An evaluation of lower-body functional limitations among long-term survivors of 11 different types of cancers

Abstract

BACKGROUND:

The authors examined potential reasons (sociodemographics, psychologic distress, health behavior, chronic health conditions, access to medical care) for increased prevalence of lower-body functional limitations among long-term (≥5 years) cancer survivors.

METHODS:

The authors used National Health Interview Survey data from 2005 through 2007, and defined lower-body functional limitation as reporting difficulty/inability to perform at least 1 of 5 activities (walking approximately one–quarter of a mile; walking up and down 10 steps without rest; standing for 2 hours; stooping, crouching, or kneeling; and lifting 10 lbs). Increased prevalence of lower-body functional limitations was compared between long-term survivors of each of 11 cancer types reported by ≥50 respondents (n = 2143) and persons without cancer history (controls; n = 72,618).

RESULTS:

Among cancer survivors, 57.0% had a lower-body functional limitation versus 26.6% of controls. The unadjusted prevalence of lower-body functional limitations varied by cancer type, ranging from 44.9% (lymphoma survivors) to 88.8% (lung cancer survivors). Long-term lung (odds ratio [OR], 7.91), uterine (OR, 2.41), thyroid (OR, 2.27), cervical (OR, 1.76), ovarian (OR, 1.75), and breast (OR, 1.35) cancer survivors had increased odds of reporting a lower-body functional limitation than controls after adjusting for sociodemographic factors (all P < .05). Differences in the prevalence of arthritis and lower-back pain and in access to medical care explained differences in lower-body functional limitation prevalence between controls and long-term breast, cervical, ovarian, and uterine cancer survivors. Long-term bladder, colorectal, lymphoma, melanoma, and prostate cancer survivors were equally likely to report a lower-body functional limitation as controls.

CONCLUSIONS:

Treatment of arthritis and lower–back pain and increasing access to medical care might help reduce the risk of lower-body functional limitations and improve quality of life among specific long-term cancer survivors. Cancer 2009. © 2009 American Cancer Society.

Approximately 1.4 million people were predicted to be diagnosed with cancer in the United States in 2008.1 This number is expected to double by the year 2050. The number of long-term cancer survivors (ie, ≥5 years postdiagnosis) also is increasing rapidly because of increased screening and improved treatment. It is estimated that nearly 11 million Americans were alive with a cancer history in 2004.1 As a result, the number of cancer survivors, including long-term survivors, is expected to increase dramatically in the next few decades.

Although cancer is increasingly being viewed as a chronic illness, and many cancer survivors are in good health, there is the potential for experiencing adverse medical and psychosocial outcomes after cancer diagnosis, including functional limitations (inability to perform a physical action, task, or activity in an efficient, typically expected or competent manner).2 Functional status predicts survival, chemotherapy toxicity, postoperative morbidity, and mortality among cancer survivors.3 Because functional limitations represent an early and key transition in the disablement pathway,4, 5 they constitute an early target for interventions or for screening to slow or prevent disability6, 7 and thereby reduce the risk of disability, morbidity, and mortality among cancer survivors.

There are several potential reasons for the increased prevalence of lower-body functional limitations among cancer survivors. First, cancer survivors are likely to be different in terms of their sociodemographic factors (eg, age, income) from persons without a cancer history, and these differences may convey increased vulnerability to lower-body functional limitations.8 Second, receipt of cancer treatment may lead to complications affecting various organs, including effects on the musculoskeletal, cardiovascular, and respiratory systems, which may subsequently lead to functional limitations.9 Third, cancer survivors are more likely to have chronic conditions and diseases than those without a cancer history, which may or may not be the result of types of cancer treatment received. Many of these conditions and diseases are risk factors for functional limitations.8 Fourth, access to and contact with the medical system may increase the likelihood of detecting functional limitations.8 Fifth, psychosocial factors such as depression, increased stress, and poor mental health status, which are risk factors for functional limitations,8 also may be elevated among cancer survivors. Sixth, cancer survivors may be more likely to engage in behaviors that increase the prevalence of functional limitations, such as smoking, or they may be less likely to engage in behaviors that reduce the prevalence of functional limitations, such as physical activity.8, 10

The aim of this analysis was to identify factors associated with increased prevalence of lower-body functional limitations by comparing long-term cancer survivors with persons without a cancer history (controls). Because of differences in the likelihood of lower-body functional limitations across cancers,11 we examined a national sample of 11 specific types of long-term cancer survivors. Identification of these factors could lead to implementation of evidence-based interventions aimed at reducing the likelihood of lower-body functional limitations and thereby preventing inability to perform routine activities as part of daily life among long-term survivors with specific types of cancers.

