Detection of later stage breast cancer in first nations women in Ontario, Canada.
Sheppard, Amanda J. ; Chiarelli, Anna M. ; Marrett, Loraine D. 等
Breast cancer is the most common cancer among women in Ontario,
with an estimated 8,700 cases diagnosed in 2009. (1) The
age-standardized incidence rate for breast cancer among Ontario women is
102 per 100,000. (1) While this rate has stabilized across the province,
that is not the case for all populations within the province. In fact,
while still significantly lower than in the general population, the
incidence of breast cancer in Ontario's First Nations (FN) people
is increasing, (2) and "survival following a diagnosis is
significantly worse" (Dr. Loraine Marrett, April 27, 2009). Little
research has been conducted investigating cancer patterns of Indigenous
populations (including FN, Metis or Inuit populations) in Canada. When
examined, lower incidence and mortality rates for Indigenous compared to
non-Indigenous populations are found for all cancers combined and for
many specific sites. (2-5)
Increasingly, there is a growing literature outside of Canada on
breast cancer prognosis comparing Indigenous to non-Indigenous
populations within the same geographical area. All of these studies
found poorer breast cancer survival among the Indigenous populations.
(6-15) Some studies comparing the distribution of stage at diagnosis
have found the Indigenous populations were more often diagnosed at a
later stage than respective non-Indigenous reference groups,
(6,9,10,12,14,15) however others have found no difference in the
distribution of breast cancer stage. (7,8,13) Since the leading
determinant of breast cancer survival is the stage in which cancer is
detected, (16,17) it is important to understand this distribution in a
population whose incidence is on the rise. (2) The purpose of this study
was to compare the distribution of stage at breast cancer diagnosis
between FN and non-FN women, and to investigate factors associated with
later diagnosis in FN women.
METHODS
Study population
As race or ethnicity is not recorded in the Ontario Cancer Registry
(OCR), it is not possible to routinely estimate cancer incidence or
mortality rates for any Aboriginal population directly from OCR data.
Therefore, a cohort of Ontario FN people was created through linking of
the OCR and mortality files with the Ontario files of FN people
'with Status' from Indian and Northern Affairs Canada (from
1968 to 1991). (2)
Our study employed a case-case design, comparing FN women diagnosed
with invasive breast cancer in 1995-2004 and seen at one of
Ontario's specialized cancer centres, known as Regional Cancer
Centres (RCCs), with a frequency-matched random sample of general
population women diagnosed in the same time period. (18) The RCCs
deliver all cancer radiotherapy in the province and are often involved
in the treatment and planning. All FN women satisfying study eligibility
criteria were included; the general population sample were frequency
matched (2:1) on the following design variables: period of diagnosis
(1995-1999 and 2000-2004), age at diagnosis (<50 vs. [greater than or
equal to]50), and RCC first attended.
Data collection
Stage and data relevant to the determinants of stage were collected
from medical charts at the RCCs. Two trained chart abstractors reviewed
centre charts to abstract all data elements manually between October
2007 to August 2008. A pilot test revealed 94% accuracy between the two
abstractors pre-study.
Variable definitions
Breast cancer was defined as a diagnosis of primary invasive breast
cancer of any histological type. We utilized the American Joint
Committee on Cancer TNM classification scheme for staging breast
cancers. (19) The TNM system is an expression of the anatomic extent of
disease and is based on three components: T- the extent of the primary
tumour, N- the absence or presence and extent of regional lymph node
involvement and M- the absence or presence of distant metastasis. Stage
at diagnosis was aggregated into a binary variable (stage I and stages
Distance to RCC was computed by calculating the straight line distance
(to the nearest kilometre) between a RCC and a woman's residence
(postal code) at diagnosis. (20) Method of detection was grouped as
'screened' for women whose breast cancer was detected through
participation in either routine mammography or the Ontario Breast
Screening Program. Those considered 'not screened' included
women whose breast cancer was detected by themselves, by a physician or
other health professional. Family history was defined as having a
first-degree relative with breast cancer and/or ovarian cancer. Body
mass index was calculated as weight in kg divided by height in [m.sup.2]
and was defined as (normal weight = BMI <25kg/[m.sup.2]; overweight =
BMI [greater than or equal to]25 to <30kg/[m.sup.2]; obese = BMI
[greater than or equal to]30kg/[m.sup.2]). (21) Comorbidity was defined
as having concurrent health conditions at the time of diagnosis that are
likely to influence the probability of mortality, decrease the adherence
to the treatment regime or result in procedural or drug
contraindication. The 17 conditions considered were those specified in
the Charlson index. (22) Because of the risk of current and/or former
active cigarette smoking on breast cancer survival, (23) we grouped all
the 'ever smokers' together.
