Abstract
Background
Few studies have evaluated the independent effect of a cancer diagnosis on vulnerability and frailty, which have been associated with adverse health outcomes in older adults.
Methods
We used data in the 2003 Medicare Current Beneficiary Survey from a nationally representative sample of 12 480 community-dwelling elders. Multivariable logistic regression models were used to evaluate whether cancer was independently associated with vulnerability and frailty. Measures of vulnerability and frailty included disability, geriatric syndromes, self-rated health, and scores on two assessment tools for elderly cancer patients—the Vulnerable Elders Survey-13 (VES-13) and the Balducci frailty criteria. All statistical tests were two-sided.
Results
Diagnosis of a non-skin cancer was reported by 18.8% of the respondents. Compared with respondents without a cancer history, respondents with a personal history of cancer had a statistically significantly higher prevalence of limitations in activities of daily living (31.9% vs 26.9%), limitations in instrumental activities of daily living (49.5% vs 42.3%), geriatric syndromes (60.8% vs 53.9%), low self-rated health (27.4% vs 20.9%), score of 3 or higher on the VES-13 (45.8% vs 39.5%), and satisfying criteria for frailty as defined by Balducci (79.6% vs 73.4%) (P < .001 for all characteristics). After adjustment for confounders, a cancer diagnosis was found to be associated with low self-rated health (adjusted odds ratio [OR] = 1.46, 95% confidence interval [CI] = 1.30 to 1.64; relative risk [RR] = 1.33), limitations in activities of daily living (adjusted OR = 1.19, 95% CI = 1.06 to 1.33; RR = 1.13), limitations in instrumental activities of daily living (adjusted OR = 1.25, 95% CI = 1.13 to 1.38; RR = 1.13), a geriatric syndrome (adjusted OR = 1.27, 95% CI = 1.15 to 1.41; RR = 1.11), VES-13 score of 3 or higher (adjusted OR = 1.26, 95% CI = 1.13 to 1.41; RR = 1.14), and frailty (adjusted OR = 1.46, 95% CI = 1.29 to 1.65; RR = 1.09) as defined by Balducci criteria.
Conclusion
Diagnosis of a non-skin cancer was associated with increased levels of having disability, having geriatric syndromes, and meeting criteria for vulnerability and frailty.
CONTEXT AND CAVEATS
Prior knowledge
Relatively few studies have evaluated the relationship between a cancer diagnosis in older adults and their vulnerability and frailty, which have been associated with adverse health outcomes in this population.
Study design
Cross-sectional study that used data from the 2003 Medicare Current Beneficiary Survey from a nationally representative sample of 12 480 community-dwelling elders. Statistical models were used to evaluate whether cancer was independently associated with vulnerability and frailty.
Contribution
Older adults with a diagnosis of a non-skin cancer were more likely to have more disabilities and geriatric syndromes and to meet criteria for vulnerability and frailty.
Implications
Assessment of underlying vulnerability and frailty should aid in the design of future clinical trials that are directed not only at improving survival but also at maintaining function, enhancing quality of life, and preventing geriatric syndromes.
Limitations
The cohort of cancer patients was heterogeneous. It could not be determined whether cancer or its treatment directly caused vulnerability or frailty in older patients. The number of patients with some cancer subtypes was small. Geriatric syndromes were self-reported. The assessment tools used in this study have not been validated in independent cohorts of older adults with cancer.
From the Editors
As the population ages, the numbers of vulnerable and frail older persons will increase dramatically (1). Increased numbers of vulnerable and frail elders will pose a real challenge to oncologists because by the year 2030, persons who are older than 65 years are projected to make up 70% of cancer patients and have 65% of cancer deaths (2). Although the fit elderly who receive cancer treatment appear to receive benefits that are similar to those in the younger population (3–5), age disparities in cancer care may be attributed to the difficulties in extrapolation of clinical trial results to community practice (6,7).
Older patients are a very heterogeneous group with respect to underlying health status. The spectrum of impairment ranges from those who are independent to those who are at moderate risk of health deterioration (ie, vulnerable or prefrail) and to those who are at high risk of functional decline or mortality (ie, frail) (1,8–10). Several clinical characteristics have synergistic influences on vulnerability and frailty in the elderly, including disability, comorbidity, and geriatric syndromes. Fried et al. (11) have proposed separate but interrelated definitions of these concepts. They defined disability as dependency in performing tasks that allow for self-care and living in the community, comorbidity as the concurrent presence of two or more chronic medical illnesses, frailty as a “state of high vulnerability for adverse health outcomes” caused by decreased physiological reserve leading to a loss of homeostatic capacity in the face of physiological stress, and geriatric syndromes (eg, dementia, depression, incontinence, or falls) as observable manifestations of impending or established frailty. Although the definitions of vulnerability and frailty are evolving, one overarching theme is an increasing risk for adverse health outcomes, such as functional or physical decline and/or mortality.
The National Comprehensive Cancer Network guidelines for older adults advocate the assessment of factors that affect vulnerability and frailty in all older patients with cancer (8,9,12,13). A clinical tool for recognizing frailty that was proposed by Balducci and Extermann (1,9,12) combines selected risk factors from several components of a comprehensive geriatric assessment, including age older than 85 years, dependence in one or more activities of daily living, the presence of three or more comorbid conditions, and the presence of one or more geriatric syndrome(s). Another assessment tool is the Vulnerable Elders Survey-13 (VES-13) (10,14), a self-administered survey derived from the 1993–1995 Medicare Current Beneficiary Survey. In the national sample of elders from the Medicare Current Beneficiary Survey, higher scores predicted an increasing risk for functional decline and/or death (15). A score of 3 or higher on the VES-13 identified 32% of individuals as vulnerable and was associated with a greater risk of death or functional decline for 2 years than a score of less than 3. The frailty criteria as proposed by Balducci (16–19) and the VES-13 (20,21) have been used to help with cancer patient selection, risk stratification, and evaluation of toxic effects.
Only a few studies (18,22–24) to date have evaluated the prevalence of vulnerability and frailty in cancer patients. Although the prevalence was typically noted to be high, few of these studies specifically included a noncancer comparison group. We used a nationally representative sample of older patients who provided detailed information on the 2003 Medicare Current Beneficiary Survey to compare the prevalence of disability, geriatric syndromes, poor self-rated health, and vulnerability and frailty between those with and without a personal history of cancer and to estimate the independent association of a cancer diagnosis with these predictors of poor health outcomes. In addition, we derived VES-13 scores, a validated measure of vulnerability that, to our knowledge, has not yet been studied in a nationally representative population-based sample of older cancer patients, from data in the 2003 Medicare Current Beneficiary Survey.
Participants and Methods
Data Source
We used cross-sectional data from the 2003 Medicare Current Beneficiary Survey, a nationally representative in-person survey of randomly sampled Medicare beneficiaries that was conducted by the Federal Centers for Medicare and Medicaid Services. Participants completed a baseline interview and three follow-up interviews per year during a 4-year period. New participants enter into a longitudinal cohort each year. Because of the cross-sectional nature of this study, the data used were either from a baseline or from a follow-up survey depending on when the respondent entered into the longitudinal cohort. The data in this analysis included persons who entered into their longitudinal cohort and completed their baseline survey in the years 2000 through 2003. Data were collected through personal interviews of the beneficiary or a proxy chosen by the beneficiary if he or she was physically or mentally unable to do the interview. On average, approximately 12% of the community interviews in each round were conducted by proxy (25). The “Access to Care” files contain information collected in the fall interview (September through December 2003) and include information on beneficiaries’ demographics, socioeconomic status, and indicators of health status and functional status. Interviewees were requested to have their medical records on hand at the time of the interview.
Study Participants
The participants in this study were drawn from respondents to the year 2003 Medicare Current Beneficiary Survey, Access to Care, by Centers for Medicare and Medicaid Services from its Medicare enrollment file. Respondents for the Medicare Current Beneficiary Survey were selected from the Medicare enrollment file to be representative of the Medicare population as a whole. The oldest group (ie, those 85 years or older) were oversampled to permit a more detailed analysis of this subpopulation. Participants were selected by using a stratified, multistage, area probability sample design. The response rates for the Medicare Current Beneficiary Survey were more than 80% (25).
