their nutritional status thresholds: Moderate to severe risk of malnutrition (M/S Risk): <92; Low risk of malnutrition (Low Risk): ≥92 to < 98; Nutritional health (No Risk): ≥98[12].

2.4. Assessment of mortality

The NCHS in the United States utilized the National Death Index data to create a publicly accessible mortality linkage file for participants in the NHANES cycles from 1999 to 2018, with a cutoff date of December 31, 2019. This study focused on key outcome events, including all-cause mortality, cancer mortality, and cardiovascular disease (CVD) mortality. All-cause mortality encompasses deaths from any cause classified by the Tenth Revision of the International Classification of Diseases (ICD-10). Cancer mortality is indicated by ICD-10 codes C00-C97, while CVD mortality is indicated by ICD-10 codes I00-I09, I11, I13, I20-I51, and I60-I69.

2.5. Assessment of covariates

Previous research has identified potential covariates including age, gender, race, marital status, education level, family income (PIR), body mass index (BMI), smoking status, alcohol consumption, alanine aminotransferase (ALT), uric acid (UA), CVD, hypertension, and diabetes mellitus (DM) [11–13, 18–21]. Racial data of the subjects is categorized into five groups: Mexican Americans, non-Hispanic blacks, non-Hispanic whites, other Hispanics, or other races. Marital status is divided into four groups: married, unmarried, cohabiting with a partner, and other situations such as being widowed or divorced. Education level is classified as less than high school, high school or equivalent, and more than high school. Family income categories are based on PIR and are divided into low income (≤ 1.3), medium income (1.31 to 3.5), and high income (> 3.5)[22]. BMI is calculated using a standard method based on weight and height. Smoking status is classified as never smokers (smoked fewer than 100 cigarettes), former smokers (smoked more than 100 cigarettes but quit), and current smokers (smoked more than 100 cigarettes and currently smoke)[22]. Alcohol consumption is categorized as never drinkers (consumed < 12 drinks in a lifetime), former drinkers (consumed ≥ 12 drinks in the past year and did not drink in the past year, or did not drink in the past year but consumed ≥ 12 drinks in a lifetime), and current drinkers (consume ≥ 1 drink per day)[23]. History of diseases such as hypertension, diabetes, and cardiovascular disease is determined based on responses in the questionnaire regarding whether a doctor has diagnosed these conditions. Subjects self-report a history of CVD, including heart failure, coronary heart disease, angina, heart attack, or stroke[11, 12].

2.6. Statistical analysis

The NHANES database employs a sophisticated multi-stage probability sampling method. To analyze the NHANES dataset accurately, it is advisable to use sampling weights to adjust the statistical estimates. When analyzing BMI, ALT, and UA measurements collected in the MEC, it is essential to incorporate sample weights, clustering, and stratification details. As per NHANES guidelines, MEC weights should be applied [24]. The calculation of sampling weights is as follows: data from the 1999–2000 and 2001–2002 cycles have a weight of wtmec4 year/5, while data from the 2003–2004, 2005–2006, 2007–2008, 2009–2010, 2011–2012, 2013–2014, 2015–2016, and 2017–2018 cycles have a weight of wtmec2 year/10. Follow-up time is calculated from the MEC examination completion date. The National Death Index is updated every 4 years, with the latest follow-up data current as of December 31, 2019. Hence, the follow-up time for each participant is determined from the MEC examination date to the date of death or the end of follow-up on December 31, 2019.

Categorical variables were represented as percentages (%), while continuous variables were expressed as mean (standard deviation, SD) or median (interquartile range, IQR) as appropriate. Group differences were analyzed using one-way ANOVA for normally distributed data, Kruskal-Wallis test for skewed data, and chi-square test for categorical variables.

To evaluate the HR and 95% confidence intervals (95% CI) for all-cause, cancer, and cardiovascular disease (CVD) mortality among cancer survivors in relation to GNRI, weighted multivariable Cox proportional hazards regression models were utilized. Proportional hazards assumption testing was performed using Schoenfeld residuals, revealing no violations. Model 1 was adjusted for age, sex, race, marital status, education level, and poverty-income ratio (PIR). Model 2 further included adjustments for body mass index (BMI), smoking status, alcohol consumption, alanine aminotransferase (ALT), and uric acid (UA). Model 3 additionally considered adjustments for CVD, hypertension, and diabetes mellitus (DM).

GNRI was categorized into three subgroups based on nutritional status thresholds and included as both a categorical and continuous variable in the model. Following adjustments in Model 3, restricted cubic spline (RCS) regression was conducted with three knots at the 10th, 50th, and 90th percentiles of GNRI to assess linearity and explore the dose-response relationship between GNRI and mortality. Kaplan-Meier survival curves were used to evaluate long-term survival rates for the three GNRI subgroups, with between-group comparisons performed using the log-rank test.