Statistical analysis
The distribution of design variables used for matching, including
age at diagnosis (<50 vs. [greater than or equal to]50), period of
diagnosis (1995-1999, 2000-2004), and RCC location (see list in Table
1), was compared between FN and non-FN women using a Pearson chi-square
test. Logistic regression analyses examined associations of FN status
with stage (I and II+) at breast cancer diagnosis, and method of
detection (screened, non-screened), since it is an important determinant
of stage, as well as potential confounders (including distance to a RCC,
BMI, comorbidity, and smoking status). Odds ratios (OR) and
corresponding 95% confidence intervals were computed from models
adjusted for matching variables. The primary OR of interest relating
stage to FN status was also estimated in multivariate analysis adjusted
for matching variables, method of detection and potential confounders.
For FN women, exploratory unadjusted and multivariate models were
developed to examine the association of stage at diagnosis with
variables of interest (including distance to a RCC, method of detection,
BMI, and comorbidity). All analyses were conducted using SAS version 9.1
(SAS Institute) and statistical significance of all tests was evaluated
using two-sided p-values.
Ethical approval was obtained from the Human Subjects Ethics Review
Committee of the University of Toronto. In addition, ethics for the
chart abstractions was obtained from universities and hospitals
affiliated with each of the RCCs.
RESULTS
We identified 309 eligible FN women in the OCR diagnosed in
1995-2004, compared to 55,501 women diagnosed with breast cancer in the
same time period and age range in the general population. Charts were
available for 287 (93%) of the FN women, and 671 (90%) of the selected
743 non-FN women.
The distribution of matching variables was similar between the FN
and non-FN women (Table 1). In logistic regression analyses, adjusted
only for the matching variables, a statistically significant association
was found for FN status and later stage at diagnosis (OR 1.55, 95% CI
1.15-2.08). FN women (66%) were more likely to be diagnosed at a later
stage than non-FN women (56%). Other significant associations revealed
that non-FN women were more likely to have their breast cancer diagnosed
by screening than FN women, FN women traveled further to a RCC compared
to non-FN women, FN women were more likely to have a higher BMI,
presence of comorbidity and a history of smoking (Table 2).
After controlling for the design variables, method of detection and
potential confounders, the OR estimate for stage is reduced from 1.55 to
1.21 and while still suggesting an increased odds of later stage of
diagnosis for FN women, this is no longer statistically significant
(Table 2). Results of distance to RCC, BMI, comorbidity and smoking
status remained significant in the multivariate analysis.
Unadjusted models demonstrated that age of diagnosis, method of
detection, BMI, comorbidity and smoking status were independently
associated with stage at diagnosis (stage II+ vs. stage I) for FN women
(Table 3). The multivariate analysis modeling stage in FN women shows
that women who were not screened were 5 times more likely to detect
their breast cancer at a later stage (OR 4.97, 95% CI 2.44-10.15). FN
women with above normal weight at diagnosis had about a 3 to 5 times
increased odds of a later diagnosis. The presence of comorbidity was
significantly associated with reduced odds of a later stage for breast
cancer diagnosis (OR 0.51, 95% CI 0.27-0.96).
DISCUSSION
This study reveals that in Ontario between 1995-2004, FN women were
diagnosed with a later stage of breast cancer more often than non-FN
women. These findings are consistent with those of previous studies
investigating other Indigenous populations in Australia, New Zealand and
the United States. (6,9,10,12,14,15,24)
Determinants of breast cancer stage at diagnosis have not
previously been studied in Canadian Indigenous populations. In the
Ontario FN population, significant associations were found for later
stage with detection method, BMI and comorbidity. Although screening by
mammography is a known determinant of stage irrespective of race or
ethnicity, (25) as was supported by our findings, we found that FN women
were more likely to be diagnosed with a non-screened cancer. This
warrants further investigation of screening practices, including
increasing education in FN communities and improving access for FN women
in Ontario.