MATERIALS AND METHODS

Data Source

We used data from the 2005 through 2007 National Health Interview Survey.12 The National Health Interview Survey is a continuous face-to-face household interview that covers a wide variety of health-related topics. It uses a complex, stratified sampling design to provide estimates for the civilian noninstitutionalized US population. Data were collected regarding all family members in approximately 40,000 households annually. More detailed information is obtained from 1 randomly sampled adult household. Participants in this study were limited to respondents who were part of the “sample adult” component of the survey who were asked about their cancer history. Data from the sample adult survey were self-reported, except when respondents were physically incapable of responding themselves. In this study, proxy-reported data accounted for 1.3% of all data (2.9% among long-term cancer survivors and 1.3% among respondents without cancer history). To obtain more stable estimates among survivors with specific types of cancer, we combined 3 years (2005-2007) of data. Response rates for the sample adult component of the National Health Interview Survey in 2005, 2006, and 2007 were 69.0%, 70.8%, and 67.8%, respectively.

Lower-Body Functional Limitations

Five items from the Nagi physical performance scale assessed lower-body functional limitations (0 = no difficulties to 1 = difficulty), which were summed to form the outcome measure (ranging from 0-5).13 This scale, which assesses lower-extremity strength and basic motor functions, has been used in studies of cancer survivors14, 15 and measures difficulties with walking one–quarter of a mile; walking up and down 10 steps without rest; standing for 2 hours; stooping, crouching, or kneeling; and lifting 10 pounds.16 Subjects who expressed any difficulty or inability to perform an activity at the time of the interview were considered to be limited in that activity.17 For this study, we defined lower-body functional limitation as reporting difficulty with or inability to perform at least 1 of the 5 physical activities.17

Cancer History

Cancer history was based on the question “Have you ever been told by a physician or other health professional that you had cancer or a malignancy of any kind?” If the respondent reported a history of cancer, they were asked the site of the cancer and the type of cancer from a list of 30 different types of cancer. Participants could report up to 3 cancer sites. On the basis of the age at the interview and age at the first cancer diagnosis, we calculated the number of years since cancer diagnosis. We excluded persons who reported >1 cancer.

We limited the analysis of cancer-specific data to 11 types of cancer (bladder, breast, cervical, colorectal, lung, lymphoma, melanoma, ovarian, prostate, thyroid, and uterine), which had been reported by at least 50 respondents in the survey and were diagnosed ≥5 years before the interview.15, 18 In addition to persons who reported >1 cancer, we excluded persons diagnosed with other cancers, more recent cancer survivors (<5 years since their diagnosis), and patients diagnosed with nonmelanoma and unknown-type skin cancers from the analysis.15, 19 The comparison (control) group included persons without a cancer history.

Factors Potentially Associated With Increased Lower-Body Functional Limitation Prevalence

We examined factors that may account for any observed association between cancer history and lower-body functional limitations, including: 1) sociodemographic factors, 2) chronic conditions, 3) access to medical care, 4) health behaviors, and 5) psychologic distress. Factors included in the analysis were patterned after other studies in the general population8, 20 and in cancer survivors.11, 15 First, sociodemographic factors included age group, race/Hispanic origin, sex, income categories, family size, receipt of government assistance because of low income, educational attainment, marital status, and home ownership. Second, chronic conditions were comprised of current body mass index categories2 and 11 chronic conditions based on participants' self-reported diagnosis by a physician. The chronic conditions included those previously associated with functional decline,8 namely having diabetes; current asthma; ever having coronary heart disease; ever having a myocardial infarction; ever having other heart disease; having hypertension on >1 occasion; ever having a stroke; ever having arthritis (osteoarthritis, rheumatoid arthritis, gout, lupus, or fibromyalgia); general joint symptoms (pain, aching, or stiffness in or around a joint) during the past 30 days; lower-back pain during past 3 months; and current trouble with vision, even when wearing glasses or contact lenses. Third, access to medical care factors was comprised of not having healthcare insurance at the time of the interview; being unable to afford prescription medicines, mental healthcare, dental care, or eyeglasses during the past 12 months; being unable to see a physician during the 12 months before the interview because of cost; and reporting delay in receipt of medical care during the past 12 months because of inability to 1) get through on the telephone, 2) get an appointment soon enough, 3) get in to see a physician soon enough, 4) get to the clinic/physician's office when it was open, or 5) find transportation to get there. Fourth, health behaviors included smoking status, amount of alcohol use during the past week, and current participation in moderate or vigorous physical activity at least 2 times per week. Fifth, nonspecific psychologic distress was measured using the K6, which asks about the frequency of 6 symptoms of mental illness or nonspecific psychologic distress.21 By using a 5-point response scale for each symptom, scores could range from 6 to 30, with higher scores indicating greater frequency of symptom experience. A score of ≥13 on the K6 was used to indicate serious psychologic distress.22