For this study, we restricted our analysis to 12 480 community-dwelling Medicare beneficiaries who were aged 65 years or older. The sample of patients with a history of cancer was created by including those who responded affirmatively to the following question: “Has a doctor ever told you that you had any kind of cancer, malignancy, or tumor other than skin cancer?” The prevalence of a cancer history among respondents was 18.8% (n = 2349).
Dependent Variables
Variables associated with adverse health outcomes in community-dwelling older adults were treated as dichotomous dependent measures. These variables included the presence of limitations in activities of daily living (26), limitations in instrumental activities of daily living (ie, the ability of a person to live independently in community) (27), low self-reported health, a geriatric syndrome, vulnerability (as measured by the VES-13) (10), and frailty as defined by Balducci (1,16–19). All of these outcome variables are associated with adverse health outcomes such as mortality in community-dwelling older adults (24,28–33). As in the geriatric literature (10,24,28,34,35), we chose to dichotomize the outcome scores at the cutoff point for each measure that has been found previously to be associated with a higher risk of morbidity and mortality in community-dwelling older adults. Scores for geriatric assessment measures have generally been dichotomized at a clinically meaningful cutoff point for ease of clinical use. By dichotomizing scores, we could more directly compare results from our study with previously published risk scores for each outcome.
We defined activities of daily living, instrumental activities of daily living, and self-reported health categories essentially as described previously (35,36). Individuals were classified as being limited in an activity of daily living if they had “a lot of difficulty” or were unable to do any of the following basic activities: bathing, eating, walking, using the toilet, or transferring in and out of a bed or a chair. Individuals were classified as being limited in an instrumental activity of daily living if they had a lot of difficulty in doing or were unable to do any of the following activities: using the telephone, doing light housework, doing heavy housework, preparing meals, shopping, or managing money (37). The self-rated health status was incorporated as a dichotomous measure (fair or poor vs good, very good, or excellent).
Self-reported history of dementia or memory loss, depression, falls, incontinence, and osteoporosis were considered geriatric syndromes (18,38). Geriatric syndromes were either captured within the Medicare Current Beneficiary Survey as yes or no or were ordinal responses (severity of symptom). The definitions of the geriatric syndromes that we adopted are consistent with those described previously (22,39,40). For the geriatric syndromes that were captured as dichotomous variables (yes or no), subjects were included as having the geriatric syndrome if they answered yes. For the ordinal variables, only respondents reporting the most severe symptoms that have been associated with adverse outcomes were included as having that syndrome (18,22,38,39). Respondents were considered to have dementia if their doctors told them that they had this diagnosis. Individuals were classified as having memory loss if they reported that memory loss interfered with daily activities. Dementia and memory loss were included as one variable. Patients were considered to have osteoporosis if they were told by their doctors that they had “soft or fragile bones” or had experienced a recent hip fracture. Depression was defined as feeling “sad, blue, or depressed” most or all of the time. Older adults who reported falling at least once within the year before survey administration were considered to have the geriatric syndrome falls. Urinary incontinence was defined as loss of control of urine at least once or more per week over the previous year.
As in the derivation studies for the VES-13 screening tool, “vulnerability” was defined as a score of 3 or higher on the VES-13 (10). The VES-13 score was calculated from age (1 point for age 75–84 years and 3 points for age 85 years or older), self-rated heath status (1 point for fair or poor), difficulty with physical activities (1 point for difficulty with each of six physical activities, with a maximum of 2 points), and difficulties with functional activities (maximum of 4 points for presence of any functional limitation). Subjects were considered to have a difficulty with functional activity if they responded that they had difficulty in doing the activity or did not do the activity because of health reasons, which was the same as the way that the activities of daily living and instrumental activities of daily living limitations were assessed. Participants were considered to be “frail” if they had one of the following characteristics: aged 85 years or older, a limitation in an activity of daily living, any geriatric syndrome, or three or more chronic medical conditions (16–19).
Independent Variables
The main independent variable of interest was self-reported personal history of cancer. We also recorded the type of cancer. The most common types of cancer were assessed individually. We included all other cancers in an “other” category that included cancers with small sample sizes and cancers whose type was not specified by the respondent. Additional beneficiary-level covariates included age, sex, race or ethnicity, income, and education. Chronic medical conditions such as self-reported histories of heart disease, pulmonary disease, and neurological diseases were also assessed from the survey data within the Medicare Current Beneficiary Survey (41,42). Although the severity of comorbidity was not available, the chronic diseases were selected by their high prevalence in older adults and measurable impact on mortality and disability (43). By use of methods essentially as described previously (18,43), variables that could influence cancer care, such as poor vision or hearing, were also considered as comorbidities. Table 1 lists the variables included in the comorbidity count. Cancer was excluded from the comorbidity score.
Table 1.
Characteristics of elderly Medicare beneficiaries, including 2349 with cancer and 10 131 without cancer*
| Characteristic | Cancer group, No. (%)† | Noncancer group, No. (%)† | P‡ |
| Age, y | <.001 | ||
| 65–69 | 388 (19.1) | 2283 (26.2) | |
| 70–74 | 484 (25.2) | 2235 (25.9) | |
| 75–79 | 556 (25.3) | 2086 (21.5) | |
| 80–84 | 528 (18.2) | 1932 (15.0) | |
| ≥85 | 393 (12.3) | 1595 (11.3) | |
| Sex | .819 | ||
| Male | 1017 (43.3) | 4365 (43.0) | |
| Female | 1332 (56.7) | 5766 (57.0) | |
| Race | <.001 | ||
| Non-Hispanic white | 1994 (84.4) | 8098 (79.7) | |
| Non-Hispanic African American | 162 (7.4) | 799 (7.9) | |
| Hispanics | 112 (4.6) | 789 (7.9) | |
| Other | 81 (3.6) | 445 (4.6) | |
| Marital status | .403 | ||
| Married | 1269 (56.3) | 5356 (55.3) | |
| Other | 1080 (43.7) | 4775 (44.7) | |
| Income, $ | .083 | ||
| <25 000 | 1331 (54.7) | 6023 (56.9) | |
| 25 000–50 000 | 749 (33.2) | 3101 (32.4) | |
| >50 000 | 269 (12.1) | 1007 (10.7) | |
| Education | <.001 | ||
| Less than ninth grade | 297 (11.5) | 1631 (14.8) | |
| Ninth grade to high school | 1075 (45.9) | 4567 (45.1) | |
| Some college | 514 (22.5) | 2094 (21.0) | |
| Associate degree or higher | 463 (20.1) | 1839 (19.1) | |
| Comorbidity | |||
| Vision | 226 (8.9) | 898 (7.8) | .119 |
| Hearing | 225 (8.7) | 737 (6.5) | <.001 |
| Hypertension | 1485 (63.0) | 6044 (58.5) | <.001 |
| Coronary artery disease | 539 (22.0) | 2132 (19.92) | .050 |
| Congestive heart failure | 226 (8.7) | 711 (6.3) | <.001 |
| Valvular heart disease | 245 (9.9) | 934 (8.8) | .075 |
| Arrhythmia | 561 (23.1) | 2102 (19.7) | <.001 |
| Other heart condition | 261 (10.7) | 1018 (9.6) | .142 |
| Cerebrovascular disease | 334 (13.8) | 1327 (12.3) | .058 |
| Arthritis | 1494 (62.2) | 5996 (57.6) | <.001 |
| Mental disorder | 358 (15.6) | 1278 (12.4) | <.001 |
| Parkinson disease | 30 (1.1) | 154 (1.4) | .310 |
| Emphysema, asthma, or COPD | 399 (16.7) | 1376 (13.4) | <.001 |
| Diabetes | 501 (21.2) | 2000 (19.8) | .233 |
| No. of comorbidities | <.001 | ||
| 0 | 164 (7.4) | 1034 (11.1) | |
| 1 | 393 (17.9) | 2032 (21.1) | |
| 2 | 566 (24.2) | 2342 (23.2) | |
| ≥3 | 1226 (50.5) | 4723 (44.5) |
COPD = chronic obstructive pulmonary disease.
The weighted prevalence was based on a weighting algorithm that is provided by the Medicare Current Beneficiary Survey (25).