Consistent with our results, BMI has also been determined to impact
later stage at diagnosis in an urban Native American population. (9) In
addition, as we found with Ontario FN women, increased levels of BMI
have been reported in other Indigenous populations and have been
recognized as a determinant of disease in these populations. (26)
Twice as many FN women had at least one comorbidity at diagnosis
compared to non-FN women, and the most reported condition was diabetes,
occurring in 23% of FN and 7% of non-FN women. However, the presence of
comorbidity reduced the risk of later stage of breast cancer diagnosis.
Similar analysis using the non-FN sample of women found that the
presence of comorbidity significantly increased the risk of a later
stage at diagnosis (results not shown). The literature has suggested
associations between certain types of comorbidities and stage at breast
cancer diagnosis. In an analysis of the Surveillance, Epidemiology and
End Results program data in the United States, women with cardiovascular
conditions had a reduced risk of later stage diagnosis, compared to
women with, for instance, diabetes who had an increased risk of later
stage breast cancer. (27) This is not consistent with our finding among
FN women and therefore this result warrants further exploration of this
relationship. Our findings do suggest possible differential use of the
health care system by FN and non-FN women in Ontario. System and
personal characteristics such as a lack of continuity of care, high
health care worker turnover, (28) perceived care options, (29) low
awareness of early symptoms, tardiness in seeking medical advice, (24)
and a lack of culturally appropriate screening programs (30) have all
been identified as barriers to cancer care in an Indigenous context.
Health promotion programs may need to better target FN women, in
particular those who are not accessing the health care system, to
promote increased awareness and participation in screening programs in
order to detect cancers when they are small and more curable.
There were a few limitations in this study. Only women who attended
a RCC were eligible for inclusion (85% of all FN and 80% of all non-FN
women with breast cancer during this study period). Ethnicity was
identified from a database linkage, therefore the non-Status FN
population could have included other FN women. However, upon abstraction
of medical charts, we identified only one non-FN woman as being
'Aboriginal'. Second, although we collected information on
distance to a RCC, we did not have access to data on distance to a
screening centre, which may have been a more important factor for stage
at diagnosis.
Stage at diagnosis is the most important determinant of breast
cancer survival and the findings from this study that Ontario FN women
are more likely to be diagnosed at a later stage than non-FN women, may
explain the poorer survival rate observed in FN women. Other factors,
with observed differences between FN and non-FN women may contribute to
the differential survival. Future study is needed to identify survival
determinants in FN women adjusting for stage at diagnosis.
Acknowledgements: The authors thank the study staff: Lindsay
Stewart, Sarah Ahmad and Tanya Cecic. This research was endorsed by the
Joint Ontario Aboriginal Cancer Committee comprised of Aboriginal
representatives from each of the nine provincial/territorial Aboriginal
organizations (PTOs), including a representative from the Independent
First Nations and representatives from Cancer Care Ontario. We also
thank the participants of the Aboriginal Breast Cancer Care Workshop
(Toronto, ON, April 2009) who shared their relevant experiences and
contributed their interpretations of the study findings.
The study was funded by a grant from the Canadian Breast Cancer
Foundation Ontario Region. Amanda Sheppard's doctoral training was
supported by an Indigenous Health Research Development Program
scholarship, a Strategic Training Program in the Trandisciplinary
Approach to the Health of Marginalized Populations Fellowship, and a
Doctoral Award from Canadian Breast Cancer Foundation--Ontario Region.
Received: August 14, 2009 Accepted: February 4, 2010
REFERENCES
(1.) Canadian Cancer Society/National Cancer Institute of Canada.
Canadian Cancer Statistics 2009, Toronto.
(2.) Marrett LD, Chaudhry M. Cancer incidence and mortality in
Ontario First Nations, 1968-1991 (Canada). Cancer Causes Control
2003;14(3):259-68.
(3.) Health Canada. A Statistical Profile on the Health of First
Nations in Canada. First Nations and Inuit Health Branch, 2003.
(4.) Mahoney MC, Michalek AM. A meta-analysis of cancer incidence
in United States and Canadian native populations. Int J Epidemiol
1991;20(2):323-27.
(5.) Rosenberg T, Martel S. Cancer trends from 1972-1991 for
Registered Indians living on Manitoba Reserves. Int J Circumpolar Health
1998;57(Suppl 1):391-98.
(6.) Condon JR, Barnes T, Armstrong BK, Selva-Nayagam S, Elwood JM.
Stage at diagnosis and cancer survival for Indigenous Australians in the
Northern Territory. Med J Aust 2005;182(6):277-80.
(7.) Samet JM, Key CR, Hunt WC, Goodwin JS. Survival of American
Indian and Hispanic cancer patients in New Mexico and Arizona, 1969-82.
