Associations of interleukin-6 with functional trajectories in older adults with cancer: Findings from the Health, Aging, and Body Composition Study

Background: Elevated markers of inflammation, such as interleukin-6 (IL-6), are associated with aging, cancer, and functional decline. We assessed the association of pre-diagnosis IL-6 levels with post-diagnosis functional trajectories among older adults with cancer. Black and White participants experience different social structures,

Background: Elevated markers of inflammation, such as interleukin-6 (IL-6), are associated with aging, cancer, and functional decline. We assessed the association of pre-diagnosis IL-6 levels with post-diagnosis functional trajectories among older adults with cancer. Black and White participants experience different social structures, therefore we sought to understand whether these associations differ between Black and White participants. Methods: We conducted secondary analysis of the Health Aging, Body, and Composition (ABC) prospective longitudinal cohort study. Participants were recruited from 4/1997 to 6/1998. We included 179 participants with a new cancer diagnosis and IL-6 level measured within 2 years before diagnosis. Primary endpoint was functional measures (self-reported ability to walk 1/4, 20-meter gait speed). Nonparametric longitudinal models were used to cluster the trajectories; multinomial and logistic regressions to model associations. Findings: Mean age was 74 (SD 2.9); 36 % identified as Black. For self-reported functional status, we identified 3 clusters: high stable, decline, low stable. For gait speed, we identified 2 clusters: resilient, decline. The relationship between cluster trajectory and IL-6 was different between Black and White participants (p for inter-action<0.05). For gait speed, among White participants, a greater log IL-6 level was associated with greater odds of being in the decline vs. resilient cluster [Adjusted Odds Ratio (AOR): 4.31,95 % CI: 1.43,17.46]. Among Black participants, a greater log IL-6 levels were associated with lower odds of being in the decline vs. resilient cluster (AOR: 0.49, 95 % CI: 0.10, 2.08). Directionality was similar for self-reported ability to walk ¼ mile (high stable vs. low stable). Among White participants, a higher log IL-6 level was associated numerically with greater odds of being in the low stable vs. high stable cluster (AOR: 1.99, 95 % CI: 0.82, 4.85). Among Black participants, a higher log IL-6 level was associated numerically with lower odds of being in the low stable cluster vs. high stable cluster (AOR: 0.78, 95 % CI: 0.30, 2.00). Interpretation: The association between IL-6 levels and functional trajectories of older adults differed by race. Future analyses exploring stressors faces by other minoritized racial backgrounds are needed to determine the association between IL-6 and functional trajectories.
Panel: Research in context: Evidence before this study: Previous research has shown that aging is the greatest risk factor for cancer and older adults with cancer experience a higher burden of comorbidities, increasing their risk of functional decline. Race has also been shown to be associated with increased risk for functional decline. Black individuals are exposed to more chronic negative social determinants, compared to White individuals. Previous work has shown that chronic exposure to negative social determinants leads to elevated levels of inflammatory markers, such as IL-6, but studies investigating the relationship between inflammatory markers and functional decline are limited. Added value of this study: Authors of this study sought to understand the association between pre-diagnosis IL-6 levels and functional trajectories post-diagnosis in older adults with cancer, and whether these associations differed between Black and White participants with cancer. Authors decided to utilize the data from the Health, Aging and Body Composition (Health ABC) Study. The Health ACB study was a prospective longitudinal cohort study that has a high representation of Black older adults and collected inflammatory cytokines and physical function data over time. Implications of all available evidence: This work adds to the literature by providing an opportunity to study the difference in the relationships between IL-6 levels and functional trajectories between older Black and White participants with cancer. Identifying factors associated with functional decline and its trajectories may inform treatment decision making and guide development of supportive care interventions to prevent functional decline. Additionally, given the disparities in clinical outcomes for Black individuals, a better understanding of the difference in functional decline based on race will allow more equitable care to be distributed.