Two-sided χ2 tests. All statistical tests were two-sided.
Statistical Analysis
The goals of our analyses were to assess the independent relationships of cancer diagnosis to other characteristics that have been associated with long-term adverse outcomes in community-dwelling older adults. Baseline differences in demographics such as age, sex, race, socioeconomic factors, and comorbidity between the noncancer and cancer groups were evaluated by use of the χ2 test of proportions. We then compared the proportion of older cancer patients reporting limitations in activities of daily living and/or in instrumental activities of daily living, fair or poor self-rated health, and the presence of geriatric syndromes with that of all other Medicare beneficiaries. We also compared the scores for the VES-13 and the Balducci frailty criteria between subjects with and without a personal cancer history. We assessed the statistical significance of these differences by use of the χ2 test of proportions.
Multivariable logistic regression was used to determine the association of cancer with each dependent variable in our study (ie, limitations in activities of daily living, limitations in instrumental activities of daily living, low self-rated health, vulnerability as defined by VES-13 score of 3 or higher, and frailty as defined by Balducci criteria). Each outcome was examined in a separate model. Models were also generated that included only the cancer-specific subtypes. Adjusted odds ratios (ORs) and their 95% confidence intervals (95% CIs) provide an indication of the independent association of each variable with the likelihood of the outcome after adjustment for the effects of all other variables in the model. In addition, because the adjusted odds ratio obtained from the logistic regression may exaggerate the risk association when the prevalence rate is high, the odds ratio estimates were converted to relative risks (RRs) to better capture the magnitude of the association between cancer and impairment characteristics and to better interpret the data for clinical use (44). The relative risk estimates depict the probability of having the outcome variable among the cancer group compared with the noncancer group. We have presented both the odds ratio estimates with confidence intervals and the relative risk estimates. We used the c-statistic, which makes use of receiver operating characteristic curve analysis, to evaluate the logistic regression models. In this analysis, the power of the model’s predicted values to discriminate between positive and negative outcomes is quantified by the area under the receiver operating characteristic curve. The area under this curve, which is the c-statistic (or concordance index), is a value that varies from 0.5 (discriminating power not better than chance) to 1.0 (perfect discriminating power). It can be interpreted as the percentage of all possible pairs of predictions in which the model assigns a higher probability to a positive outcome than to a negative outcome.
The VES-13 and Balducci frailty scores were examined as dichotomous dependent variables. Because comorbidity (ie, three or more clinically significant comorbidities) is a category within the frailty criteria, comorbidities were excluded from the adjusted models for frailty score. In addition, because age is incorporated into the scoring procedures for the VES-13 and Balducci frailty criteria, age was excluded from these multivariable analyses.
Because the Medicare Current Beneficiary Survey uses a stratified, multistage sampling scheme and oversamples certain population groups, each person had an unequal probability of being included in the survey. To obtain nationally representative population estimates, we conducted the analyses with SAS Survey Procedures (version 9.13; SAS Institute, Inc, Cary, NC) and used the Medicare Current Beneficiary Survey cross-sectional weight to adjust for the complex multistage sample design and the number of nonresponses. Because we were interested in whether cancer was independently associated with each of the individual characteristics (rather than if cancer subjects as a population were inherently different from noncancer subjects), no formal adjustment was made for multiple comparisons. However, the maximum P value for statistical significance was set conservatively at .01. All statistical tests were two-sided. The large sample size yielded high statistical power to make comparisons between noncancer and cancer subjects and to analyze outcomes in multivariable models (45). The University of Rochester's Institutional Review Board approved these analyses with exempt status.
Results
A personal history of non-skin cancer was reported by 2349 (18.8%) of the 12 480 respondents to the Medicare Current Beneficiary Survey. The weighted unadjusted demographic and clinical characteristics in the cancer and noncancer groups are shown in Table 1. Respondents with a personal cancer history were statistically significantly older (mean age 76.2 vs 75.2 years; P < .001), were statistically significantly more likely to be non-Hispanic white (84.4% vs 79.7%; P < .001), and to have received some college and education to the associate degree level or higher (42.6% vs 40.1%; P < .001). A statistically significantly higher proportion of cancer patients had three or more comorbid conditions compared with those without a personal history of cancer (50.5% vs 44.5%; P < .001). Among the elderly cancer patients, colon, breast, and prostate cancers were the most prevalent diagnoses (Table 2).
Table 2.
Cancer prevalence and population estimates*
| Cancer site (n = 2349 patients) | No. of patients | Cancer prevalence, %† (95% CI) | US population estimates (millions)† |
| Lung | 119 | 0.99 (0.81 to 1.18) | 0.31 |
| Colon | 328 | 2.43 (2.15 to 2.71) | 0.77 |
| Breast | 601 | 4.72 (4.31 to 5.13) | 1.49 |
| Cervical or uterine | 273 | 2.17 (1.90 to 2.44) | 0.68 |
| Prostate | 525 | 4.11 (3.73 to 4.50) | 1.30 |
| Bladder | 121 | 0.91 (0.71 to 1.11) | 0.29 |
| Ovarian | 84 | 0.68 (0.53 to 0.82) | 0.21 |
| Other‡ | 604 | 4.84 (4.38 to 5.30) | 1.53 |
| Total | 2349 | 18.45 (17.68 to 19.21) | 5.82 |
CI = confidence interval.
The Medicare Current Beneficiary Survey weighting algorithm was used to calculate the cancer prevalence (both by percentage and by US population estimate of cancer) in the 2003 US population (25).
Other includes cancers with a small sample size or no specified site provided.
We calculated unadjusted differences in functional limitations (ie, activities of daily living and instrumental activities of daily living), self-rated health status, prevalence of geriatric syndromes, and prevalence of vulnerability and frailty between the cancer and noncancer groups by use of the VES-13 and Balducci criteria, respectively (Table 3). The cancer group had statistically significantly higher proportions of all five characteristics than the noncancer group (for all characteristics, P < .001). The cancer group had a statistically significantly higher prevalence of activity of daily living limitations (cancer group = 31.9%, 95% CI = 29.9% to 33.9%, and noncancer group = 26.9%, 95% CI = 25.8% to 28.1%; P < .001) and instrumental activity of daily living limitations (cancer group = 49.5%, 95% CI = 47.2% to 51.7%, and noncancer group = 42.3%, 95% CI = 41.0% to 43.7%; P < .001) than those without cancer. More respondents with a personal history of cancer reported poor or fair self-rated health than noncancer participants (cancer group = 27.4%, 95% CI = 25.4% to 29.4%, and noncancer group = 20.9%, 95% CI = 20.0% to 21.8%; P < .001). The overall prevalence of geriatric syndromes was high in both cancer (60.8%, 95% CI = 58.5% to 63.1%) and noncancer groups (53.9%, 95% CI = 52.8% to 55.1%), but it was statistically significantly higher in the cancer group (P < .001). Among the geriatric syndromes evaluated, only dementia or memory loss was not statistically significantly higher in cancer patients than in noncancer participants. Vulnerability as measured by both mean VES-13 score (cancer group = 3.33, 95% CI = 3.20 to 3.47, and noncancer group = 2.86, 95% CI = 2.78 to 2.94; P < .001) and prevalence of those with a VES-13 score of 3 or higher (cancer group = 45.8%, 95% CI = 43.6% to 48.0%, and noncancer group = 39.5%, 95% CI = 38.4% to 40.7%; P < .001) was statistically significantly higher in the cancer group than in the noncancer group. The prevalence of those subjects with frailty by the Balducci criteria was statistically significantly higher in the cancer group than in the noncancer group (cancer group = 79.6%, 95% CI = 77.9% to 81.4%, and noncancer group = 73.4%, 95% CI = 72.4% to 74.4%; P < .001).
Table 3.