Statistical Analysis

The National Health Interview Survey collects data through a complex sampling design, involving stratification, clustering, and multistage sampling. To obtain nationally representative estimates, all data were weighted by adjusting for the probability of inclusion in the sample, taking into consideration survey nonresponse as well as the sex, age, and race/ethnicity of survey respondents. We used logistic regression to examine whether long-term cancer survivors had greater odds of prevalent lower-body functional limitations compared with persons without cancer history and also to identify factors potentially associated with increased lower-body functional limitation prevalence among the former population. Variance estimates for proportions and logistic regression models were calculated using the Taylor series approximation. All P values were 2-sided.

The sociodemographic factors, chronic conditions, access to medical care, health behavior, and psychologic distress factors were included as groups of variables in the logistic regression models to examine their effects on the prevalence odds ratio [OR] associated with a specific cancer diagnosis. Because of the differences in sociodemographic factors between persons with and without a cancer history, we examined in separate models the effect of each of the 4 other factors (chronic conditions, access to medical care, health behaviors, and psychologic distress) in separate models after adjustment for sociodemographic factors. A change in the prevalence ORs was considered evidence of the effect of that factor on the association between cancer history and lower-body functional limitation prevalence23 after adjusting for sociodemographic factors. We examined specific variables within a particular factor when that factor demonstrated a significant reduction in OR (at least 20%) for cancer history status; we report these data for specific variables. We used SAS 9.1 (SAS Institute Inc, Cary, NC) and SUDAAN 9.1 (Research Triangle Institute, Research Triangle Park, NC) statistical software to take into account the complex sampling design.

RESULTS

The 2005 through 2007 National Health Interview Survey dataset included 79,096 adults who completed the “sample adult” interview. A total of 4335 (5.5%) respondents were excluded because they met ≥1 of the following exclusion criteria: 1) their only cancer was nonmelanoma (n = 1004) or unknown-type (n = 494) skin cancer, 2) they were diagnosed with cancer <5 years before the interview (n = 2191), 3) their cancer history was unknown (n = 229), or 4) their cancer type was reported by <50 respondents (n = 576). Data regarding 74,761 adults were included in the analysis. Of these, 2143 (2.7%) were long-term cancer survivors and 72,618 were never diagnosed with cancer (controls). Average time since diagnosis was 15.6 years (95% confidence interval [95% CI], 14.9-16.2 years). Of the long-term cancer survivors, 4.5% were diagnosed before age 18 years.

Table 1 describes selected characteristics of the study population. In unadjusted analysis, persons who were older, were female, had lower incomes, had limited access to medical care, were current or former smokers, were not participating in physical activity at least twice a week, had a greater number of chronic conditions, or who reported serious psychologic distress each had increased odds of lower-body functional limitations. Of the long-term cancer survivors, 57.0% (95% CI, 54.6-59.4%) reported ≥1 lower-body functional limitations, compared with 26.6% (95% CI, 26.1-27.0%) of persons without a cancer history. Long-term cancer survivors were 3.66 times (95% CI, 3.32-4.05 times) more likely to report having at least 1 lower-body functional limitation compared with the control group in unadjusted analysis. As shown in Figure 1, the most frequent lower-body functional limitation mentioned by cancer survivors was the inability to stoop, crouch, or kneel (48.8%), followed by standing for 2 hours (44.8%), and walking approximately one–quarter of a mile (39.7%). The prevalence of lower-body functional limitations for each of the 5 types of lower-body functional limitation was consistently higher among long-term cancer survivors compared with the control group. Figure 2 displays the distribution of lower-body functional limitation prevalence in the 2 groups.