Introduction
Aging is the greatest risk factor for cancer development, and 60 % of new cancer cases occur in older adults (Pal et al., 2010). Compared to those without cancer, older adults with cancer have a higher burden of comorbidities, geriatric syndromes, polypharmacy, and impairments in physical function, cognition, and nutrition (Mohile et al., 2009;O'Donovan and Leech, 2020). These vulnerabilities may lead to increased susceptibility to functional decline (Kenis et al., 2017). In addition to age, racial disparities in functional decline between Black and White individuals have been identified (Moody-Ayers et al., 2005;Weuve et al., 2018). In some individuals, physical function is a modifiable vulnerability, and some older adults value their physical function over survival (Fried et al., 2002;Fusco et al., 2012). Therefore, identifying factors associated with functional decline and its trajectories may inform treatment decision making and guide supportive care interventions to prevent functional decline.
Cross-sectional studies demonstrate that elevated levels of inflammatory markers are associated with impairments in physical function in older adults (Windham et al., 2016;Brinkley et al., 2009;Cohen et al., 1997). Elevated levels of inflammatory markers are also associated with functional decline in the general and non-cancer populations (Figaro et al., 2006;McDermott et al., 2005;Dupont et al., 2021). In the US, minoritized populations often experience constrained socioeconomic conditions and disparities in social determinants of health (e.g., economic stability, education), resulting in health disadvantages compared to the majority White population. Chronic exposure to stressors and negative social determinants experienced by individuals from minoritized racial backgrounds, rather than biological differences, leads to elevated levels of inflammatory markers, specifically in Black individuals compared to those who are White (Simons et al., 2018). Studies evaluating the association of inflammatory markers with functional decline and its trajectories among older adults with cancer are limited. In addition, associations of inflammatory markers with functional decline in Black compared to White older adults with cancer have not been investigated.
Interleukin-6 (IL-6) is one of the most robust inflammatory markers associated with physical function (Adriaensen et al., 2014). Elevated levels of IL-6 are also associated with illnesses such as cardiovascular disease and osteoporosis that contribute to impairments in physical function (McDermott et al., 2005;Scheidt-Nave et al., 2001). The Health, Aging and Body Composition (Health ABC) Study was a prospective longitudinal cohort study that has a high representation of Black older adults and collected inflammatory cytokines and physical function data over time. Therefore, it provides an opportunity to study difference in the relationships between IL-6 levels and functional trajectories between older Black and White participants with cancer. Given the disparities in clinical outcomes in Black individuals, a better understanding of the difference in relationships will allow equitable care to be distributed.
We assessed the association of pre-diagnosis IL-6 levels with functional trajectories post-diagnosis in older adults with cancer, and whether these associations differed between Black and White older participants with cancer. We hypothesized that higher levels of IL-6 prior to diagnosis were associated with greater declines in functional trajectories in older adults with cancer, and older Black participants with cancer who have higher levels of IL-6 were more likely to experience greater declines in functional trajectories than older White participants with cancer.