Prevalence of impairment characteristics in the cancer group (N = 2349) and the noncancer group (N = 10 131)*
| Characteristic | Cancer group, No. (%)† | Noncancer group, No. (%)† | P‡ |
| Functional limitations | |||
| ADL limitations | 794 (31.9) | 2959 (26.9) | <.001 |
| Bathing or showering | 335 (13.3) | 1172 (10.2) | <.001 |
| Dressing | 212 (8.4) | 728 (6.4) | <.001 |
| Eating | 81 (3.3) | 271 (2.5) | .033 |
| Getting in or out of bed or chair | 352 (14.2) | 1369 (12.5) | .022 |
| Walking | 671 (27.0) | 2570 (23.2) | <.001 |
| Using toilet | 146 (5.7) | 586 (5.3) | .373 |
| IADL limitations | 1220 (49.5) | 4606 (42.3) | <.001 |
| Using telephone | 218 (8.2) | 812 (6.9) | .023 |
| Doing light housework | 424 (17.2) | 1562 (13.9) | <.001 |
| Doing heavy housework | 1083 (44.0) | 4035 (36.9) | <.001 |
| Preparing meals | 409 (15.9) | 1540 (13.5) | .003 |
| Shopping | 443 (17.2) | 1716 (14.9) | .007 |
| Managing money | 266 (10.0) | 1086 (9.5) | .434 |
| Geriatric syndromes | 1456 (60.8) | 5661 (53.9) | <.001 |
| Dementia and memory loss | 288 (11.5) | 1190 (10.5) | .181 |
| Depression | 614 (26.1) | 2481 (23.8) | .039 |
| Falls | 633 (25.9) | 2288 (21.6) | <.001 |
| Incontinence | 375 (15.6) | 1220 (11.1) | <.001 |
| Osteoporosis | 593 (24.3) | 2103 (19.8) | <.001 |
| Vulnerability or frailty | |||
| Vulnerability (VES-13 ≥ 3) | 1184 (45.8) | 4513 (39.5) | <.001 |
| Frailty§ | 1910 (79.6) | 7722 (73.4) | <.001 |
| Poor or fair self-rated health status | 661 (27.4) | 2191 (20.9) | <.001 |
ADL = activity of daily living; IADL = instrumental activity of daily living; VES-13 = Vulnerable Elders Survey-13.
The weighted prevalence was based on a weighting algorithm that is provided by the Medicare Current Beneficiary Survey (25).
Two-sided χ2 tests. All statistical tests were two-sided.
Individuals were considered to be frail who met any of the following criteria: aged 85 years and older, one or more ADL limitations, presence of a geriatric syndrome(s), and/or three or more chronic medical conditions.
We used multivariable analyses to evaluate the independent association of a cancer diagnosis with functional limitations (Table 4), geriatric syndromes, and low self-rated health (Table 5), as well as being vulnerable or frail (Table 6). The c-statistics for the models that evaluated the association of cancer with the presence of limitations in activities of daily living, the presence of limitations in instrumental activities of daily living, the presence of a geriatric syndrome, scoring 3 or more on the VES-13, and meeting Balducci criteria for frailty ranged from 0.65 to 0.74, indicating excellent model prediction of the dichotomous outcomes (46). After adjustment for known potential clinical and demographic confounders, a cancer diagnosis was statistically significantly associated with a limitation in activities of daily living (adjusted OR = 1.19, 95% CI = 1.06 to 1.33; P = .002; RR = 1.13), a limitation in instrumental activities of daily living (adjusted OR = 1.25, 95% CI = 1.13 to 1.38; P < .001; RR = 1.13), a geriatric syndrome (adjusted OR = 1.27, 95% CI = 1.15 to 1.41; P < .001; RR = 1.11), and poor or fair self-rated health (adjusted OR = 1.46, 95% CI = 1.30 to 1.64; P < .001; RR = 1.33). In adjusted analyses that used cancer subtype instead of any cancer as the independent variable, a diagnosis of ovarian or “other” cancer subtype was associated with statistically significantly increased risk of limitations in activities of daily living (P = .009 for ovarian and P < .001 for “other” cancer subtype). A diagnosis of lung, ovarian, or “other” cancer subtype was associated with statistically significantly increased risk of limitations in instrumental activities of daily living compared with no cancer diagnosis (P = .003 for lung, P = .007 for ovarian, and P < .001 for “other” cancer subtype). A diagnosis of prostate cancer was statistically significantly associated with increased risk of a geriatric syndrome compared with no cancer diagnosis (P = .004). A diagnosis of lung cancer or “other” cancer subtype was statistically significantly associated with risk of low self-rated health compared with no cancer diagnosis (P < .001 for lung and for “other” cancer subtype).
Table 4.
Associations between a cancer diagnosis and risk of functional limitation among Medicare beneficiaries (N = 12 480)*
| ADL limitation |
IADL limitation |
|||||
| Independent variable | OR (95% CI) | P | RR | OR (95% CI) | P | RR |
| Age, y | ||||||
| 65–69 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| 70–74 | 1.04 (0.88 to 1.22) | .665 | 1.03 | 1.09 (0.96 to 1.23) | .190 | 1.05 |
| 75–79 | 1.23 (1.06 to 1.43) | .007 | 1.16 | 1.43 (1.25 to 1.63) | <.001 | 1.21 |
| 80–84 | 1.74 (1.51 to 2.02) | <.001 | 1.45 | 2.28 (2.00 to 2.60) | <.001 | 1.48 |
| ≥85 | 2.90 (2.47 to 3.39) | <.001 | 1.92 | 4.35 (3.70 to 5.11) | <.001 | 1.80 |
| Sex | ||||||
| Male | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Female | 1.28 (1.15 to 1.42) | <.001 | 1.19 | 1.31 (1.18 to 1.45) | <.001 | 1.16 |
| Race or ethnicity | ||||||
| Non-Hispanic white | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Non-Hispanic African American | 1.21 (1.03 to 1.42) | .022 | 1.15 | 1.19 (1.00 to 1.42) | .046 | 1.10 |
| Hispanics | 1.33 (1.07 to 1.66) | .012 | 1.22 | 1.40 (1.18 to 1.66) | <.001 | 1.20 |
| Other | 1.39 (1.12 to 1.73) | .003 | 1.26 | 1.40 (1.13 to 1.74) | .002 | 1.20 |
| Marital status | ||||||
| Married | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Other | 1.08 (0.96 to 1.22) | .210 | 1.06 | 0.94 (0.85 to 1.05) | .263 | 0.96 |
| Income, $ | ||||||
| ≤25 000 | 1.90 (1.55 to 2.33) | <.001 | 1.53 | 1.39 (1.17 to 1.65) | <.001 | 1.19 |
| 25 000–50 000 | 1.44 (1.18 to 1.75) | <.001 | 1.29 | 1.09 (0.93 to 1.28) | .311 | 1.05 |
| ≥50 000 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Education | ||||||
| ≤Eighth grade | 1.40 (1.15 to 1.69) | <.001 | 1.26 | 1.53 (1.31 to 1.79) | <.001 | 1.25 |
| Ninth grade to high school | 1.07 (0.92 to 1.24) | .370 | 1.05 | 1.02 (0.90 to 1.15) | .749 | 1.01 |
| Some college | 1.12 (0.96 to 1.30) | .158 | 1.09 | 1.04 (0.90 to 1.19) | .619 | 1.02 |
| Associate degree or higher | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Comorbidity, No. | ||||||
| 0 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| 1 | 2.13 (1.61 to 2.81) | <.001 | 1.63 | 1.36 (1.13 to 1.63) | .001 | 1.18 |
| 2 | 3.60 (2.73 to 4.74) | <.001 | 2.12 | 2.16 (1.83 to 2.55) | <.001 | 1.45 |
| 3 or more | 8.69 (6.65 to 11.4) | <.001 | 2.83 | 5.18 (4.42 to 6.08) | <.001 | 1.87 |
| Cancer | 1.19 (1.06 to 1.33) | .002 | 1.13 | 1.25 (1.13 to 1.38) | <.001 | 1.13 |
| c-statistics | 0.731 | 0.739 | ||||
| Cancer subtype† | ||||||
| Lung | 1.57 (1.07 to 2.30) | .021 | 1.36 | 2.29 (1.34 to 3.92) | .003 | 1.48 |
| Colon | 0.91 (0.69 to 1.22) | .536 | 0.93 | 1.24 (0.96 to 1.59) | .099 | 1.13 |
| Breast | 1.02 (0.84 to 1.25) | .820 | 1.02 | 1.17 (0.98 to 1.40) | .089 | 1.09 |
| Cervical or uterine | 1.15 (0.87 to 1.52) | .326 | 1.11 | 0.96 (0.70 to 1.31) | .784 | 0.98 |
| Prostate | 1.04 (0.83 to 1.29) | .755 | 1.03 | 1.04 (0.86 to 1.26) | .693 | 1.02 |
| Bladder | 1.48 (0.95 to 2.30) | .086 | 1.31 | 1.11 (0.72 to 1.72) | .629 | 1.06 |
| Ovarian | 1.86 (1.16 to 2.97) | .009 | 1.51 | 2.05 (1.22 to 3.46) | .007 | 1.42 |
| Other cancer | 1.42 (1.16 to 1.72) | <.001 | 1.28 | 1.50 (1.23 to 1.83) | <.001 | 1.24 |
All statistical tests were two-sided. The statistical test of significance applied for each variable's coefficient within the multivariable logistic regression was the Wald test. ADL = activities of daily living; CI = confidence interval; IADL = instrumental activities of daily living; OR = odds ratio; ref = referent; RR = relative risk.
