Quality of Life and Impact of Cancer: Differences in Rural and Non-Rural Non-Hodgkin’s Lymphoma Survivors

Abstract

Purpose:

People living in rural areas experience greater health disparities than their non-rural counterparts, but little is known about the association between rural status and quality of life (QOL) in non-Hodgkin’s lymphoma (NHL) survivors. We compared self-reported quality of life and impact of cancer in rural and non-rural NHL survivors.

Methods:

This study is a secondary analysis of 566 NHL cancer survivors recruited from cancer registries at 2 large academic medical centers in 1 state. Standardized measures collected information on demographics and clinical characteristics, quality of life (QOL; SF-36), and the Impact of Cancer (IOCv2). Rural residence was determined by Rural-Urban Commuting Area (RUCA) codes designated as non-metropolitan. Multiple linear regression analysis, adjusted for demographic and clinical covariates, was used to evaluate the relationship between rural residence and QOL and impact of cancer.

Findings:

Among the 566 participants (83% response rate), rural residence was independently associated with lower SF-36 physical component summary scores and the physical function subscale (all P < .05). Rural residence was also associated with higher IOCv2 positive impact scores and the subscales of altruism/empathy and meaning of cancer scores in the adjusted models (all P < .05).

Conclusions:

Given documented rural cancer disparities and the lack of resources in rural communities, study findings support the continued need to provide supportive care to rural cancer survivors to improve their QOL. Consistent with previous research, rural residence status is associated with increased positive impact following cancer diagnosis.

Non-Hodgkin’s lymphoma (NHL) is a common cancer in the US with over 74,200 people being diagnosed in 2018.¹ Improved treatment has led to increased survival with an overall 5-year survival of 72%.² With this increased survival comes specific challenges that put lymphoma cancer survivors at risk for later physical and psychosocial adverse effects, including both long-term and late-effects that can negatively impact quality of life (QOL).³ NHL survivors report decreased QOL following cancer treatment that for some persists or worsens over time.^4-8 Decreased QOL in NHL has been associated with lower income, poor health behaviors, post-traumatic stress disorder (PTSD), health comorbidities, and lower social support.^7,9-12

Geographic location, including living in a rural area, has been associated with QOL in other types of cancers; however, less is known about the relationship between location and QOL in long-term NHL cancer survivors. In general, the research findings related to rural residence and QOL among other cancer types have been inconsistent. Some studies report lower QOL in rural patients while others report higher QOL compared to urban patients.^13-17 Many rural cancer patients have low socioeconomic status (SES), decreased educational opportunities, and lack of access to health care providers—all of which have been associated with decreased QOL.¹⁸ However, rural cancer survivors also may benefit from strong social networks and increased post-traumatic growth (or the positive change that occurs as a result of the struggle with highly challenging life crises such as cance), both of which may improve coping skills and QOL.^19,20,21 The unique relationship between rural residence and QOL in NHL long-term cancer survivors is a gap in the current literature. Given that NHL survivors often report high levels of PTSD following cancer treatment that negatively affect QOL,²² and the documented access to care issues that rural cancer survivors often face, this study can inform targeted outreach, assessment and treatment initiatives for this group of cancer survivors that are often underserved.

Rurality has also been related to the Impact of Cancer. The Impact of Cancer (IOC) scale was developed to measure the negative effects and the positive effects or life changes among cancer survivors.²³ Cancer survivors report both positive and negative impacts of cancer on their lives, with variations found in rural and urban populations.^19,23 McNutly and colleagues (2015)²⁴ reported that urban survivors had greater negative impact scores compared to their rural counterparts. Research also supports an increase in post-traumatic growth among rural cancer survivors compared to urban survivors. This suggests that the impact of cancer may be more positive for rural cancer survivors.¹⁹ Measuring the Impact of Cancer on survivors is an important tool for assessing survivor concerns that can impact QOL following cancer. The information gained can be used to develop interventions that improve QOL through addressing the negative impact as well as harnessing the positive impact following cancer. Given that QOL has been reported to be low in NHL survivors in general⁸ and the marked disparities in rural populations in cancer morbidity, mortality and access to services,¹⁸ understanding differences in the Impact of Cancer and QOL in rural-urban NHL cancer survivors has implications for targeting those most in need of increased resources and services to promote optimal well-being. We hypothesized that rural NHL cancer survivors would report: 1) lower levels of physical and mental health status and functioning (QOL) as measured by the SF36, and 2) higher levels of positive impact of cancer as measured by the Impact of Cancer Scale version 2 (IOCv2).