Subgroup analyses based on gender (male vs. female), smoking status (never vs. former vs. current), history of CVD (no vs. yes), history of hypertension (no vs. yes), and history of DM (no vs. yes) were conducted using multivariable Cox proportional hazards regression models with covariate adjustments similar to those in Model 3. Interactions between subgroups were assessed using likelihood ratio tests.

Additionally, subgroup analyses were conducted based on gender (male vs. female), smoking status (never vs. former vs. current), history of CVD (no vs. yes), history of hypertension (no vs. yes), and history of DM (no vs. yes) using multivariable Cox proportional hazards regression models with the same adjusted covariates as in Model 3. Interactions between subgroups were assessed using likelihood ratio tests.

Several sensitivity analyses were performed to assess the stability of the conclusions. Firstly, participants who died within 2 years of inclusion were excluded to mitigate potential reverse causality. Secondly, multiple imputation was applied for missing variables, and the relationship between GNRI and all-cause mortality was re-examined in the complete imputed data.

Statistical power calculations were not performed in advance due to the sample size being determined by existing data. Data analysis was conducted using R software, R survey package, and Free Statistics software, with statistical significance set at a two-sided p-value of less than 0.05. The data analysis took place from January to March 2024.

2.7. Standard protocol approval, registration, and patient consent

The NHANES procedures and protocols were sanctioned by the NCHS Ethics Review Board, with written informed consent obtained from all participants. As this study entails secondary analysis, no further approval from an institutional review board is required.

3.1. Baseline characteristics of study participants

Out of the initial cohort of 3791 elderly cancer survivors aged 60 years and older, a total of 1209 individuals were excluded from the analysis due to unavailable GNRI data (n = 679), loss to follow-up (n = 1), or missing covariate data (n = 529). Consequently, the final sample size for analysis consisted of 2582 elderly cancer survivors (Fig. 1).

At the baseline, among the 2,582 participants included in the analysis, there were 50 (1.9%) individuals classified as mildly malnourished, 20 (0.8%) individuals classified as moderately to severely malnourished, and 2,512 (97.3%) individuals classified as relatively well-nourished. Table 1 provides an overview of the baseline characteristics of the participants. The mean age of the participants was 73.0 (± 7.1) years, and 1,404 (54.4%) of them were male.

Table 1

Baseline characteristics of participants by risk category (GNRI score).
Variables	Total (n = 2582)	No Risk (n = 2512)	Low Risk (n = 50)	M/S Risk (n = 20)	P-value
Age, Mean ± SD	73.0 ± 7.1	72.9 ± 7.0	75.8 ± 8.4	74.7 ± 6.3	0.009
Sex, n (%)					0.005
Male	1404 (54.4)	1354 (53.9)	33 (66.0)	17 (85.0)
Female	1178 (45.6)	1158 (46.1)	17 (34.0)	3 (15.0)
Race, n (%)					0.098
White	1923 (74.5)	1873 (74.6)	34 (68.0)	16 (80.0)
Black	327 (12.7)	310 (12.3)	13 (26.0)	4 (20.0)
Mexican	149 (5.8)	149 (5.9)	0 (0)	0 (0)
Hispanic	105 (4.1)	104 (4.1)	1 (2.0)	0 (0)
Other Races	78 (3.0)	76 (3.0)	2 (4.0)	0 (0)
Marital status, n (%)					0.008
Married	1543 (59.8)	1513 (60.2)	18 (36.0)	12 (60.0)
Never married	82 (3.2)	81 (3.2)	1 (2.0)	0 (0)
Living with partner	42 (1.6)	41 (1.6)	0 (0)	1 (5.0)
Others	915 (35.4)	877 (34.9)	31 (62.0)	7 (35.0)
Educational level, n (%)					0.687
Less than high school	614 (23.8)	596 (23.7)	14 (28.0)	4 (20.0)
High school or equivalent	599 (23.2)	583 (23.2)	13 (26.0)	3 (15.0)
Above high school	1369 (53.0)	1333 (53.1)	23 (46.0)	13 (65.0)
PIR, n (%)					0.181
≤ 1.30	530 (20.5)	514 (20.5)	13 (26.0)	3 (15.0)
1.31–3.50	1167 (45.2)	1130 (45.0)	23 (46.0)	14 (70.0)
> 3.50	885 (34.3)	868 (34.6)	14 (28.0)	3 (15.0)
Smoking status, n (%)					< 0.001
Never	1114 (43.1)	1095 (43.6)	15 (30.0)	4 (20.0)
Former	1209 (46.8)	1177 (46.9)	20 (40.0)	12 (60.0)
Now	259 (10.0)	240 (9.6)	15 (30.0)	4 (20.0)
Drinking status, n (%)					0.215
Never	365 (14.1)	355 (14.1)	7 (14.0)	3 (15.0)
Former	709 (27.5)	682 (27.1)	18 (36.0)	9 (45.0)
Now	1508 (58.4)	1475 (58.7)	25 (50.0)	8 (40.0)
BMI (kg/m²), Mean ± SD	28.5 ± 5.9	28.7 ± 5.7	20.3 ± 2.4	19.1 ± 3.2	< 0.001
ALT (IU/L), Median(IQR)	20.0 (16.0, 25.0)	20.0 (16.0, 25.0)	15.0 (12.0, 22.8)	18.5 (14.0, 24.2)	0.001
UA ( mg/dl), Mean ± SD	5.8 ± 1.5	5.8 ± 1.5	5.1 ± 1.4	5.1 ± 1.8	< 0.001
CVD, n (%)					0.330
No	1812 (70.2)	1761 (70.1)	34 (68.0)	17 (85.0)
Yes	770 (29.8)	751 (29.9)	16 (32.0)	3 (15.0)
Hypertension, n (%)					0.302
No	712 (27.6)	687 (27.3)	18 (36.0)	7 (35.0)
Yes	1870 (72.4)	1825 (72.7)	32 (64.0)	13 (65.0)
DM, n (%)					0.001
No	1816 (70.3)	1753 (69.8)	46 (92.0)	17 (85.0)
Yes	766 (29.7)	759 (30.2)	4 (8.0)	3 (15.0)
Status, n (%)					< 0.001
Alive	1415 (54.8)	1401 (55.8)	10 (20.0)	4 (20.0)
Death	1167 (45.2)	1111 (44.2)	40 (80.0)	16 (80.0)
Abbreviations: GNRI, Geriatric Nutrition Risk Index; PIR, Ratio of family income to poverty; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); ALT, alanine transaminase; UA, uric acid; CVD, cardiovascular disease; DM, diabetes.