Analysis used cancer subtype instead of cancer as independent variables while controlling for the covariates in the table.
Table 5.
Associations between a cancer diagnosis and risk of low self-reported health or geriatric syndromes among Medicare beneficiaries (N = 12 480)*
| Geriatric syndromes |
Low self-rated health |
|||||
| Independent variable | OR (95% CI) | P | RR | OR (95% CI) | P | RR |
| Age, y | ||||||
| 65–69 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| 70–74 | 0.99 (0.87 to 1.13) | .883 | 1.00 | 0.80 (0.69 to 0.93) | .004 | 0.84 |
| 75–79 | 1.11 (0.97 to 1.26) | .125 | 1.05 | 0.72 (0.62 to 0.84) | <.001 | 0.77 |
| 80–84 | 1.56 (1.37 to 1.78) | <.001 | 1.20 | 0.74 (0.64 to 0.86) | <.001 | 0.78 |
| ≥85 | 1.90 (1.65 to 2.19) | <.001 | 1.28 | 0.71 (0.59 to 0.84) | <.001 | 0.76 |
| Sex | ||||||
| Male | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Female | 2.56 (2.37 to 2.77) | <.001 | 1.39 | 0.97 (0.87 to 1.08) | .563 | 0.98 |
| Race or ethnicity | ||||||
| Non-Hispanic white | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Non-Hispanic African American | 0.65 (0.54 to 0.77) | <.001 | 0.80 | 1.68 (1.43 to 1.98) | <.001 | 1.47 |
| Hispanics | 0.96 (0.80 to 1.15) | .664 | 0.98 | 1.10 (0.86 to 1.39) | <.001 | 1.08 |
| Other | 1.13 (0.92 to 1.39) | .259 | 1.06 | 1.43 (1.19 to 1.73) | .462 | 1.31 |
| Marital status | ||||||
| Married | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Other | 1.14 (1.04 to 1.25) | .004 | 1.06 | 0.90 (0.79 to 1.01) | .075 | 0.92 |
| Income, $ | ||||||
| ≤25 000 | 1.20 (1.03 to 1.40) | .022 | 1.08 | 2.36 (1.90 to 2.93) | <.001 | 1.84 |
| 25 000–50 000 | 1.07 (0.92 to 1.24) | .381 | 1.03 | 1.40 (1.12 to 1.76) | .003 | 1.29 |
| ≥50 000 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Education | ||||||
| ≤Eighth grade | 1.29 (1.07 to 1.54) | .006 | 1.12 | 2.31 (1.91 to 2.80) | <.001 | 1.81 |
| Ninth grade to high school | 1.00 (0.88 to 1.13) | .991 | 1.00 | 1.47 (1.23 to 1.75) | <.001 | 1.34 |
| Some college | 1.04 (0.90 to 1.20) | .592 | 1.02 | 1.23 (1.02 to 1.47) | .026 | 1.17 |
| Associate degree or higher | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Comorbidity, No. | ||||||
| 0 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| 1 | 1.38 (1.19 to 1.60) | <.001 | 1.15 | 2.34 (1.76 to 3.23) | <.001 | 1.83 |
| 2 | 1.62 (1.41 to 1.87) | <.001 | 1.21 | 4.18 (3.12 to 5.60) | <.001 | 2.51 |
| 3 or more | 3.33 (2.93 to 3.79) | <.001 | 1.48 | 11.8 (8.93 to 15.7) | <.001 | 3.62 |
| Cancer | 1.27 (1.15 to 1.41) | <.001 | 1.11 | 1.46 (1.30 to 1.64) | <.001 | 1.33 |
| c-statistics | 0.709 | 0.740 | ||||
| Cancer subtype† | ||||||
| Lung | 1.23 (0.76 to 1.98) | .395 | 1.09 | 3.04 (2.09 to 4.44) | <.001 | 2.13 |
| Colon | 1.29 (1.01 to 1.67) | .046 | 1.12 | 1.45 (1.05 to 2.01) | .024 | 1.33 |
| Breast | 1.07 (0.88 to 1.31) | .489 | 1.03 | 1.12 (0.88 to 1.43) | .363 | 1.09 |
| Cervical or uterine | 1.26 (0.93 to 1.71) | .138 | 1.10 | 1.01 (0.75 to 1.36) | .948 | 1.01 |
| Prostate | 1.37 (1.11 to 1.69) | .004 | 1.14 | 1.26 (0.99 to 1.60) | .065 | 1.20 |
| Bladder | 1.40 (0.91 to 2.16) | .129 | 1.15 | 1.65 (1.07 to 2.55) | .024 | 1.45 |
| Ovarian | 1.79 (1.08 to 2.96) | .023 | 1.26 | 1.49 (0.91 to 2.46) | .117 | 1.35 |
| Other cancer | 1.13 (0.93 to 1.37) | .237 | 1.06 | 1.75 (1.44 to 2.12) | <.001 | 1.51 |
All statistical tests were two-sided. The statistical test of significance applied for each variable's coefficient within the multivariable logistic regression was the Wald test. CI = confidence interval; OR = odds ratio; ref = referent; RR = relative risk.
Analysis used cancer subtype instead of cancer as independent variables while controlling for the covariates in the table.
Table 6.