Methods

Design and Sample

The current study is a secondary analysis of a NHL survivor cohort study that was conducted in 2010 and approved by the Duke Cancer Institute and the University of North Carolina Lineberger Institutional Review Boards and detailed previously.⁷ This secondary analysis was based on 566 NHL survivors (83% response rate) who were at least 7 years post-diagnosis and 23 years of age or older.⁸ Informed written consent was obtained from all participants according to the Declaration of Helsinki.

Measures

Rural residence was determined by Rural-Urban Commuting Area (RUCA) codes designated as non-metropolitan (RUCA code of 4 or above).²⁵ We used the ZIP Code approximation of the Census tract-based RUCA codes.²⁶ Demographics and clinical characteristics were self-reported and included age, gender, race/ethnicity, marital status, education, employment status, treatment status, disease severity (Indolent vs. Aggressive), and years since diagnosis.

Comorbidities were assessed using the 11-item self-administered version of the Charlson Co-morbidity Index.²⁷ Life Style Factors included questions to evaluate healthy body mass index (BMI) and tobacco use. The calculation for BMI was derived from self-reported height and weight using the formula: weight (kg) / [height (m)]².²⁸ Tobacco use was assessed using questions from the Centers for Disease Control (CDC) Behavioral Risk Factor Surveillance System (BRFSS). The questions asked were: Have you smoked at least 100 cigarettes in your entire life?; Do you now smoke cigarettes every day, some days, or not at all?²⁹ For this study, responses were dichotomized into current smoker and non-smoker.

QOL domains related to health status and functioning were assessed using the Medical Outcomes Study Short Form-36 (SF-36 v2.0). This scale contains 36 items assessing 8 health concepts: physical functioning; role limitations due to physical health; bodily pain; social functioning; general mental health; role limitations due to emotional health; vitality; and general health perceptions. Mental component summary (MCS) and physical component summary (PCS) scores were arrived at by multiplying subscale z scores by norm-based factor coefficients for the MCS and PCS, respectively. Scores were normed across a 0–100 range. Scoring was completed by QualityMetric Health Outcomes scoring software (QualityMetric, Lincoln, RI) to generate norm-based scores where 50 represents the norm-based mean score for each subscale and summary score with a standard deviation of 10.^30,31 Higher scores indicate better health/function. Internal consistency has been reported to be high (≥ 0.70).³²

Impact of Cancer was assessed with the IOCv2, a 37-item survey that measures participants’ perceptions of the positive and negative impacts of cancer on their lives.³³ The IOCv2 consists of 4 negative (Appearance Concerns, Body Concerns, Life Interferences, and Worry) and 4 positive (Altruism/Empathy, Health Awareness, Meaning of Cancer, and Positive Self-evaluation) subscales that total to 2 summary scores (positive and negative). The range of scores for the subscales are 1–5 and the responses are averaged. Higher scores on the positive subscale indicate greater positive impacts, and higher scores on the negative subscale indicate greater negative impacts. Internal consistency has been reported to be high, ranging between 0.7 and 0.9 for each of the subscales.³⁴

Statistical Analyses

Descriptive statistics were used to examine participants’ demographic and clinical characteristics, SF-36 scores, and IOCv2 scores. Bivariate analyses (chi-square analyses for categorical variables and analysis of variance for continuous variables) were conducted to examine relationships between demographics, clinical characteristics, SF-36, IOCv2, and place of residence (rural vs. non-rural). All P values were 2-sided. A multivariable linear regression adjusted for demographic (age, race, marital status, income, education) and clinical characteristic (treatment status, disease severity, years since diagnosis, and comorbidities) variables was used to examine the relationship between rural and non-rural residence and SF-36 and IOCv2 subscales. Variables were chosen from their known associations with physical and mental health from prior research in NHL cancer survivors.^7,8 All analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC).

Results

In this sample of 566 NHL cancer survivors, 33% of the sample was rural, most of the sample was over 65 years old (61%), and there were slightly more females (51%) compared to males. Table 1 outlines the demographic and clinical characteristics by place of residence status. Bivariate results comparing rural and non-rural status on demographic and clinical characteristics are also outlined in Table 1. Survivors living in a rural area were more likely to be female (P < .05) and have a lower income (P < .001) and lower education (P < .001) compared to those living in a non-rural area.