Comparing the malnourished group to the well-nourished group, the malnourished group tended to be older (mean ages of 72.9 [± 7.0] years, 75.8 [± 8.4] years, and 74.7 [± 6.3] years, respectively), have a higher proportion of males (1,354 [53.9%], 33 [66.0%], and 17 [85.0%], respectively), a higher proportion of individuals who were widowed or divorced (877 [34.9%], 31 [62.0%], and 7 [35.0%], respectively), a higher proportion of current smokers (240 [9.6%], 15 [30.0%], and 4 [20.0%], respectively), lower mean BMI values (28.7 [± 5.7] kg/m², 20.3 [± 2.4] kg/m², and 19.1 [± 3.2] kg/m², respectively), lower mean UA levels (5.8 [± 1.5] mg/dl, 5.1 [± 1.4] mg/dl, and 5.1 [± 1.8] mg/dl, respectively), but a lower proportion with a history of diabetes (759 [30.2%], 4 [8.0%], and 3 [15.0%], respectively). Additionally, the malnourished groups had a higher proportion of deaths (1,111 [44.2%], 40 [80.0%], and 16 [80.0%], respectively).

At the end of the follow-up period, 1,415 (54.8%) of the 2,582 participants were alive, while 1,167 (45.2%) had died. Supplementary Table 1 provides additional baseline data for both survivors and non-survivors..

3.2. Associations between GNRI and mortality among cancer survivors

Over the course of a 15–21 year follow-up period from 1999 to 2018 in the NHANES database, a total of 1167 all-cause deaths, 338 cancer-related deaths, and 262 deaths attributed to cardiovascular disease were identified. The median follow-up duration was 80 (45, 128) months. In the Cox proportional hazards regression model with weighting for a single factor, it was observed that for each incremental rise in the GNRI, there was a corresponding 2% reduction in the hazard of all-cause mortality among individuals who have survived cancer (HR = 0.98, 95% CI 0.97–0.99, P < 0.001) (See Supplementary Table 2). In the multivariable Cox proportional hazards regression model incorporating GNRI as a continuous variable and adjusting for relevant covariates, an incremental increase in GNRI was found to be significantly associated with a 5% decrease in the hazard ratio for all-cause and cancer-related mortality among cancer survivors (HR = 0.95, 95% CI 0.93–0.97, HR = 0.95, 95% CI 0.92–0.98, model 3) and a 4% decrease in the hazard ratio for cardiovascular disease mortality (HR = 0.96, 95% CI 0.93–0.99, model 3). When GNRI was treated as a categorical variable and adjusted for relevant covariates in model 3, patients with mild malnutrition exhibited a 1.76 (95% CI 0.79–3.93) increased risk of all-cause mortality, a 1.29 (95% CI 0.44–3.77) increased risk of cancer-related mortality, and a 2.86 (95% CI 0.71–11.43) increased risk of CVD mortality compared to cancer survivors in good nutritional health. Patients with moderate to severe malnutrition demonstrated even higher risks, with an all-cause mortality risk of 3.86 (95% CI 2.31–6.45), a cancer-related mortality risk of 2.01 (95% CI 0.57–7.15), and a CVD mortality risk of 6.67 (95% CI 2.27–19.64) (Table 2).