Study design, setting, and population
This was a secondary analysis of the Health ABC Study, a prospective longitudinal cohort study that aimed to assess differences in onset of functional limitation, disability, and longevity between older men and women and between Black and White individuals (45 % of the women; 33 % of the men were Black individuals). Details of the study sample and design of the Health ABC study were previously reported (Introducing the Health ABC Study: the dynamics of health, aging, and body com-positionAccessed January 25, 2022). There were 3075 community dwelling older adults recruited to the Health ABC Study from 4/1997 to 6/1998, and follow-up continued for up to 16 years. Participants were recruited from a random sample of White Medicare beneficiaries and all age-eligible Black community residents in designated zip codes areas around Pittsburgh, Pennsylvania, and Memphis, Tennessee. Inclusion criteria were: 1) age 70-79 years; 2) no difficulty walking one-quarter of a mile or climbing 10 steps without resting; 3) no difficulty in performing activities of daily living; and 4) no active treatment for cancer in the prior three years. All participants provided informed consent. The study was approved by the institutional review boards at the University of Pittsburgh, the University of Tennessee, and the University of California, San Francisco.
In the Health ABC, self-reported functional status was collected every six months in the clinic, via home-based interviews, or via phone calls. Objective physical function, specifically 20-meter gait speed, was collected annually in certain years at study visits 1, 2, 3, 4, 5, 6, 8, 10, and 16. Blood samples for measurement of IL-6 levels were obtained at study visits 1, 2, 4, and 6. A report of a new cancer diagnosis during follow-up assessments prompted an adjudication protocol whereby pathology and cytology reports as well as supportive radiologic and laboratory data were obtained (except for non-melanoma skin cancers). For this study, we restricted our analysis to participants who received a new cancer diagnosis after enrollment into Health ABC and who had IL-6 levels measured within two years before their cancer diagnosis. Participants also had at least three functional measures within 2000 days after cancer diagnosis, with at least one being at least 1000 days following the cancer diagnosis. These criteria were selected in consideration of the resulting proportion of participants who were alive (i.e., >80 % of patients were alive at 2000 days), because a significant proportion of death in the sample would risk biasing our estimation procedure, and to reduce the extent of trajectory extrapolation beyond the time of the subject's final functional measurement (Supplemental Fig. 1).

Independent/predictor variables
Overnight fasting blood samples were collected, and IL-6 levels were analyzed from frozen stored serum at the Health ABC Core Laboratory at the University of Vermont (Penninx et al., 2003). IL-6 was measured in duplicate using enzyme-linked immunosorbent assay (ELISA) kits from R&D Systems (Minneapolis, MN). The detectable limit for IL-6 (by HS600 Quantikine kit) was 0.10 pg/mL (Cauley et al., 2007) and detection range was 0.156-17.0 pg/mL. Interassay coefficient of variation was 7 % (Visser et al., 2002). In older adults, levels of IL-6 measured at one point in time have been shown to be reliable, reproducible, and representative over extended periods of time (Macy et al., 1997).
Objective physical function was assessed using 20-meter gait speed. A higher gait speed was considered to indicate better objective physical function. Changes in gait speed of 0.05 to 0.20 m/s are considered clinically meaningful (Hass et al., 2014;Bohannon and Glenney, 2014).

Covariates
We selected covariates that might be associated with inflammatory markers and functional status. Socio-demographic variables included age, sex (female, male), educational attainment (less than high school, high school graduate, postsecondary), race (Black, White), marital status (married, never/previously married), smoking status (never, current/ past smoker), alcohol use (never, current/former), health insurance (Medicare and/or Medicaid only, other supplemental insurance) (Rooks et al., 2008), family income (<$25,000, ≥$25,000 to <$50,000, ≥$50,000), number of household members, and financial status at enrollment to Health ABC. Higher financial status was defined as yes if they endorsed two or more of the following: the amount of money they have meets their needs very well; they have money left over at the end of the month; or they own their apartment or house (Kaup et al., 2015). Clinical variables included comorbidities (diabetes, heart attack, hypertension, stroke) (Lucas et al., 2020), body mass index (BMI), cancer type (breast, gastrointestinal, genitourinary, other), and stage (limited, metastases, unknown). We used the number of comorbidities, which included diabetes, heart attack, hypertension, and stroke, as a continuous variable (0-4). We also included study site (Pittsburgh, Memphis).