Figure 1. Lower-body functional limitations (LBFL) are shown by type among long-term survivors and persons without a cancer history (United States, 2005-2007).

Figure 2. Distribution of lower-body functional limitations among long-term survivors and persons without cancer history (United States, 2005-2007) is shown.

Table 1. Selected Characteristics of the Study Population and Unadjusted Association (OR, 95% CI) With Prevalence of Lower-Body Functional Limitations, 2005 Through 2007

Characteristic	Long-Term Cancer Survivors* (n=7,241,117), %	No Cancer History (n=204,111,198), %	OR	95% CI
OR indicates odds ratio; 95% CI, 95% confidence interval. * Types of cancers that had been reported by at least 50 respondents: bladder, breast, cervical, colorectal, lung, lymphoma, melanoma, ovarian, prostate, thyroid, and uterine. † P < .05 between survivors and controls. ‡ Adjusted for number of family members.
Sociodemographics
Age, y†
18-39	8.9	42.6	1.00
40-64	39.1	43.7	3.00	2.85-3.15
65-84	46.0	12.2	9.02	8.46-9.62
85+	6.1	1.5	25.38	22.27-28.92
Sex†
Male	32.2	48.7	1.00
Female	67.8	51.3	1.57	1.52-1.65
Race†
White	89.4	81.0	1.00
African American	7.1	12.2	0.97	0.92-1.03
Other	3.5	6.8	0.68	0.62-0.74
Family income‡
$0-$34,999	33.6	28.6	2.24	2.10-2.40
$35,000-$74,999	27.4	27.6	1.28	1.19-1.38
≥$75,000	18.6	23.8	1.00
Unknown	20.4	20.0	1.31	1.21-1.41
Healthcare access
Usual place for medical care
Yes	92.0	82.1	0.50	0.47-0.53
No	5.9	15.8	1.00
>1 place	1.3	1.1	1.17	0.97-1.42
Unknown	0.8	1.0	0.69	0.53-0.90
One or more system barriers†
Yes	11.8	9.5	2.51	2.37-2.67
No	88.2	89.5	1.00
Unknown	1.0	1.0	0.97	0.76-1.22
Behavior
Smoking†
Current	19.0	19.1	1.47	1.40-1.55
Former	36.9	19.8	2.03	1.93-2.13
Never	43.1	61.1	1.00
Other/unknown	1.0	1.4	1.01	0.81-1.26
Physical activity≥2 times/wk†
Yes	47.6	53.2	1.00
No	51.2	45.5	1.79	1.71-1.87
Other/unknown	1.2	1.3	0.73	0.58-0.90
Chronic conditions
Arthritis (vs no)†	45.9	19.3	9.18	8.74-9.65
Asthma (vs no)†	10.4	7.1	2.55	2.38-2.74
Coronary heart disease (vs no)†	12.4	3.7	6.67	6.05-7.35
Diabetes (vs no)†	15.0	7.1	4.66	4.35-5.00
General joint symptoms (vs no)†	48.3	27.9	4.67	4.35-5.00
Having trouble seeing (vs no)†	16.8	9.0	4.45	4.20-4.72
Hypertension (vs no)†	43.4	21.5	4.44	4.21-4.63
Lower-back pain (vs no)†	37.7	26.3	4.63	4.44-4.82
Myocardial infarction (vs no)†	8.8	2.8	6.81	6.09-7.61
Other heart disease (vs no)†	16.5	6.4	3.99	3.71-4.28
Stroke (vs no)†	6.7	2.1	10.19	8.86-11.72
Body mass index
<18.5	2.6	1.7	1.42	1.20-1.68
18.5-24.9	34.4	35.7	1.00
25.0-29.9	33.4	33.5	1.37	1.30-1.44
≥30.0	29.6	29.1	2.58	2.46-2.72
Nonspecific psychologic distress†
Yes	11.1	7.8	4.44	4.14-4.76
No	87.1	90.7	1.00
Unknown	1.8	1.5	1.63	1.37-1.93

Factors Associated With Increased Lower-Body Functional Limitation Among Long-Term Survivors of Specific Cancer Types

The unadjusted prevalence of lower-body functional limitations varied considerably by cancer type, ranging from a low of 44.9% among lymphoma survivors to a high of 88.8% among lung cancer survivors (Table 2). When adjusting for sociodemographic factors, long-term breast, cervical, lung, ovarian, thyroid, uterine, and multiple/other cancer survivors had increased odds of reporting lower-body functional limitations than controls (Model 1, Table 3). Long-term bladder, colorectal, lymphoma, melanoma, and prostate cancer survivors were equally as likely to report at least 1 lower-body functional limitation as respondents in the control group after adjusting for sociodemographic factors.