Table 2

Weighted association between GNRI and all-cause, cancer, and CVD mortality among US cancer survivors in multiple regression model.
	No	Low	M/S	P for Trend	GNRI
All-cause Mortality
Unadjusted	1.00	3.06(1.86,5.04)	4.15(2.23,7.73)	<0.001	0.98(0.97,0.99)
Model 1	1.00	2.06(1.00,4.24)	3.35(1.89,5.94)	<0.001	0.99(0.98,1.00)
Model 2	1.00	1.84(0.89,3.81)	3.47(2.09,5.78)	<0.001	0.95(0.93,0.96)
Model 3	1.00	1.76(0.79,3.93)	3.86(2.31,6.45)	<0.001	0.95(0.93,0.97)
Cancer Mortality
Unadjusted	1.00	1.85(0.73,4.72)	2.89(0.81,10.22)	0.025	0.98(0.97,0.99)
Model 1	1.00	1.47(0.54,4.02)	2.10(0.62,7.08)	0.143	0.99(0.98,1.00)
Model 2	1.00	1.26(0.43,3.70)	2.00(0.57,7.03)	0.271	0.95(0.92,0.98)
Model 3	1.00	1.29(0.44,3.77)	2.01(0.57,7.15)	0.254	0.95(0.92,0.98)
CVD Mortality
Unadjusted	1.00	4.02(1.53,10.53)	4.74(1.49,15.09)	<0.001	0.98(0.97,1.00)
Model 1	1.00	2.80(0.89,8.83)	3.30(1.11,9.83)	0.003	1.00(0.98,1.02)
Model 2	1.00	3.36(1.09,10.31)	4.62(1.54,13.88)	<0.001	0.95(0.92,0.98)
Model 3	1.00	2.86(0.71,11.43)	6.67(2.27,19.64)	0.002	0.96(0.93,0.99)
Data are presented as HR (95% CI)
Model 1: adjust for age, sex, race, marital status, educational level, PIR
Model 2: adjust for age, sex, race, marital status, educational level, PIR, BMI, smoking status, drinking status, ALT, UA
Model 3: adjust for age, sex, race, marital status, educational level, PIR, BMI, smoking status, drinking status, ALT, UA, CVD, Hypertension, DM

The restricted cubic spline (RCS) curve fitting analysis demonstrated a curvilinear relationship between GNRI and both all-cause mortality and cardiovascular disease (CVD) mortality among cancer survivors. In contrast, the association between GNRI and cancer-specific mortality followed a linear trend. Higher GNRI values were associated with lower rates of overall mortality, cancer-specific mortality, and CVD mortality (Fig. 2). Furthermore, the weighted survival analysis indicated that individuals with good nutritional health had the highest overall survival rates across all mortality outcomes examined (Fig. 3).

3.3. Subgroup analyses and sensitivity analyses

Subgroup analyses showed that the relationship between GNRI and all-cause mortality among cancer survivors remained consistent across various subgroups, with a 3–6% reduction in all-cause mortality risk per unit increase in GNRI. Additionally, the association between GNRI and CVD mortality was statistically significant in male individuals (Hazard Ratio [HR] = 0.94, 95% Confidence Interval [CI] 0.90–0.98), those with a history of hypertensive disease (HR = 0.95, 95% CI 0.92–0.98), and the overall cancer population (Fig. 4).

The findings from the sensitivity analyses are presented in supplementary tables 3 and 4, along with supplementary Fig. 1. Following the exclusion of participants who passed away within a 2-year follow-up period, the adjusted HR for all-cause mortality among cancer survivors with moderate to severe malnutrition, in comparison to those with good nutritional status, was 2.99 (95% CI 1.66–5.38, p = 0.024, model 3). Additionally, the adjusted HR for CVD mortality was 8.50 (95% CI 2.65–27.24, p = 0.013, model 3). A total of 529 participants were eliminated from the analysis due to missing relevant covariates, with a maximum covariate missing rate of 10.77%. Following multiple imputation, the GNRI exhibited a consistent association with the risk of all-cause mortality, as evidenced by a HR of 0.95 for the entire cohort of 3111 participants, aligning with the findings observed prior to imputation.

The results of our study suggest that malnutrition is associated with an increased likelihood of negative outcomes in individuals who have survived cancer. Our conclusions are supported by subgroup and sensitivity analyses, which reinforce the reliability of our findings.

The GNRI has become a widely accepted tool for predicting mortality risk in older patients. Bouillanne et al. introduced the GNRI in 2005 as a method for evaluating the nutritional status of elderly individuals, highlighting its effectiveness in quantifying the risk of mortality[10].In comparison to other screening tools, including the Nutritional Risk Screening 2002 (NRS-2002), Malnutrition Universal Screening Tool (MUST), and Malnutrition-Inflammation Score (MIS), the GNRI is noted for its ease of use and enhanced accuracy[9, 25]. The GNRI, a novel index based on body mass index and serum albumin levels, demonstrates a significant improvement in mortality risk prediction compared to utilizing solely body mass index and serum albumin levels[26].Moreover, the GNRI can serve as a valuable tool in everyday clinical settings and as a means of monitoring patients over an extended period, facilitating the early detection of individuals susceptible to malnutrition[11–13].