Statistical analyses
Functional principal components analysis (FPCA) was performed to identify the time-dependent mean and major modes of variation in the outcomes. For both outcomes, the leading three variation functions, which explained >99 % of the variation, were retained for subsequent modeling. Smooth functional trajectories were estimated from each subject's discretely observed outcomes as the sum of the overall mean and a weighted combination of the variation functions. The weights were estimated using the best linear predictor given each subject's observed data. The estimated subject-level trajectories were subsequently clustered. For self-reported ability to walk one-quarter of a mile, K-means clustering was applied to the variation function weight vector. For the 20-meter gait speed, the shape-based elastic K-means method in the R package "fdasrvf" was applied to the trajectories (Tucker et al., 2013). The number of clusters (K) was determined using a combination of the elbow method applied to the per-K sum squared error along with manual evaluation based on clinical significance (Supplemental Fig. 2). This procedure resulted in three clusters for self-reported ability to walk one-quarter of a mile and two clusters for 20-meter gait speed. To compare the cluster mean trajectories, simultaneous confidence bands (SCBs) were formed using an asymptotic approximation under a shared covariance model. The clusters were also compared in terms of the mean decline-that is, the expected difference between the performance at diagnosis (day = 0) and at 2000 days after diagnosis. Under a Gaussian process model, we formed an unbiased estimator of this quantity, referred to as the "decline", and we created confidence intervals using the results of the FPCA along with the raw longitudinal performance measurements.
The cluster labels were used as outcomes in a regression model to investigate the association of the baseline inflammatory marker (natural log transformed IL-6 level) and race with post-diagnosis functional trajectories, adjusting for covariates. First, we included IL-6 level and race. Second, we included IL-6 level, race, and their interaction term. Third, due to significant interaction between IL-6 and race, we stratified the analyses based on race. For self-reported ability to walk one-quarter of a mile, we used a multinomial logistic regression model with all baseline covariates included. For 20-meter gait speed, we used a logistic regression model with an initial reduced set of baseline covariates (age and cancer stage) due to sample size restrictions. A stepwise forward selection procedure with an Akaike information criterion (AIC) was subsequently performed to select additional covariates to include in the model (Heinze et al., 2018).

Demographics
Since enrollment to Health ABC, 857 participants had a new cancer diagnosis (average days from enrollment to cancer diagnosis = 2256 days, SD 1506); 344 of these individuals had IL-6 levels measured within two years prior to the diagnosis. A total of 179 had at least three selfreported functional status measures, and 81 had at least three objective physical function tests after the cancer diagnosis (Supplemental Fig. 3). Tables 1 and Supplemental Table 1 show the demographics. In the larger sample (N = 179), mean age was 73.4 (SD 2.8) years, 41 % were female, and 36 % were Black. Fig. 1A and B demonstrate mean trajectories in self-reported ability to walk one-quarter of a mile and 20-meter gait speed. There was a gradual decline in both self-reported ability to walk one-quarter of a mile and in 20-meter gait speed over the first 2000 days post-cancer diagnosis. Fig. 1C-F illustrate the trajectories of individuals and cluster-means for the two outcomes. For self-reported ability to walk

IL-6 levels
In the larger sample (N = 179), mean and median times between IL-6 measurement and cancer diagnosis were 353 days (SD 192) and 340 days (range 9-725 days), respectively (Supplemental Table 2). Mean IL-6 levels were 3.49 pg/mL (SD 3.72) and 3.45 pg/mL (SD 3.42), respectively, in the samples with self-reported ability to walk onequarter of a mile (N = 179) and 20-meter gait speed (N = 81). Mean IL-6 levels prior to cancer diagnosis were higher in Black patients than in White participants (4.01 vs. 3.20 pg/mL, P = 0.30). IL-6 levels by clusters and race for both self-reported ability to walk onequarter of a mile and 20-meter gait speed are shown in Fig. 2. For self-reported ability to walk one-quarter of a mile, log IL-6 levels were higher among White in the low stable cluster compared to the high stable cluster. Among those in the low stable cluster, White participants Fig. 1. A) Mean trajectory in self-reported ability to walk one-quarter of a mile; B) mean trajectory in 20-meter gait speed; C) individual trajectories by cluster for self-reported ability to walk one-quarter of a mile; D) individual trajectories by cluster for 20-meter gait speed; E) Cluster-mean trajectories for self-reported ability to walk one-quarter of a mile; and F) cluster-mean trajectories for 20-meter gait speed Abbreviations: BMI, body mass index; GI, gastrointestinal, GU, genitourinary; SD, standard deviation.
had higher log-IL-6 levels compared to Black participants. For 20-meter gait speed, log IL-6 levels were higher among white in the decline compared to the resilient cluster. Among those in the resilient cluster, Black participants higher log-IL-6 levels compared to White participants.