J Natl Cancer Inst 1987;79(3):457-63.
(8.) Dennis TD. Cancer stage at diagnosis, treatment, and survival
among American Indians and non-American Indians in Montana. Cancer
2000;89(1):181-86.
(9.) Tillman L, Myers S, Pockaj B, Perry C, Bay RC, Al-kasspooles
M. Breast cancer in Native American women treated at an urban-based
Indian health referral center 1982-2003. Am J Surg 2005;190(6):895-902.
(10.) Wampler NS, Lash TL, Silliman RA, Heeren TC. Breast cancer
survival of American Indian/Alaska Native women, 1973-1996. Soz
Praventivmed 2005;50(4):230-37.
(11.) Jeffreys M, Stevanovic V, Tobias M, Lewis C,
Ellison-Loschmann L, Pearce N, et al. Ethnic inequalities in cancer
survival in New Zealand: Linkage study. Am J Public Health
2005;95(5):834-37.
(12.) Maskarinec G, Pagano IS, Yamashiro G, Issell BF. Influences
of ethnicity, treatment, and comorbidity on breast cancer survival in
Hawaii. J Clin Epidemiol 2003;56(7):678-85.
(13.) Sugarman JR, Dennis LK, White E. Cancer survival among
American Indians in western Washington State (United States). Cancer
Causes Control 1994;5(5):440-48.
(14.) Frost F, Tollestrup K, Hunt WC, Gilliland F, Key CR, Urbina
CE. Breast cancer survival among New Mexico Hispanic, American Indian,
and non-Hispanic white women (1973-1992). Cancer Epidemiol Biomarkers
Prev 1996;5(11):861-66.
(15.) Valery PC, Coory M, Stirling J, Green AC. Cancer diagnosis,
treatment, and survival in Indigenous and non-Indigenous Australians: A
matched cohort study. Lancet 2006;367(9525):1842-48.
(16.) Rosenberg J, Chia YL, Plevritis S. The effect of age, race,
tumor size, tumor grade, and disease stage on invasive ductal breast
cancer survival in the U.S. SEER database. Breast Cancer Res Treat
2005;89(1):47-54.
(17.) McBride R, Hershman D, Tsai WY, Jacobson JS, Grann V, Neugut
AI. Within stage racial differences in tumor size and number of positive
lymph nodes in women with breast cancer. Cancer 2007;110(6):1201-8.
(18.) Begg CB, Zhang ZF. Statistical analysis of molecular
epidemiology studies employing case-series. Cancer Epidemiol Biomarkers
Prev 1994;3(2):173-75.
(19.) American Joint Committee on Cancer. Cancer Staging Manual.
New York, 2002.
(20.) Groome PA, Schulze KM, Keller S, Mackillop WJ. Demographic
differences between cancer survivors and those who die quickly of their
disease. Clin Oncol (R Coll Radiol) 2008;20(8):647-56.
(21.) National Institutes of Health: Clinical Guidelines on the
Identification, Evaluation, and Treatment of Overweight and Obesity in
Adults-The Evidence Report. National Institutes of Health. Obes Res
1998;6(Suppl 2):51S-209S.
(22.) Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of
classifying prognostic comorbidity in longitudinal studies: Development
and validation. J Chronic Dis 1987;40(5):373-83.
(23.) Sagiv SK, Gaudet MM, Eng SM, Abrahamson PE, Shantakumar S,
Teitelbaum SL, et al. Active and passive cigarette smoke and breast
cancer survival. Ann Epidemiol 2007;17(5):385-93.
(24.) Condon JR, Cunningham J, Barnes T, Armstrong BK,
Selva-Nayagam S. Cancer diagnosis and treatment in the Northern
Territory: Assessing health service performance for indigenous
Australians. Intern Med J 2006;36(8):498-505.
(25.) Jacobellis J, Cutter G. Mammography screening and differences
in stage of disease by race/ethnicity. Am J Public Health
2002;92(7):1144-50.
(26.) Gracey M, King M. Indigenous health part 1: Determinants and
disease patterns. Lancet 2009;374(9683):65-75.
(27.) Fleming ST, Pursley HG, Newman B, Pavlov D, Chen K.
Comorbidity as a predictor of stage of illness for patients with breast
cancer. Med Care 2005;43(2):132-40.