Prior research has investigated the correlation between GNRI levels and mortality among various cancer patients. Xie et al. conducted a meta-analysis incorporating data from 9 studies involving 2153 gastrointestinal cancer patients, revealing a significant association between lower GNRI levels and reduced overall survival (HR = 1.94, 95% CI 1.65–2.28, p < 0.001). These findings suggest that GNRI can function as an independent prognostic indicator for complications and long-term outcomes in patients with gastrointestinal cancer[27]. Yiu et al. conducted a systematic review of 10 studies encompassing a total of 2793 head and neck cancer patients. Their findings indicated a significant correlation between lower GNRI levels and decreased overall survival rates (HR = 2.84, 95% CI 2.07–3.91, P < 0.00001). This suggests that GNRI may serve as a valuable prognostic tool in routine clinical practice for predicting unfavorable outcomes in patients with head and neck cancer[28]. Shen et al. conducted a comprehensive literature review on the prognostic significance of GNRI in non-small cell lung cancer patients. Their findings indicate that lower GNRI levels are associated with decreased overall survival (HR = 1.96, 95% CI 1.66–2.30, p < 0.00001) and disease-free survival (HR = 1.74, 95% CI: 1.36–2.23, p < 0.0001). These results suggest the potential for GNRI to serve as a valuable tool for patient stratification and the development of personalized treatment strategies in this population[29]. Xu et al. highlighted the significance of malnutrition as an autonomous prognostic indicator for individuals with gastric cancer, correlating with suboptimal tumor treatment outcomes and heightened complication rates[30]. Our investigation, utilizing data from the National Health and Nutrition Examination Survey encompassing older adults in the United States, specifically examined survivors of various cancer types spanning from 1999 to 2018. By conducting an observational analysis on this extensive cohort, we substantiated the link between diminished GNRI values and elevated mortality rates, thereby enhancing the comprehension of the interplay between GNRI and cancer prognosis.

Yu et al. conducted a longitudinal study spanning six years, which revealed a significant association between severe malnutrition in the elderly and increased all-cause mortality. Utilizing the GNRI as a tool to evaluate nutritional status, the researchers observed that individuals aged 60–69 classified as malnourished had a hazard ratio (HR) for death of 2.86 (95% CI 1.44–5.68) compared to those with adequate nutrition. Similarly, in the 70–79 age group, malnourished individuals had a HR of 2.60 (95% CI 1.39–4.85) for mortality compared to their well-nourished counterparts[31]. Huo et al. conducted a study involving 10,037 elderly hypertensive patients from the NHANES database, which revealed a significant correlation between moderate to severe malnutrition, as assessed by GNRI, and reduced survival rates. Upon treating GNRI as a continuous variable and controlling for pertinent covariates, the researchers observed that lower GNRI levels were linked to elevated risks of all-cause mortality and cardiovascular mortality, with hazard ratios of 0.958 (95% CI 0.949–0.967) and 0.956 (95% CI 0.941–0.972), respectively. When treating GNRI as a categorical variable, individuals with moderate to severe malnutrition exhibited significantly higher risks of all-cause mortality and cardiovascular mortality compared to those with good nutrition (HR = 2.112, 95% CI 1.377–3.240, HR = 2.604, 95% CI 1.603–4.229)[11]. In their study, Shen et al. analyzed 4400 elderly diabetic patients from the NHANES database and, after adjusting for relevant covariates, observed that for every one-unit increase in GNRI, the risk of all-cause mortality decreased by 5% (HR = 0.95, 95% CI 0.942–0.966) and the risk of cardiovascular mortality decreased by 4% (HR = 0.96, 95% CI 0.935–0.989). Upon stratifying GNRI, it was observed that the group exhibiting moderate to severe malnutrition displayed the lowest survival rate[12]. In a study conducted by Chai et al., it was discovered that malnutrition was linked to a heightened long-term all-cause mortality risk in a cohort of 579 elderly individuals diagnosed with chronic obstructive pulmonary disease. Following adjustments for all pertinent covariates, the malnourished cohort exhibited a twofold increase in the risk of all-cause mortality in comparison to the nutritionally healthy group (HR = 2.47, 95% CI 1.36–4.5)[13]. This study is the first to examine the correlation between GNRI and mortality in elderly cancer patients. Our findings align with previous research, underscoring the widespread applicability of GNRI as a prognostic indicator for disease outcomes.