Table 2. Frequency (Unweighted and Weighted), Prevalence of LBFL (95% CI), and Unadjusted Prevalence OR (and 95% CI) by Type of Long-Term Cancer Survivor, 2005 Through 2007*

Type of Cancer	Unweighted Frequency	Weighted Frequency	Prevalence of LBFL	95% CI	OR†	95% CI
LBFL indicates lower-body functional limitations; 95% CI, 95% confidence interval; OR, odds ratio. * Types of cancers that had been reported by at least 50 respondents. † Versus persons without cancer.
Bladder	60	177,280	51.6	36.1-66.8	2.94	1.56-5.55
Breast	583	1,372,577	60.6	56.1-64.9	4.24	3.53-5.10
Cervical	299	784,913	47.1	40.5-53.8	2.46	1.88-3.22
Colorectal	199	469,900	61.8	54.1-68.9	4.46	3.25-6.13
Lung	53	125,412	88.8	77.1-94.9	21.97	9.34-51.71
Lymphoma	77	246,780	44.9	32.6-57.8	2.25	1.34-3.79
Melanoma	197	562,446	46.5	38.3-54.9	2.46	1.71-3.36
Ovarian	95	207,565	52.8	41.3-64.0	3.09	1.95-4.92
Prostate	337	904,315	59.3	52.7-65.5	4.02	3.08-5.25
Thyroid	67	201,500	57.6	42.3-71.7	3.76	2.03-6.98
Uterus	194	518,419	67.3	58.9-74.7	5.68	3.95-8.17

Table 3. ORs* (95% CI) for 5 Groups of Factors Examining Odds of LBFL Among Long-Term Survivors of 11 Types of Cancer Compared With Persons Without a Cancer History, 2005 Through 2007†

Type of Cancer	Model 1: Adjusted for Sociodemographic Factors		Model 2: Adjusted for Sociodemographic Factors and Access to Care		Model 3: Adjusted for Sociodemographic Factors and Chronic Conditions		Model 4: Adjusted for Sociodemographic Factors and Health Behaviors		Model 5: Adjusted for Sociodemographic Factors and Psychologic Distress
	OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI
OR indicates odds ratio; 95% CI, 95% confidence interval; LBFL, lower-body functional limitations. * Versus persons without cancer. † Types of cancers that had been reported by at least 50 respondents.
Bladder	1.23	0.64-2.38	1.05	0.61-1.82	0.77	0.41-1.45	1.13	0.59-2.15	1.25	0.62-2.54
Breast	1.35	1.12-1.63	1.38	1.14-1.67	1.34	1.05-1.70	1.36	1.13-1.64	1.37	1.13-1.66
Cervical	1.76	1.30-2.37	1.34	0.98-1.81	1.12	0.75-1.68	1.52	1.12-2.06	1.58	1.14-2.19
Colorectal	1.28	0.89-1.85	1.27	0.89-1.80	1.19	0.81-1.74	1.21	0.85-1.77	1.27	0.87-1.86
Lung	7.91	3.45-18.15	8.78	3.72-20.75	8.55	3.50-0.88	6.82	2.93-15.84	7.72	3.35-17.81
Lymphoma	1.19	0.68-2.09	1.17	0.69-2.01	1.15	0.57-2.29	1.22	0.70-2.12	1.16	0.67-2.02
Melanoma	1.31	0.94-1.83	1.18	0.82-1.69	1.21	0.78-1.90	1.35	0.96-1.88	1.30	0.92-1.83
Ovarian	1.75	1.04-2.94	1.46	0.89-2.42	1.03	0.52-2.05	1.61	0.95-2.71	1.42	0.85-2.36
Prostate	1.24	0.93-1.64	1.29	0.98-1.69	1.38	0.98-1.94	1.24	0.94-1.64	1.28	0.96-1.69
Thyroid	2.27	1.15-4.48	2.10	0.97-4.54	2.10	1.04-4.23	2.35	1.19-4.62	2.19	1.13-4.25
Uterus	2.41	1.63-3.58	2.11	1.40-3.18	1.50	0.94-2.40	2.18	1.49-3.20	2.69	1.54-3.33