Association of a cancer diagnosis with vulnerability and frailty in elderly Medicare beneficiaries (N = 12 480)*
| Vulnerability |
Frailty |
|||||
| Independent variable | OR (95% CI) | P | RR | OR (95% CI) | P | RR |
| Sex | ||||||
| Male | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Female | 1.26 (1.15 to 1.39) | <.001 | 1.14 | 1.67 (1.51 to 1.85) | <.001 | 1.12 |
| Race or ethnicity | ||||||
| Non-Hispanic white | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Non-Hispanic African American | 0.94 (0.81 to 1.08) | .376 | 0.96 | 0.78 (0.66 to 0.93) | .007 | 0.93 |
| Hispanics | 0.97 (0.79 to 1.19) | .061 | 0.98 | 0.80 (0.65 to 0.99) | .044 | 0.94 |
| Other | 0.84 (0.70 to 1.01) | .789 | 0.90 | 0.90 (0.71 to 1.16) | .422 | 0.97 |
| Marital status | ||||||
| Married | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Other | 1.64 (1.48 to 1.82) | <.001 | 1.31 | 1.39 (1.25 to 1.54) | <.001 | 1.08 |
| Income, $ | ||||||
| ≤25 000 | 1.97 (1.65 to 2.35) | <.001 | 1.42 | 1.61 (1.35 to 1.92) | <.001 | 1.11 |
| 25 000–50 000 | 1.27 (1.08 to 1.49) | .004 | 1.15 | 1.14 (0.97 to 1.34) | .101 | 1.03 |
| ≥50 000 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Education | ||||||
| ≤Eighth grade | 2.37 (2.00 to 2.81) | <.001 | 1.54 | 1.61 (1.35 to 1.92) | <.001 | 1.11 |
| Ninth grade to high school | 1.31 (1.14 to 1.50) | <.001 | 1.17 | 1.14 (0.97 to 1.34) | .244 | 1.03 |
| Some college | 1.21 (1.04 to 1.40) | .013 | 1.12 | 1.63 (1.36 to 1.96) | .067 | 1.12 |
| Associate degree or higher | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||
| Comorbidity, No. | ||||||
| 0 | 1 (ref) | 1 (ref) | NA | NA | ||
| 1 | 1.80 (1.49 to 2.18) | <.001 | 1.37 | NA | NA | |
| 2 | 3.21 (2.65 to 3.88) | <.001 | 1.71 | NA | NA | |
| 3 or more | 8.36 (6.97 to 10.0) | <.001 | 2.14 | NA | NA | |
| Cancer | 1.26 (1.13 to 1.41) | <.001 | 1.14 | 1.46 (1.29 to 1.65) | <.001 | 1.09 |
| c-statistics | 0.742 | 0.646 | ||||
| Cancer subtype† | ||||||
| Lung | 2.00 (1.33 to 3.01) | .001 | 1.43 | 1.27 (0.72 to 2.26) | .406 | 1.06 |
| Colon | 1.33 (1.01 to 1.74) | .043 | 1.18 | 1.40 (1.02 to 1.93) | .040 | 1.08 |
| Breast | 1.05 (0.87 to 1.27) | .616 | 1.03 | 1.32 (1.01 to 1.72) | .039 | 1.07 |
| Cervical or uterine | 0.98 (0.73 to 1.31) | .887 | 0.99 | 1.36 (0.90 to 2.04) | .140 | 1.08 |
| Prostate | 1.22 (1.00 to 1.48) | .050 | 1.12 | 1.45 (1.16 to 1.80) | .001 | 1.09 |
| Bladder | 1.62 (1.03 to 2.55) | .036 | 1.30 | 2.04 (1.21 to 3.43) | .007 | 1.16 |
| Ovarian | 1.26 (0.76 to 2.10) | .373 | 1.14 | 1.19 (0.64 to 2.20) | .578 | 1.04 |
| Other cancer | 1.31 (1.08 to 1.60) | .006 | 1.17 | 1.25 (0.96 to 1.63) | .092 | 1.06 |
All statistical tests were two-sided. The statistical test of significance applied for each variable's coefficient within the multivariable logistic regression was the Wald test. CI = confidence interval; OR = odds ratio; ref = referent; RR = relative risk.
Analysis used cancer subtype instead of cancer as independent variables while controlling for the covariates in the table.
Table 6 examines associations of cancer with the VES-13 and Balducci criteria for frailty, measures that have been advocated for screening for impairment in geriatric oncology after adjustment for known clinical and demographic factors. A cancer diagnosis, compared with no cancer diagnosis, was statistically significantly associated with having a VES-13 score of 3 or higher (adjusted OR = 1.26, 95% CI = 1.13 to 1.41; P < .001; RR = 1.14) and with an increased risk of frailty (adjusted OR = 1.46, 95% CI = 1.29 to 1.65; P < .001; RR = 1.09). A diagnosis of lung cancer or “other” cancer, compared with no cancer diagnosis, was statistically significantly associated with having a VES-13 score of 3 or higher (adjusted OR for lung cancer = 2.00, 95% CI = 1.33 to 3.01; P = .001; RR = 1.43; and adjusted OR for “other” cancer = 1.31, 95% CI = 1.08 to 1.60; P = .006; RR = 1.17). A diagnosis of prostate or bladder cancer, compared with no cancer diagnosis, was statistically significantly associated with frailty (adjusted OR for prostate cancer = 1.45, 95% CI = 1.16 to 1.80; P = .001; RR = 1.09; and adjusted OR for bladder cancer = 2.04, 95% CI = 1.21 to 3.43; P = .007; RR = 1.16).
Discussion
In a nationally representative sample, elders with a cancer diagnosis had a high prevalence of having a limitation in an activity of daily living (31.9%), a limitation in an instrumental activity of daily living (49.5%), a common geriatric syndrome (60.8%), and low self-rated health (27.4%). A high proportion of elders with a history of cancer also scored as “vulnerable” on the VES-13 (45.8%) and as “frail” as defined by Balducci criteria (79.6%). The high prevalence of geriatric syndromes appeared to contribute largely to the designation of frailty. We also found that a cancer diagnosis was statistically significantly associated with the presence of functional limitations (as measured by activities of daily living and instrumental activities of daily living), the presence of a geriatric syndrome, low self-rated health, having a VES-13 score of 3 or higher, and satisfying the Balducci criteria for frailty (P < .001 for each characteristic).
Vulnerability and frailty have emerged as important conditions to assess as decisions are made about the care of older persons in general and older patients with cancer in particular. Potential adverse outcomes resulting from vulnerability and frailty include falls, injuries, susceptibilities to acute illnesses, hospitalization, frank disability, and dependence as well as institutionalization and death (17). Many models have been developed to assess vulnerability and frailty. A commonly used predictive model was developed through the Cardiovascular Health Study (47). In this study, Fried et al. defined frailty as having at least three of the following characteristics: unintentional weight loss, self-reported exhaustion, poor grip strength, slow walking speed, and low physical activity. The overall prevalence of frailty was 6.9% by this definition, and frailty was associated with falling, worsening mobility or functional disability, hospitalization, or death during the next 3 years. Other approaches for assessing vulnerability and frailty include the VES-13 and measures that incorporate age, comorbidity, geriatric syndromes, and limitations in the activities of daily living. In an earlier cohort of Medicare Current Beneficiary Survey participants from 1993 to 1995, a score of 3 or higher on the VES-13 indicated a higher likelihood of decline or death during the next 2 years (10). In another large cohort of community-dwelling older adults (34), the 2-year mortality rate was 27% for older persons with limitations in activities of daily living and 14% for those with limitations in instrumental activities of daily living. Among older persons, each geriatric syndrome has been associated with increased risk for functional decline and death (39,40,48). Because the Medicare Current Beneficiary Survey captured age, comorbidity, geriatric syndromes, and limitations in activities of daily living, these variables were used in this study to identify older persons with a history of cancer potentially at risk for adverse outcomes. Further research in the assessment of the older cancer patient should investigate the associations of these underlying measures or tools for predicting adverse outcomes for older cancer patients with specific subtypes and stages of cancer.
Our study results that show a high prevalence of functional limitations and geriatric syndromes in older cancer patients are consistent with those of others (18,22,24,35). Stafford and Cyr (35) analyzed data from the 1991 cohort of the Medicare Current Beneficiary Survey and reported that cancer was associated with reduced physical function. Korouckian et al. (18) used the Ohio Cancer Incidence Surveillance System to evaluate the prevalence of comorbidity, disability, and geriatric syndromes among elders with cancer who were receiving home health care. They found that the proportion of persons with no comorbidity, disability, or geriatric syndromes was only 26% among breast cancer patients, 14% among colon cancer patients, and 12% among prostate cancer patients. Flood et al. (22) showed that geriatric syndromes were highly prevalent in hospitalized cancer patients. Fifty percent of older patients with prostate cancer who were receiving androgen deprivation therapy were reported to have scored as “vulnerable” on the VES-13 (24). In these four studies, it is unclear whether a personal history of cancer or other comorbidities was independently associated with the increase in factors that are related to vulnerability in older persons. In our analyses, we found that a cancer diagnosis was associated with an increased likelihood of having limitations in activities of daily living (adjusted OR = 1.19, 95% CI = 1.06 to 1.33; RR = 1.13), limitations in instrumental activities of daily living (adjusted OR = 1.25, 95% CI = 1.13 to 1.38; RR = 1.13), a geriatric syndrome (adjusted OR = 1.27, 95% CI = 1.15 to 1.41; RR = 1.11), VES-13 score of 3 or higher (adjusted OR = 1.26, 95% CI = 1.13 to 1.41; RR = 1.14), and frailty as defined by Balducci criteria (adjusted OR = 1.46, 95% CI = 1.29 to 1.65; RR = 1.09) compared with those without cancer. It is important also to note the strong associations of increasing age and comorbidities with having functional limitations, geriatric syndromes, vulnerability, and frailty. For example, persons aged 85 years or older had a 92% higher probability of having a limitation in activities of daily living than those aged 65–69 years. The category of having three or more comorbidities was most strongly associated with having functional limitations, geriatric syndromes, low self-rated health, and vulnerability. These findings highlight the importance of considering these covariates when assessing an older person's risk of adverse health outcomes (48).