The etiology of heightened mortality rates linked to malnutrition among individuals with cancer is intricate and multifaceted. Primarily, cancer patients may exhibit anorexia and diminished appetite while hospitalized as a result of diverse treatment reactions, resulting in ongoing deterioration of their nutritional well-being[32, 33]. This phenomenon is particularly pronounced in elderly cancer patients, who are more vulnerable to malnutrition and face an elevated likelihood of experiencing negative outcomes due to prolonged illness and compromised immune function during hospitalization[34, 35]. Secondly, within the realm of cancer, systemic inflammation contributes to heightened metabolism and enhanced breakdown of fats and proteins[36]. This inflammatory response not only adversely affects blood vessels, causing endothelial dysfunction and smooth muscle cell proliferation and migration, but also heightens the likelihood of cardiovascular events in individuals with cancers[37, 38]. Prolonged cancer may ultimately lead to cachexia, a state characterized by severe metabolic disruption, negative protein and energy balance, and weight loss exceeding 5%[39]. In cachectic individuals, the inflammatory response is heightened, resulting in increased catabolic processes in multiple tissues, which worsens the adverse effects of cancer treatments and associated complications[40]. The persistent presence of cachexia and inflammation may also play a role in diminishing survival rates among cancer patients[41, 42]. Additionally, research on immune cells and cytokines within the tumor microenvironment has demonstrated a clear link between immune metabolism and cancer-induced cachexia. The dysregulated immune metabolism in cancer patients further contributes to the development of cachexia[43]. In conditions of malnutrition, there is a reduction in the quantity and efficacy of T cells in the body, resulting in the deactivation of immune cells and immune suppression[44, 45]. This can exacerbate the primary tumor or give rise to additional complications, ultimately impacting the patient's mortality[46].

Our study is constrained by the inherent limitations of the NHANES database, primarily due to its observational nature which precludes definitive establishment of causal relationships between GNRI levels and mortality rates among cancer survivors. Utilizing data from elderly cancer survivors in the United States, our study aimed to investigate the association between GNRI and mortality rates. However, further prospective studies are warranted to validate our findings in this domain.

Based on the inherent characteristics of the NHANES database, our study has several limitations. Firstly, as an observational study, we are unable to establish a definitive causal relationship between GNRI levels and various mortality rates among cancer survivors. Our research utilized data from elderly cancer survivors in the United States to explore this relationship, necessitating future prospective studies to validate our findings. Secondly, our observational design may be subject to selection bias, which could potentially affect the robustness of our conclusions. Despite controlling for relevant covariates based on previous literature and clinical evidence, the influence of confounding factors cannot be entirely excluded. To mitigate this, we employed a multivariable Cox proportional hazards regression model and conducted subgroup and sensitivity analyses. Thirdly, the NHANES database does not provide specific information regarding cancer characteristics such as type, grade, and stage, limiting our ability to assess the relationship between GNRI levels and mortality rates for different cancer types or stages. Nevertheless, our findings align with previous studies on the relationship between GNRI and mortality rates in individual cancers, suggesting the broader applicability of our results across all cancers. Fourthly, GNRI data in our study was based on a single measurement, and baseline data for patients may change over time, potentially affecting the stability of the observed relationship between GNRI and mortality rates. Finally, our analysis represents a secondary analysis of NHANES data, which generally provides lower evidence strength compared to primary study designs. However, secondary analysis allows for the comprehensive utilization of original data, and our application of weighting techniques enhances the generalizability of our findings.

Our study found that among older cancer survivors in the US, those with poorer nutritional status (lower GNRI) had significantly higher all-cause mortality, cancer mortality, and cardiovascular disease mortality compared to those with better nutritional status. This suggests that regularly assessing the nutritional status of older cancer patients and implementing appropriate nutritional support measures are of great importance for improving their long-term prognosis. Future research is still needed to further explore the impact of nutritional intervention on the prognosis of older cancer patients, to provide a basis for developing more targeted nutritional support strategies.

Author contributions

The conception and design: Min Tang; analysis and interpretation of the data: Jingyi Li; the drafting of the paper: Bo Su and Fangfang Chen; revising it critically for intellectual content: Min Tang; and the final approval of the version to be published: all authors; and that all authors agree to be accountable for all aspects of the work.

Funding

No funding had received to conduct this study.

Clinical trial number

Not applicable.

Data availability statement

The data that support the findings of this study are a combination of publicly available and restricted data. Please see the NHANES website for more information: https://www.cdc.gov/nchs/nhanes/index.htm.

Ethics approval and consent to participate

Human Ethics and Consent to Participate declarations: not applicable.

Consent for publication

Not application.

Competing interests

The authors declare that they have no conflicts of interest.