For long-term breast cancer survivors, adding the access to care variables (Model 2, Table 3) to Model 1 did not appear to reduce the OR appreciably, suggesting that access to care variables did not explain differences in lower-body functional limitation prevalence relative to controls. Also, the OR changed little for breast cancer survivors when each of the other 3 groups of factors was added separately to Model 1 (Table 3). However, adjusting for the presence of arthritis alone in Model 1 indicated that breast cancer survivors were equally likely to report at least 1 lower-body functional limitation compared with controls (OR, 1.22; 95% CI, 0.99-1.51).

Long-term cervical cancer survivors had increased odds of reporting a lower-body functional limitation compared with controls (Model 1, Table 3). Adding the access to medical care factors (Model 2) to Model 1 reduced the odds of lower-body functional limitations for cervical cancer survivors compared with controls, as did adding chronic conditions. Inability to afford prescription medicines, mental healthcare, dental care, or eyeglasses during the past 12 months and presence of lower-back pain during the past 30 days reduced the OR to 1.18 (95% CI, 0.86-1.62) relative to Model 1 for cervical cancer survivors compared with controls. The 3 health behaviors (Model 4) and psychosocial distress (Model 5) reduced the OR much less than the access to care and chronic condition variables.

Model 1 (Table 3) shows that lung cancer survivors had increased odds of reporting a lower-body functional limitation compared with controls. Access to medical care (Model 2), chronic conditions (Model 3), and psychosocial distress (Model 5) did not reduce the OR for lung cancer survivors. The health behavior factors (Model 5) reduced the OR to 6.82 relative to Model 1. This was mainly because of increased smoking among lung cancer survivors (OR, 7.17; 95% CI, 3.11-17.92). However, lung cancer survivors had increased odds of reporting a lower-body functional limitation relative to controls.

Women with ovarian cancer had increased odds of reporting a lower-body functional limitation (Model 1, Table 3). The largest reduction in OR relative to Model 1 was observed when adding chronic conditions (Model 3), followed by the psychologic distress (Model 5), access to care (Model 2), and health behavior factors (Model 4). The access to care variables responsible for the reduction in OR to 1.45 (95% CI, 0.90-2.34) included the inability to afford prescription medicines, mental healthcare, dental care, or eyeglasses during the past 12 months. Among the chronic conditions, the presence of arthritis alone reduced the OR to 1.32 (95% CI, 0.72-2.43) when added to Model 1. When the presence of arthritis and reduced access to medical care were all added to Model 1, the OR was reduced to 1.14 (95% CI, 0.64-2.04).

Thyroid cancer survivors had increased odds of reporting a lower-body functional limitation than controls (Model 1, Table 3). The OR for thyroid cancer survivors changed little when adjusting for any of the other factors.

Uterine cancer survivors had increased odds of reporting a lower-body functional limitation (Model 1, Table 3) than controls. Access to medical care (Model 2), health behavior (Model 4), and psychologic distress (Model 5) did not appear to affect this OR appreciably. However, increased presence of arthritis (OR, 1.80; 95% CI, 1.20-2.72) and lower-back pain (OR, 2.00; 95% CI, 1.28-3.11) each reduced the OR relative to Model 1 for uterine cancer survivors. Arthritis and lower-back pain combined reduced the OR to 1.61 (95% CI, 1.01-2.57) for uterine cancer survivors compared with controls when added to Model 1.

DISCUSSION

Among the long-term survivors of 11 types of cancer included in the current study sample, 57.0% had ≥1 lower-body functional limitation compared with 26.6% of controls. Lower-body functional limitation prevalence and associated factors varied by cancer type. There were no differences in lower-body functional limitations noted among long-term bladder, colorectal, lymphoma, melanoma, and prostate cancer survivors compared with controls after adjustment for sociodemographic factors. An increased odds of lower-body functional limitations among long-term lung, uterine, and thyroid cancer survivors were not explained by factors we examined. Increased odds of lower-body functional limitations among long-term cervical, ovarian, and breast cancer survivors were partly explained by having arthritis, lower-back pain, and access to care.