Table 1.

Demographic and Clinical Characteristics byGgeographic Location

	Total (n = 566) n (%)	Rural (n = 187) n (%)	Non-rural (n = 379) n (%)	P
Demographics
Age
<50 years	53 (9.4)	13 (7.0)	40 (10.6)	.355
50–64 years	174 (30.7)	61 (32.6)	113 (29.8)
≥ 65 years	339 (59.9)	113 (59.6)	226 (60.4)
Gender
Male	272 (48.1)	77 (41.2)	195 (51.5)	.021
Female	294 (51.9)	110 (58.8)	184 (48.5)
Race
White	494 (87.3)	156 (83.4)	338 (89.2)	.053
Non-white	72 (12.7)	31 (16.6)	41 (10.8)
Ethnicity
Hispanic	7 (1.2)	1 (1.0)	6 (1.6)	.289
Non-Hispanic	559 (98.8)	186 (99.0)	373 (98.4)
Income
< $30,000	131 (25.4)	66 (35.3)	65 (17.2)	<.001
≥$30,000	435 (76.86)	121 (64.7)	314 (82.9)
Education
Less than college	316 (55.8)	132 (70.6)	184 (48.6)	<.001
College or post-grad	250 (44.2	55 (29.4)	195 (51.5)
Marital Status
Married or living with partner	431 (77)	138 (75.4)	293 (77.7)	.543
Not married	129 (23)	45 (24.6)	84 (22.3)
Health Insurance Coverage
Yes	538 (96.1)	174 (94.6)	364 (96.8)	.199
No	22 (3.9)	10 (5.4)	12 (3.2)
Clinical Characteristics
Lymphoma Type
Indolent	270 (50.3)	92 (52.6)	178 (49.2)	.460
Aggressive	267 (49.7)	83 (47.4)	184 (50.8)
Treatment Status
Not in treatment	516 (92.8)	172 (93.5)	344 (92.5)	.666
Receiving treatment	40 (7.2)	12 (6.5)	28 (7.5)
Had a recurrence	109 (19.3)	35 (18.7)	74 (19.5)	.819
Years Since Diagnosis m(SD)	15.2 (7.19)	15.6 (7.2)	15.1 (7.2)	.429
Current smoker	34 (6.0)	9 (4.8)	25 (6.6)	.401
	m (SD)	m (SD)	m (SD)
Co-morbidities	2.78 (2.1)	2.98 (2.1)	2.67 (2.1)	.103
Body Mass Index	27.13 (5.7)	27.65 (6.3)	26.88 (5.4)	.133
Medical Outcomes Study Short Form-36
Physical Component Score	45.00 (10.96)	42.33 (11.60)	46.27 (10.41)	<.001
Mental Component Score	49.95 (10.86)	50.38 (11.11)	49.75 (10.75)	.519
Impact of Cancer
Positive Impact Score	3.38 (0.79)	3.56 (0.76)	3.29 (0.79)	<.001
Negative Impact Score	1.90 (0.75)	1.92 (0.80)	1.89 (0.73)	.678

Note: Bold P values designate statistically significant

Regarding QOL, rural survivors reported lower mean SF-36 PCS scores compared to non-rural survivors (P < .001). There were no significant differences between rural and non-rural survivors on mean SF-36 MCS scores (Table 1). There were significant differences in mean PCS, with rural residents reporting lower scores compared to non-rural survivors for physical function (P < .001); role limitations, physical (P = .001); bodily pain (P = .015); and general health (P = .039). (Figure 1, available online only.)

Figure 1.

Differences in QOL Subscale Scores by Rural Residence

PF= Physical Function; RP= Role Limitations, Physical; BP= Bodily Pain; GH= General Health; VT= Vitality; SF= Social Functioning; RE= Role Limitations, Emotional; MH= Mental Health; PCS= Physical Component Score; MCS= Mental Component Score

Higher scores reflect better QOL

Significant differences were reported for the mean IOCv2 positive impact summary scores between rural and non-rural survivors, with rural survivors reporting an overall higher score compared to non-rural survivors (P = <.001). (See Table 1.) There were no significant differences in the mean overall Negative Impact Summary score between rural and non-rural survivors. As displayed in Figure 2 (available online only), there were significant differences in the mean IOCv2 subscales, with rural survivors reporting higher scores for altruism and empathy (P = <.001), health awareness (P = .045), meaning of cancer (P = <.001), and positive self-evaluation (P = .021).