Gilbertson RJ. Mapping cancer origins. Cell. 2011;145(1):25-29.
Malvezzi M, Santucci C, Boffetta P, et al. European cancer mortality predictions for the year 2023 with focus on lung cancer. Ann Oncol. 2023;34(4):410-419.
Siegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. CA Cancer J Clin. 2024;74(1):12-49.
Qi J, Li M, Wang L, et al. National and subnational trends in cancer burden in China, 2005-20: an analysis of national mortality surveillance data. Lancet Public Health. 2023;8(12):e943-e955.
Ness KK, Wogksch MD. Frailty and aging in cancer survivors. Transl Res. 2020;221:65-82.
Anderson AS, Martin RM, Renehan AG, et al. Cancer survivorship, excess body fatness and weight-loss intervention-where are we in 2020?. Br J Cancer. 2021;124(6):1057-1065.
Dent E, Wright ORL, Woo J, Hoogendijk EO. Malnutrition in older adults. Lancet. 2023;401(10380):951-966.
Schuetz P, Seres D, Lobo DN, Gomes F, Kaegi-Braun N, Stanga Z. Management of disease-related malnutrition for patients being treated in hospital. Lancet. 2021;398(10314):1927-1938.
Bellanti F, Lo Buglio A, Quiete S, Vendemiale G. Malnutrition in Hospitalized Old Patients: Screening and Diagnosis, Clinical Outcomes, and Management. Nutrients. 2022;14(4):910.
Bouillanne O, Morineau G, Dupont C, et al. Geriatric Nutritional Risk Index: a new index for evaluating at-risk elderly medical patients. Am J Clin Nutr. 2005;82(4):777-783.
Huo X, Wu M, Gao D, et al. Geriatric nutrition risk index in the prediction of all-cause and cardiovascular mortality in elderly hypertensive population: NHANES 1999-2016. Front Cardiovasc Med. 2023;10:1203130.
Shen X, Yang L, Gu X, Liu YY, Jiang L. Geriatric Nutrition Risk Index as a predictor of cardiovascular and all-cause mortality in older Americans with diabetes. Diabetol Metab Syndr. 2023;15(1):89.
Chai X, Chen Y, Li Y, Chi J, Guo S. Lower geriatric nutritional risk index is associated with a higher risk of all-cause mortality in patients with chronic obstructive pulmonary disease: a cohort study from the National Health and Nutrition Examination Survey 2013-2018. BMJ Open Respir Res. 2023;10(1):e001518.
Zipf G, Chiappa M, Porter KS, Ostchega Y, Lewis BG, Dostal J. National health and nutrition examination survey: plan and operations, 1999-2010. Vital Health Stat 1. 2013;(56):1-37.
Cao C, Friedenreich CM, Yang L. Association of Daily Sitting Time and Leisure-Time Physical Activity With Survival Among US Cancer Survivors. JAMA Oncol. 2022;8(3):395-403.
Dai H, Xu J. Preoperative geriatric nutritional risk index is an independent prognostic factor for postoperative survival after gallbladder cancer radical surgery. BMC Surg. 2022;22(1):133.
Komatsu M, Okazaki M, Tsuchiya K, Kawaguchi H, Nitta K. Geriatric Nutritional Risk Index Is a Simple Predictor of Mortality in Chronic Hemodialysis Patients. Blood Purif. 2015;39(4):281-287.
Qiao Y, Liu F, Peng Y, et al. Association of serum Klotho levels with cancer and cancer mortality: Evidence from National Health and Nutrition Examination Survey. Cancer Med. 2023;12(2):1922-1934.
Fulgoni VL 3rd, Drewnowski A. No Association between Low-Calorie Sweetener (LCS) Use and Overall Cancer Risk in the Nationally Representative Database in the US: Analyses of NHANES 1988-2018 Data and 2019 Public-Use Linked Mortality Files. Nutrients. 2022;14(23):4957.
Yao J, Chen X, Meng F, Cao H, Shu X. Combined influence of nutritional and inflammatory status and depressive symptoms on mortality among US cancer survivors: Findings from the NHANES. Brain Behav Immun. 2024;115:109-117.
Duan W, Xu C, Liu Q, et al. Levels of a mixture of heavy metals in blood and urine and all-cause, cardiovascular disease and cancer mortality: A population-based cohort study. Environ Pollut. 2020;263(Pt A):114630.
Liu H, Wang L, Chen C, Dong Z, Yu S. Association between Dietary Niacin Intake and Migraine among American Adults: National Health and Nutrition Examination Survey. Nutrients. 2022;14(15):3052.
Rattan P, Penrice DD, Ahn JC, et al. Inverse Association of Telomere Length With Liver Disease and Mortality in the US Population. Hepatol Commun. 2022;6(2):399-410.
Liu H, Zhang S, Gong Z, et al. Association between migraine and cardiovascular disease mortality: A prospective population-based cohort study. Headache. 2023;63(8):1109-1118.