Having arthritis partly explained lower-body functional limitations among women diagnosed with breast, ovarian, and uterine cancer. Arthritis is 1 of the most prevalent chronic conditions and 1 of the leading causes of disability and functional limitations in the general population.24 Interventions to prevent lower-body functional limitations among cancer survivors might include increasing physical activity to increase muscle strength and for breast cancer survivors, modifying systemic breast cancer treatment such as taxanes and aromatase inhibitors, which are known to increase the risk of joint-like symptoms.25, 26 Depressive symptoms also are associated with lower-body functional limitations among persons with arthritis27, 28; therefore, diagnosing and treating symptoms of depression early could reduce the risk of developing lower-body functional limitations among breast, ovarian, or uterine cancer survivors who also have arthritis. Proinflammatory cytokines also may play a role in some cancers, arthritis, and lower-body functional limitations,29, 30 but such information was not available in the National Health Interview Survey data.

Lower-back pain partly explained the increased odds of lower-body functional limitations among women with cervical, ovarian, and uterine cancer. Radiotherapy may play a role in low back pain among women with cervical cancer,31, 32 although conflicting reports have been published.33 Pelvic radiation may result in chronic radiation enteritis, symptoms of which include diarrhea, incontinence, and abdominal bloating/discomfort and can negatively affect the quality of physical, psychologic, and social aspects of life among women with ovarian and uterine cancer.34 It also appears that lack of access to medical care continues to be an issue after the diagnosis of cervical cancer, as it typically is before diagnosis.35

The Institute of Medicine recommends that every cancer survivor has a comprehensive care summary and follow-up plan once they complete their primary cancer care.36 The plan should address post-treatment needs to improve the survivor's health and quality of life. This plan could include arthritis, lower-back pain, and access to medical care, which were found to be associated with an increased odds of lower-body functional limitations. Primary care physicians and oncologists should ensure that cancer survivors with lower-body functional limitations benefit as soon as possible from rehabilitation to regain physical functioning by focusing on the impact of arthritis, lower-back pain, and access to medical care.

Limitations of the current study include the restriction that only noninstitutionalized adults were eligible to participate in the National Health Interview Survey, so cancer survivors whose physical limitation resulted in institutionalization were not included in this study. Second, because some cancers associated with significant loss of lower-body function also are more rapidly fatal, the association between long-term cancer survivorship and lower-body functional limitations may be underestimated. Third, because of the cross-sectional study design, in some cases it is difficult to evaluate the temporal direction of the association between factors associated with the presence of lower-body functional limitations and cancer occurrence depending on the time frame referred to in a question concerning chronic conditions. However, results regarding the increased prevalence of functional limitations among cancer survivors were unchanged in 2 studies when comparing a cross-sectional approach with a longitudinal approach.11, 37 Future studies should use a prospective approach to discern the temporal correlation between arthritis and lower-back pain among women who developed breast, cervical, ovarian, and uterine cancer. Fourth, our analysis relied on self-reported cancer history. The accuracy of self-reported cancer history has been found to vary by cancer type, with breast cancer being reported most accurately and cervical cancer most likely to be underreported.38 As a result, the findings pertaining to lower-body functional limitations may have been underestimated in the current study. However, self-reported chronic conditions were of adequate accuracy.39 Fifth, cancer treatment data were not available in the National Health Interview Survey, and therefore we cannot attribute lower-body functional limitations to the treatment of cancer survivors. However, we did examine the association between lower-body functional limitations and common conditions possibly resulting from cancer treatment, such as heart disease, arthritis, joint problems, and asthma. Finally, limitations in pulmonary function and joint mobility or upper body strength also might result in other functional limitations that are not necessarily specific to the lower body. Therefore, overall functional limitations could be even greater than the lower-body functional limitations reported using the 5-item Nagi scale.

In conclusion, arthritis, lower-back pain, and access to medical care can be targeted to reduce the prevalence of lower-body functional limitations and improve the quality of life of long-term survivors of breast, cervical, ovarian, and uterine cancers, possibly with the aid of a comprehensive assessment and/or survivorship care plan. Additional factors associated with elevated lower-body functional limitation prevalence should be examined for long-term survivors of lung, thyroid, and uterine cancer.

Conflict of Interest Disclosures

Supported in part by grants from the National Cancer Institute (CA112159 and CA91842).