Figure 2.

Differences in Impact of Cancer Scores by Rural Residence

PIS= Positive Impact Score; AE= Altruism and Empathy; HA= Health Awareness; MC= Meaning of Cancer; PE= Positive Self-Evaluation; NIS= Negative Impact Score; AC= Appearance Concerns; BC= Body Changes Concern; LI= Life Interferences; WO= Worry

Higher scores indicate more positive or negative impact

Multivariable adjusted models show relationships between the SF-36 PCS and the IOCv2 positive impact summary scores and rural residence (see Tables 2 and 3). In the adjusted model, rural residence was significantly associated with lower SF-36 PCS scores (t = 2.30, P = .0217) and higher IOCv2 positive impact summary scores (t = −2.41, P = .0164). In the adjusted models for the SF-36 PCS subscales, rural residence was significantly correlated with lower physical function (t = 2.53, P = .0116) and was no longer associated with role limitations, bodily pain, and general health. In the adjusted models for the IOCv2 positive impact subscales, rural residence was significantly associated with higher altruism/empathy (t = −2.64, P = .0086) and meaning of cancer scores (t = −3.28, P = .0011). Rural residence was no longer significant in the adjusted models for general health and bodily pain or for the positive self-evaluation subscale.

Table 2.

SF-36^a Mean Adjusted Scores

	Estimateb	T value	P value
SF-36 Physical Component Score
Rural	39.97	2.30	.0217
Non-rural	41.80
SF-36 Physical Function Subdomain
Rural	38.53	2.53	.0116
Non-rural	40.79
SF-36 Role Limitations, Physical Subdomain
Rural	40.53	1.53	.1271
Non-rural	41.69
SF-36 Bodily Pain Subdomain
Rural	46.20	0.22	.8221
Non-rural	46.40
SF-36 General Health Subdomain
Rural	42.75	0.62	.5329
Non-rural	43.33

^aSF-36: Medical Outcomes Survey Short Form-36

^bAdjusted for age, race, gender, income, partnered status, comorbidities, lymphoma type, treatment status and years since diagnosis.

Note: Bold P values designate statistically significant

Table 3.

IOC^a Mean Adjusted Scores

	Estimateb	T value	P value
IOC Positive Impact Score
Rural	3.54	-2.41	.0164
Non-rural	3.35
IOC Altruism/Empathy Subdomain
Rural	3.94	-2.64	.0086
Non-rural	3.71
IOC Health Awareness Subdomain
Rural	3.78	-1.60	0.1094
Non-rural	3.63
IOC Meaning of Cancer Subdomain
Rural	2.80	-3.28	.0011
Non-rural	2.48
IOC Positive Self Evaluation Subdomain
Rural	3.92	-1.54	.1247
Non-rural	3.77

^aIOC: Impact of Cancer

^bAdjusted for age, race, gender, income, partnered status, comorbidities, lymphoma type, treatment status and years since diagnosis.

Note: Bold P values designate statistically significant

Discussion

The results of this study partially confirmed our hypotheses, that rural NHL cancer survivors report: 1) lower levels of physical QOL as measured by SF36; and 2) higher levels of positive impact of cancer as measured by the IOCv2. In this sample of 566 NHL cancer survivors, we found that rural survivors had lower SF-36 PCS scores and this association remained significant after adjusting for covariates. However, the mean MCS scores were not different based on rural residence as we had hypothesized. Consistent with our hypothesis, rural cancer survivors had higher IOCv2 positive impact scores, and rural residence remained significantly associated with IOCv2 positive impact scores after adjusting for other possible covariates, although there were no differences in negative impact summary scores between the 2 groups.