This study had several limitations that should be considered when interpreting its findings. The first limitation was that our cancer sample was heterogeneous in that it likely included patients at each disease stage, from long-term survivors to those receiving active treatment for metastatic disease. The risk profile associated with a recent cancer diagnosis is likely to be substantially different from that associated with a distant personal history of cancer with no evidence of disease. A second limitation was that this study was cross-sectional and we could not determine whether or not cancer or its treatment directly causes vulnerability or frailty in older patients. A third limitation was that the number of patients with some cancer subtypes was small, and so the corresponding subgroup analysis may be underpowered to detect statistically significant associations. A fourth limitation was that we could not perform age-adjusted analyses to evaluate the association of cancer with scoring 3 or higher on the VES-13 or meeting the Balducci criteria for frailty because age is included within the scoring systems of these two measures. A fifth limitation was that we were able to report only on those geriatric syndromes that were included in the Medicare Current Beneficiary Survey; syndromes such as delirium, failure to thrive, and self-neglect were not captured. Because we were restricted to the questions asked on the Medicare Current Beneficiary Survey and the self-report of a cancer diagnosis, our comorbidity and geriatric syndrome lists have not been previously validated in independent cohorts. Self-report, however, has been shown to reliably identify comorbidity in other settings (49,50).
Despite the above limitations, this study establishes the increased baseline prevalence of factors among cancer patients that have been associated with adverse health outcomes, such as mortality, in the more general geriatrics population. However, until assessment tools such as the VES-13 and Balducci frailty criteria are validated in independent cohort datasets for specific cancer subtypes and stages, clinicians and researchers should be careful when using this information for clinical decision making or risk stratification within clinical trials for cancer treatment. We consider this study to be a first step in highlighting the importance of “staging the aging” (ie, assessing factors that characterize the physiological and functional capacity) among cancer patients by the use of geriatric assessment (48). The development and validation of tools that assess underlying vulnerability and frailty should aid in the future design of clinical trials that are directed not only at improving survival but also at maintaining function, enhancing quality of life, and preventing geriatric syndromes.
Funding
Hartford Geriatrics Health Outcomes Research Scholars Award (S.G.M.)
Footnotes
The authors had full responsibility for the design of the study, the collection of the data, the analysis and interpretation of the data, the decision to submit the manuscript for publication, and the writing of the manuscript.
References
- 1.Balducci L, Extermann M. Management of the frail person with advanced cancer. Crit Rev Oncol Hematol. 2000;33(2):143–148. doi: 10.1016/s1040-8428(99)00063-3. [DOI] [PubMed] [Google Scholar]
- 2.Balducci L. Epidemiology of cancer and aging. J Oncol Manag. 2005;14(2):47–50. [PubMed] [Google Scholar]
- 3.Repetto L, Comandini D, Mammoliti S. Life expectancy, comorbidity and quality of life: the treatment equation in the older cancer patients. Crit Rev Oncol Hematol. 2001;37(2):147–152. doi: 10.1016/s1040-8428(00)00104-9. [DOI] [PubMed] [Google Scholar]
- 4.Muss HB, Cohen HJ, Lichtman SM. Clinical research in the older cancer patient. Hematol Oncol Clin North Am. 2000;14(1):283–291. doi: 10.1016/s0889-8588(05)70289-6. [DOI] [PubMed] [Google Scholar]
- 5.Balducci L. The geriatric cancer patient: equal benefit from equal treatment. Cancer Control. 2001;8(2 suppl):1–25. quiz 27–28. [PubMed] [Google Scholar]
- 6.Balducci L, Extermann M. Cancer and aging. An evolving panorama. Hematol Oncol Clin North Am. 2000;14(1):1–16. doi: 10.1016/s0889-8588(05)70274-4. [DOI] [PubMed] [Google Scholar]
- 7.Repetto L, Balducci L. A case for geriatric oncology. Lancet Oncol. 2002;3(5):289–297. doi: 10.1016/s1470-2045(02)00730-1. [DOI] [PubMed] [Google Scholar]
- 8.Extermann M, Aapro M, Bernabei R, et al. Use of comprehensive geriatric assessment in older cancer patients: recommendations from the task force on CGA of the International Society of Geriatric Oncology (SIOG) Crit Rev Oncol Hematol. 2005;55(3):241–252. doi: 10.1016/j.critrevonc.2005.06.003. [DOI] [PubMed] [Google Scholar]
- 9.Balducci L, Cohen HJ, Engstrom PF, et al. Senior adult oncology clinical practice guidelines in oncology. J Natl Compr Canc Netw. 2005;3(4):572–590. doi: 10.6004/jnccn.2005.0032. [DOI] [PubMed] [Google Scholar]
- 10.Saliba D, Elliott M, Rubenstein LZ, et al. The Vulnerable Elders Survey: a tool for identifying vulnerable older people in the community. J Am Geriatr Soc. 2001;49(12):1691–1699. doi: 10.1046/j.1532-5415.2001.49281.x. [DOI] [PubMed] [Google Scholar]
- 11.Fried LP, Ferrucci L, Darer J, Williamson JD, Anderson G. Untangling the concepts of disability, frailty, and comorbidity: implications for improved targeting and care. J Gerontol A Biol Sci Med Sci. 2004;59(3):255–263. doi: 10.1093/gerona/59.3.m255. [DOI] [PubMed] [Google Scholar]
- 12.Balducci L, Extermann M. Management of cancer in the older person: a practical approach. Oncologist. 2000;5(3):224–237. doi: 10.1634/theoncologist.5-3-224. [DOI] [PubMed] [Google Scholar]
- 13.Hurria A, Gupta S, Zauderer M, et al. Developing a cancer-specific geriatric assessment: a feasibility study. Cancer. 2005;104(9):1998–2005. doi: 10.1002/cncr.21422. [DOI] [PubMed] [Google Scholar]
- 14.Saliba D, Orlando M, Wenger NS, Hays RD, Rubenstein LZ. Identifying a short functional disability screen for older persons. J Gerontol A Biol Sci Med Sci. 2000;55(12):M750–M756. doi: 10.1093/gerona/55.12.m750. [DOI] [PubMed] [Google Scholar]
- 15.Min LC, Elliott MN, Wenger NS, Saliba D. Higher vulnerable elders survey scores predict death and functional decline in vulnerable older people. J Am Geriatr Soc. 2006;54(3):507–511. doi: 10.1111/j.1532-5415.2005.00615.x. [DOI] [PubMed] [Google Scholar]
- 16.Balducci L. Aging, frailty, and chemotherapy. Cancer Control. 2007;14(1):7–12. doi: 10.1177/107327480701400102. [DOI] [PubMed] [Google Scholar]
- 17.Ferrucci L, Guralnik JM, Cavazzini C, et al. The frailty syndrome: a critical issue in geriatric oncology. Crit Rev Oncol Hematol. 2003;46(2):127–137. doi: 10.1016/s1040-8428(02)00177-4. [DOI] [PubMed] [Google Scholar]
- 18.Koroukian SM, Murray P, Madigan E. Comorbidity, disability, and geriatric syndromes in elderly cancer patients receiving home health care. J Clin Oncol. 2006;24(15):2304–2310. doi: 10.1200/JCO.2005.03.1567. [DOI] [PubMed] [Google Scholar]
- 19.Basso U, Tonti S, Bassi C, et al. Management of Frail and Not-Frail elderly cancer patients in a hospital-based geriatric oncology program. Crit Rev Oncol Hematol. 2008;66(2):163–170. doi: 10.1016/j.critrevonc.2007.12.006. [DOI] [PubMed] [Google Scholar]