Yamada K, Furuya R, Takita T, et al. Simplified nutritional screening tools for patients on maintenance hemodialysis. Am J Clin Nutr. 2008;87(1):106-113.
Takahashi H, Ito Y, Ishii H, et al. Geriatric nutritional risk index accurately predicts cardiovascular mortality in incident hemodialysis patients. J Cardiol. 2014;64(1):32-36.
Xie H, Tang S, Wei L, Gan J. Geriatric nutritional risk index as a predictor of complications and long-term outcomes in patients with gastrointestinal malignancy: a systematic review and meta-analysis. Cancer Cell Int. 2020;20(1):530.
Yiu CY, Liu CC, Wu JY, et al. Efficacy of the Geriatric Nutritional Risk Index for Predicting Overall Survival in Patients with Head and Neck Cancer: A Meta-Analysis. Nutrients. 2023;15(20):4348.
Shen F, Ma Y, Guo W, Li F. Prognostic Value of Geriatric Nutritional Risk Index for Patients with Non-Small Cell Lung Cancer: A Systematic Review and Meta-Analysis. Lung. 2022;200(5):661-669.
Xu R, Chen XD, Ding Z. Perioperative nutrition management for gastric cancer. Nutrition. 2022;93:111492.
Yu Z, Kong D, Peng J, Wang Z, Chen Y. Association of malnutrition with all-cause mortality in the elderly population: A 6-year cohort study. Nutr Metab Cardiovasc Dis. 2021;31(1):52-59.
Fielding RA, Landi F, Smoyer KE, Tarasenko L, Groarke J. Association of anorexia/appetite loss with malnutrition and mortality in older populations: A systematic literature review. J Cachexia Sarcopenia Muscle. 2023;14(2):706-729.
Norman K, Pichard C, Lochs H, Pirlich M. Prognostic impact of disease-related malnutrition. Clin Nutr. 2008;27(1):5-15.
Zhang X, Edwards BJ. Malnutrition in Older Adults with Cancer. Curr Oncol Rep. 2019;21(9):80.
Guy GP Jr, Yabroff KR, Ekwueme DU, Rim SH, Li R, Richardson LC. Economic Burden of Chronic Conditions Among Survivors of Cancer in the United States. J Clin Oncol. 2017;35(18):2053-2061.
Morton M, Patterson J, Sciuva J, et al. Malnutrition, sarcopenia, and cancer cachexia in gynecologic cancer. Gynecol Oncol. 2023;175:142-155.
Hunt KJ, Jaffa MA, Garrett SM, et al. Plasma Connective Tissue Growth Factor (CTGF/CCN2) Levels Predict Myocardial Infarction in the Veterans Affairs Diabetes Trial (VADT) Cohort. Diabetes Care. 2018;41(4):840-846.
Jeong K, Kim JH, Murphy JM, et al. Nuclear Focal Adhesion Kinase Controls Vascular Smooth Muscle Cell Proliferation and Neointimal Hyperplasia Through GATA4-Mediated Cyclin D1 Transcription. Circ Res. 2019;125(2):152-166.
Fearon K, Strasser F, Anker SD, et al. Definition and classification of cancer cachexia: an international consensus. Lancet Oncol. 2011;12(5):489-495.
Baracos VE, Martin L, Korc M, Guttridge DC, Fearon KCH. Cancer-associated cachexia. Nat Rev Dis Primers. 2018;4:17105.
Setiawan T, Sari IN, Wijaya YT, et al. Cancer cachexia: molecular mechanisms and treatment strategies. J Hematol Oncol. 2023;16(1):54.
Nishikawa H, Goto M, Fukunishi S, Asai A, Nishiguchi S, Higuchi K. Cancer Cachexia: Its Mechanism and Clinical Significance. Int J Mol Sci. 2021;22(16):8491.
Baazim H, Antonio-Herrera L, Bergthaler A. The interplay of immunology and cachexia in infection and cancer. Nat Rev Immunol. 2022;22(5):309-321.
Gerriets VA, MacIver NJ. Role of T cells in malnutrition and obesity. Front Immunol. 2014;5:379.
Barbeito-Andrés J, Pezzuto P, Higa LM, et al. Congenital Zika syndrome is associated with maternal protein malnutrition. Sci Adv. 2020;6(2):eaaw6284.
Schreiber RD, Old LJ, Smyth MJ. Cancer immunoediting: integrating immunity's roles in cancer suppression and promotion. Science. 2011;331(6024):1565-1570.

No competing interests reported.

Supplementary.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Association Between Geriatric Nutritional Risk Index and Mortality Outcomes in Elderly Cancer Survivors in the United States

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

1. Introduction

2. Methods

2.1. Study population

2.2. Diagnosis of cancer

2.3. The Geriatric Nutrition Risk Index

2.4. Assessment of mortality

2.5. Assessment of covariates

2.6. Statistical analysis

2.7. Standard protocol approval, registration, and patient consent

3. Results

3.1. Baseline characteristics of study participants

3.2. Associations between GNRI and mortality among cancer survivors

3.3. Subgroup analyses and sensitivity analyses

4. Discussion

5. Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1