Our results add to the evidence on the importance of considering place (eg, rural residence) and its relationship to QOL.^18,20,35-37 Although previous literature has been mixed on this relationship in cancer survivors, our study is one of the first to examine this association in long-term NHL cancer survivors. NHL survivors in general have been shown to have lower levels of QOL compared to the age-adjusted population mean.⁷ In our analysis, lower mean SF-36 physical component scores were found to be significantly associated, even after adjusting for covariates with rural residence. Cancer survivors are more likely to report poorer QOL and functional impairments compared to adults without cancer.³⁷ Importantly, our analysis added that this is more likely to occur in those NHL survivors living in rural areas. Lack of health services in rural communities and transportation challenges including long travel times to health services^18,37 may explain the increased physical symptom burden in rural NHL survivors. Interventions addressing the physical symptoms that rural NHL survivors experience are imperative to improving QOL in this group. Given the lack of health resources in many rural communities, developing technology-based and navigator-based outreach programs that can reach underserved rural NHL survivors with lingering physical symptoms may decrease morbidity and increase QOL in these rural cancer survivors.

Unexpectedly there were no significant differences between the mean SF-36 mental health component scores between rural and non-rural cancer survivors. This conflicts with prior research reporting poorer mental health in rural compared to urban cancer survivors.^20,22,36,37 Further, these results are surprising given physical health tends to directly influence mental health and rural NHL survivors in this study reported lower overall general health status and higher physical component scores. These results may be explained by the fact that most of the survivors in this study were not in treatment and had moved beyond the immediate post-treatment period. Prior research in NHL survivors indicates that survivors who are off treatment report better mental health component scores compared to those in active treatment.⁷ However, in other groups of cancer survivors results have been mixed, reporting both improvements and declines in mental health longitudinally post-treatment.^38,39 These inconsistencies support the need for further research in this area and the consideration of rural/urban residence as a variable in these findings. Continued monitoring of mental health symptoms in rural NHL survivors is still warranted, as well as supportive treatment to continue to bolster mental health wellness in NHL survivors.

Our results also add to the evidence on the importance of considering place (eg, rural residence) and its relationship to the impact of cancer on survivors. Rural survivors in our study reported higher scores on the IOCv2 positive impact overall summary and the positive impact subscales of altruism/empathy, health awareness, positive self-evaluation, and meaning of cancer. The association of rural residence and the subscales of altruism/empathy and meaning of cancer remained significant in the adjusted regression models. This finding suggests that post-traumatic growth or finding positive meaning in the traumatic cancer experience is occurring in rural NHL survivors. Posttraumatic growth has been reported in other cancer survivor groups,⁴⁰ most notably breast cancer survivors. Further, rural lung cancer survivors reported significantly higher levels of post-traumatic growth compared to urban survivors.¹⁹ Our findings in rural NHL survivors could be explained by the fact that many rural communities have strong community support networks and connections to faith-based groups.^38,40 Rural survivors in our sample may have benefited from these connections, thus improving their overall quality of life during and after cancer treatment. Interventions should continue to harness the positive impacts of cancer in rural NHL survivors to increase overall QOL.

Limitations

Strengths of this study include a large sample of NHL survivors, including a modest portion of rural survivors to conduct rural and non-rural comparisons. To our knowledge this is the first study to examine the relationship between rural residence and QOL and impact of cancer in long-term NHL survivors. Limitations of this study include the cross-sectional design, which limited our ability to conclude causality between rural residence and QOL and impact of cancer. Further, although we adjusted for many covariates, there may be other variables that confound the relationship between rural residence and QOL and impact of cancer not accounted for in our models. Finally, our sample includes NHL survivors from only 2 comprehensive cancer centers in the south, limiting the generalizability of our results to rural NHL survivors in other US regions.

Implications and Conclusions

Understanding the impact of cancer in long-term NHL survivors has implications for QOL and psychosocial service and educational needs in this population. Our study serves as a starting point to tailor these services to the needs of rural NHL survivors. Given rural disparities in cancer mortality and a documented lack of resources and services that are available in rural communities,^18,37 the results of this study also indicate the continued need to provide supportive care to rural cancer survivors to improve their physical well-being. Partnering with and leveraging existing social and community supports in rural communities remain viable strategies to improve QOL in rural cancer survivors. In conclusion, the results of the current study indicate that rural residence is associated with decreased SF-36 PCS as well as increased IOCv2 positive impact scores compared to non-rural NHL survivors. Consistent with previous research we found that rural residence continues to convey benefits associated with positive impacts following a cancer diagnosis. Future studies should continue to examine the relationship between rural residence and outcomes among cancer survivors in order to target interventions and resources equitability.