- 20.Biganzoli L, Aapro M, Balducci L, Crivellari D, Minisini A, Piccart M. Adjuvant therapy in elderly patients with breast cancer. Clin Breast Cancer. 2004;5(3):188–195. doi: 10.3816/cbc.2004.n.022. discussion 196–197. [DOI] [PubMed] [Google Scholar]
- 21.Balducci L. Management of cancer pain in geriatric patients. J Support Oncol. 2003;1(3):175–191. [PubMed] [Google Scholar]
- 22.Flood KL, Carroll MB, Le CV, Ball L, Esker DA, Carr DB. Geriatric syndromes in elderly patients admitted to an oncology-acute care for elders unit. J Clin Oncol. 2006;24(15):2298–2303. doi: 10.1200/JCO.2005.02.8514. [DOI] [PubMed] [Google Scholar]
- 23.Bylow K, Mohile SG, Stadler WM, Dale W. Does androgen-deprivation therapy accelerate the development of frailty in older men with prostate cancer? A conceptual review. Cancer. 2007;110(12):2604–2613. doi: 10.1002/cncr.23084. [DOI] [PubMed] [Google Scholar]
- 24.Mohile SG, Bylow K, Dale W, et al. A pilot study of the Vulnerable Elders Survey-13 compared with the comprehensive geriatric assessment for identifying disability in older patients with prostate cancer who receive androgen ablation. Cancer. 2007;109(4):802–810. doi: 10.1002/cncr.22495. [DOI] [PubMed] [Google Scholar]
- 25.Centers for Medicare and Medicaid Services. Medicare Current Beneficiary Survey, Access to Care 2003. http://www.cms.gov/mcbs. [Google Scholar]
- 26.Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW. Studies of illness in the aged. The index of ADL: a standardized measure of biological and psychosocial function. JAMA. 1963;185:914–919. doi: 10.1001/jama.1963.03060120024016. [DOI] [PubMed] [Google Scholar]
- 27.Lawton MP. Scales to measure competence in everyday activities. Psychopharmacol Bull. 1988;24(4):609–614. [PubMed] [Google Scholar]
- 28.Inouye SK, Peduzzi PN, Robison JT, Hughes JS, Horwitz RI, Concato J. Importance of functional measures in predicting mortality among older hospitalized patients. JAMA. 1998;279(15):1187–1193. doi: 10.1001/jama.279.15.1187. [DOI] [PubMed] [Google Scholar]
- 29.Guralnik JM, Ferrucci L, Simonsick EM, Salive ME, Wallace RB. Lower-extremity function in persons over the age of 70 years as a predictor of subsequent disability. N Engl J Med. 1995;332(9):556–561. doi: 10.1056/NEJM199503023320902. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Extermann M, Overcash J, Lyman GH, Parr J, Balducci L. Comorbidity and functional status are independent in older cancer patients. J Clin Oncol. 1998;16(4):1582–1587. doi: 10.1200/JCO.1998.16.4.1582. [DOI] [PubMed] [Google Scholar]
- 31.Juurlink DN, Mamdani M, Kopp A, Laupacis A, Redelmeier DA. Drug-drug interactions among elderly patients hospitalized for drug toxicity. JAMA. 2003;289(13):1652–1658. doi: 10.1001/jama.289.13.1652. [DOI] [PubMed] [Google Scholar]
- 32.Stump TE, Callahan CM, Hendrie HC. Cognitive impairment and mortality in older primary care patients. J Am Geriatr Soc. 2001;49(7):934–940. doi: 10.1046/j.1532-5415.2001.49184.x. [DOI] [PubMed] [Google Scholar]
- 33.Neal RD, Allgar VL. Sociodemographic factors and delays in the diagnosis of six cancers: analysis of data from the “National Survey of NHS Patients: Cancer. Br J Cancer”. 2005;92(11):1971–1975. doi: 10.1038/sj.bjc.6602623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Reuben DB, Rubenstein LV, Hirsch SH, Hays RD. Value of functional status as a predictor of mortality: results of a prospective study. Am J Med. 1992;93(6):663–669. doi: 10.1016/0002-9343(92)90200-u. [DOI] [PubMed] [Google Scholar]
- 35.Stafford RS, Cyr PL. The impact of cancer on the physical function of the elderly and their utilization of health care. Cancer. 1997;80(10):1973–1980. doi: 10.1002/(sici)1097-0142(19971115)80:10<1973::aid-cncr15>3.0.co;2-v. [DOI] [PubMed] [Google Scholar]
- 36.O’Malley AS, Forrest CB, Feng S, Mandelblatt J. Disparities despite coverage: gaps in colorectal cancer screening among Medicare beneficiaries. Arch Intern Med. 2005;165(18):2129–2135. doi: 10.1001/archinte.165.18.2129. [DOI] [PubMed] [Google Scholar]
- 37.Roehrig B, Hoeffken K, Pientka L, Wedding U. How many and which items of activities of daily living (ADL) and instrumental activities of daily living (IADL) are necessary for screening. Crit Rev Oncol Hematol. 2007;62(2):164–171. doi: 10.1016/j.critrevonc.2006.10.001. [DOI] [PubMed] [Google Scholar]
- 38.Naeim A, Reuben D. Geriatric syndromes and assessment in older cancer patients. Oncology (Williston Park) 2001;15(12):1567–1577. 1580; discussion 1581, 1586, 1591. [PubMed] [Google Scholar]
- 39.Anpalahan M, Gibson SJ. Geriatric syndromes as predictors of adverse outcomes of hospitalization. Intern Med J. 2008;38(1):16–23. doi: 10.1111/j.1445-5994.2007.01398.x. [DOI] [PubMed] [Google Scholar]
- 40.Inouye SK, Studenski S, Tinetti ME, Kuchel GA. Geriatric syndromes: clinical, research, and policy implications of a core geriatric concept. J Am Geriatr Soc. 2007;55(5):780–791. doi: 10.1111/j.1532-5415.2007.01156.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Federman AD, Adams AS, Ross-Degnan D, Soumerai SB, Ayanian JZ. Supplemental insurance and use of effective cardiovascular drugs among elderly Medicare beneficiaries with coronary heart disease. JAMA. 2001;286(14):1732–1739. doi: 10.1001/jama.286.14.1732. [DOI] [PubMed] [Google Scholar]
- 42.Wolff JL, Boult C, Boyd C, Anderson G. Newly reported chronic conditions and onset of functional dependency. J Am Geriatr Soc. 2005;53(5):851–855. doi: 10.1111/j.1532-5415.2005.53262.x. [DOI] [PubMed] [Google Scholar]
- 43.Boyd CM, Weiss CO, Halter J, Han KC, Ershler WB, Fried LP. Framework for evaluating disease severity measures in older adults with comorbidity. J Gerontol A Biol Sci Med Sci. 2007;62(3):286–295. doi: 10.1093/gerona/62.3.286. [DOI] [PubMed] [Google Scholar]
- 44.Zhang J, Yu KF. What's the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA. 1998;280(19):1690–1691. doi: 10.1001/jama.280.19.1690. [DOI] [PubMed] [Google Scholar]
- 45.Vaeth M, Skovlund E. A simple approach to power and sample size calculations in logistic regression and Cox regression models. Stat Med. 2004;23(11):1781–1792. doi: 10.1002/sim.1753. [DOI] [PubMed] [Google Scholar]
- 46.Norris CM, Ghali WA, Knudtson ML, Naylor CD, Saunders LD. Dealing with missing data in observational health care outcome analyses. J Clin Epidemiol. 2000;53(4):377–383. doi: 10.1016/s0895-4356(99)00181-x. [DOI] [PubMed] [Google Scholar]
- 47.Fried LP, Tangen CM, Walston J, et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56(3):M146–M156. doi: 10.1093/gerona/56.3.m146. [DOI] [PubMed] [Google Scholar]
- 48.Rodin MB, Mohile SG. A practical approach to geriatric assessment in oncology. J Clin Oncol. 2007;25(14):1936–1944. doi: 10.1200/JCO.2006.10.2954. [DOI] [PubMed] [Google Scholar]
- 49.Stewart AL, Greenfield S, Hays RD, et al. Functional status and well-being of patients with chronic conditions. Results from the Medical Outcomes Study. JAMA. 1989;262(7):907–913. [PubMed] [Google Scholar]
- 50.Chaudhry S, Jin L, Meltzer D. Use of a self-report-generated Charlson Comorbidity Index for predicting mortality. Med Care. 2005;43(6):607–615. doi: 10.1097/01.mlr.0000163658.65008.ec. [DOI] [PubMed] [Google Scholar]