Associations of IL-6 with clusters for self-reported ability to walk onequarter of a mile
On multinomial regression for self-reported ability to walk one- Abbreviations: IL-6, interleukin-6; CI, confidence interval. All models adjusted for age, sex, site, marital status, education, number of comorbidities, cancer stage, smoking, alcohol, income, financial status, health insurance, number of household members, and body mass index.

Table 3
Logistic regression models for 20-meter gait speed. Abbreviations: IL-6, interleukin-6; CI, confidence interval. All models adjusted for age, sex, education, and number of comorbidities.

Fig. 2.
Log IL-6 levels by cluster groups and race for A) self-reported ability to walk one-quarter of a mile and B) 20-meter gait speed.
quarter of a mile, there was a significant interaction between race and log IL-6 level (p = 0.02) when comparing the low vs. high stable clusters (  Table 2). Among Black participants, a higher log IL-6 level was associated numerically with lower odds of being in the low stable cluster vs. high stable cluster (AOR: 0.78, 95 % CI: 0.30, 2.00) ( Table 2).

Associations of IL-6 with clusters for 20-meter gait speed
On logistic regression for 20-meter gait speed, there was a significant interaction between race and log IL-6 level (p = 0.01; Table 3 and Supplemental Fig. 4B). Among White participants, a greater log IL-6 level was associated with greater odds of being in the decline vs. resilient cluster (AOR: 4.31, 95 % CI: 1.43, 17.46) ( Table 3). Among Black participants, a greater log IL-6 levels were associated with lower odds of being in the decline cluster vs. resilient cluster (AOR: 0.49, 95 % CI: 0.10, 2.08) ( Table 3), but this was not statistically significant.

Discussion
We identified distinct functional trajectories that are clinically meaningful within 2000 days after cancer diagnosis in older adults with cancer. For both self-reported functional status and objective physical function, the relationship between cluster trajectory and IL-6 was different between Black and White participants. For self-reported functional status, the association between higher log IL-6 levels and greater odds of belonging to the low stable vs. high stable cluster was stronger in White than Black participants, albeit not statistically significant. For objective physical function, the association between higher log IL-6 levels and greater odds of belonging to the decline vs. resilient cluster was stronger in White patients than for Black participants.
Resilience refers to the process of adapting well in the face of adversity or significant sources of stress, or simply the ability to recover to baseline (Newman, 2005). While resilience is more commonly described in psychosocial literature, physical resilience has emerged as an important construct (Whitson et al., 2016). Identifying factors associated with physical resilience or functional decline can guide treatment discussions and inform supportive care interventions such as exercise programs to mitigate decline (Buffart et al., 2017). Previous studies have demonstrated that higher levels of pro-inflammatory cytokines are associated with functional declines and disability in the general geriatric and geriatric oncology populations (Penninx et al., 2003;Hass et al., 2014). Among White participants in our study, those with greater levels of IL-6 were more likely in the decline cluster (vs. resilient) for objective physical function. They were also more likely to be in the low stable cluster (vs. high stable) for self-reported functional status, albeit this was not statistically significant. Our findings reinforce the potential utility of measuring IL-6 in clinical practice, despite IL-6 being measured within two years prior to the cancer diagnosis and not shortly before or after cancer diagnosis. Specifically, our findings suggest the need to investigate interventions that may reduce IL-6 levels in racial subgroups in order to mitigate functional decline. In future studies, we will consider more sophisticated models to detect clinically relevant cutoffs for baseline IL-6 and functional decline.
Studies have shown differences in the levels of inflammatory cytokines among Black compared to White individuals. Black individuals (vs. White) and those with constrained socioeconomic conditions exhibit higher levels of inflammatory markers (Lam et al., 2021). In our study, absolute IL-6 levels were higher in Black compared to White participants. This may be due to social determinants of health instead of biologic differences (McClendon et al., 2021). IL-6 levels were less associated with functional trajectories in Black vs. White participants. While there were some differences in socio-demographics (e.g., personal family income, health insurance status) and health behaviors (e.g., smoking and alcohol consumption) in Black compared to White participants, these covariates were not significant in the models evaluating the predictive potential of IL-6 on functional trajectories. Future studies are needed to better understand the relationships of inflammatory cytokines and outcomes. In order to better understand the differences in the predictive potential of inflammatory cytokines on physical function by an individual's race, careful consideration needs to be paid to social determinants of health. As we aim to provide equitable care and reduce disparities in clinical outcomes for Black individuals with cancer, we need to understand how social determinants of health, such as poverty, education, access to health care, and social isolation contribute to disparities in clinical outcomes. Our study highlights the need to consider racial differences when developing models to predict outcomes, and that a "one size fits all" approach should not be used when using biological markers to inform care.
Our study has several strengths. A reasonable proportion of our sample were Black participants (around one-third), thus allowing for racial comparisons. There were multiple years of functional measures, which allowed examination of functional trajectories over time. Our study has limitations. First, our sample was small and may not be generalizable to all individuals diagnosed with cancer. We included a heterogeneous group of older adults with cancer, and we do not have cancer treatment information. Second, IL-6 levels were not measured within a consistent time frame before a cancer diagnosis. However, we did not find statistically significant differences in time between IL-6 levels and cancer diagnosis between White and Black participants. Physical function outcomes were sparse and measured irregularly. Third, we acknowledge that inflammatory markers measured shortly before or at cancer diagnosis are likely more meaningful. Inflammatory markers can fluctuate significantly in the context of acute illnesses (e.g., hospitalization) which could not be accounted for in this study. Fourth, while the Health ABC collected information regarding individual's social determinants of health, it did not provide insight into the allostatic load of chronic life stressors that influence levels of inflammatory markers such as IL-6. For example, allostatic load considers daily events that may lead to health-related consequences such as poor sleep hygiene or unhealthy diet (Guidi et al., 2021). Fifth, the Health ABC study only included White and Black individuals. Therefore, further studies investigating the effect of social stressors experienced by other minoritized races on IL-6 and functional decline are warranted. Nonetheless, our study provides rationale for a future study to assess the longitudinal association of IL-6 levels and functional trajectories.
In conclusion, we found the association between IL-6 levels and functional trajectories of older adults with a new cancer diagnosis differed by race. Further studies are needed to examine reasons for these racial differences and to determine whether interventions to reduce IL-6 levels might improve functional trajectories in the racial subgroups.

Funding
Dr. Loh is supported by the National Cancer Institute (NCI; R00CA237744), Conquer Cancer American Society of Clinical Oncology and Walther Cancer Foundation Career Development Award, and the Wilmot Research Fellowship Award. Dr. Isom is supported by the NCI (P30CA012197). Drs. Houston and Kritchevsky are supported by the Wake Forest University Claude D. Pepper Older Americans Independence Center (P30AG021332). Dr. Klepin is supported by the Wake Forest University Claude D. Pepper Older Americans Independence Center (P30AG021332), NCI (P30CA012197), and National Institute of Aging (NIA; R33AG059206). This research was supported by NIA Contracts N01-AG-6-2101; N01-AG-6-2103; N01-AG-6-2106; NIA grant R01-AG028050, and National Institute of Nursing Research grant R01-NR012459. This research was funded in part by the Intramural Research Program of the NIA.