(28.) Minore B, Boone M, Katt M, Kinch P, Birch S, Mushquash C. The
effects of nursing turnover on continuity of care in isolated First
Nation communities. Can JNurs Res 2005;37(1):86-100.
(29.) Minore B, Boone M, Katt M, Kinch P, Cromarty H. How clients
choices influence cancer care in northern Aboriginal communities. Int J
Circumpolar Health 2004;63(Suppl 2):129-32.
(30.) Friedman DB, Hoffman-Goetz L. Assessment of cultural
sensitivity of cancer information in ethnic print media. J Health Commun
2006;11(4):425-47.
Amanda J. Sheppard, MSc, [1,2] Anna M. Chiarelli, PhD, [1-3]
Loraine D. Marrett, PhD, [1-3] Lucia Mirea, MSc, [2,3] E. Diane Nishri,
MSc, [2] Maureen E. Trudeau, MD, [4] and the Aboriginal Breast Cancer
Study Group [5]
Author Affiliations
[1.] Institute of Medical Science, University of Toronto, Toronto,
ON
[2.] Population Studies and Surveillance, Cancer Care Ontario,
Toronto, ON
[3.] Dalla Lana School of Public Health, University of Toronto,
Toronto, ON
[4.] Division of Medical Hematology-Oncology, Odette Cancer Centre,
Toronto, ON
[5.] Members include: Norman Boyd, Princess Margaret Hospital; Paul
Ferner, London Regional Cancer Centre; Amanda Hey, Hopital regional de
Sudbury Regional Hospital--Regional Cancer Program; Caroline
Lidstone-Jones, Cancer Care Ontario; Yolanda Madarnas, Cancer Centre of
Southeastern Ontario; Alison McMullen, Thunder Bay Regional Health
Sciences Centre; John Shaw, Windsor Regional Cancer Program; Carol Rand,
Juravinski Cancer Centre; Vincent Young, Ottawa Hospital Cancer Centre
Correspondence: Amanda J. Sheppard, Population Studies and
Surveillance, Cancer Care Ontario, 620 University Avenue, Toronto, ON
M5G 2L7, Tel: 416-971-9800, Fax: 416-971-6888, E-mail:
amanda.sheppard@cancercare.on.ca
Table 1. Distribution of Design Variables Used for Matching of First
Nations and Non-First Nations Women Diagnosed with Breast Cancer
between 1995-2004 in Ontario, Canada
Factors First Nations Non-First
Women Nations Women
(n=287) (n=671)
N (%) N (%)
Age at diagnosis (years)
<50 104 (36) 242 (36)
[greater than or equal to] 50 183 (64) 429 (64)
Mean (sd) 55.9 (12.8) 57.7 (14.1)
Year of diagnosis
1995-1999 136 (47) 317 (47)
2000-2004 151 (53) 354 (53)
Regional Cancer Centre
Southern Centres
Ottawa Hospital Cancer Centre 21 (7) 51 (8)
Cancer Centre of Southeastern
Ontario (Kingston) 34 (12) 78 (12)
Odette Cancer Centre (Toronto) 16 (6) 33 (5)
Princess Margaret 18 (6) 32 (5)
Hospital (Toronto)
Juravinski Cancer 54 (19) 117 (17)
Centre (Hamilton)
Southwest Regional Cancer 23 (8) 57 (8)
Centre (London)
Windsor Regional Cancer 2 (1) 5 (1)
Centre
Northern Centres
Northeast Regional Cancer 61 (21) 141 (21)
Centre (Sudbury)
Northwestern Regional Cancer
Centre (Thunder Bay) 58 (20) 157 (23)
Table 2. Odds Ratios (OR) and 95% Confidence Intervals (CI) Comparing
Stage at Diagnosis (Stage I vs. II+) and Risk Factors between First
Nations and Non-First Nations Women Diagnosed with Breast Cancer
between 1995-2004 in Ontario, Canada (n=958)
Diagnostic Factors First Nations Women Non-First
(n=287) Nations Women
N (%) (n=671)
N (%)
Stage at diagnosis
Stage I 95 (34) 284 (44)
Stage II+ 188 (66) 366 (56)
Unstageable 4 21
Method of detection
Screened 64 (23) 192 (30)
Non-screened 217 (77) 456 (70)
Unknown 6 23
Distance to Regional
Cancer Centre
0-15 km 53 (18) 266 (39)
>15-100 km 98 (34) 217 (32)
>100 km 136 (47) 188 (28)
Body mass index (BMI)
Normal weight 44 (17) 211 (37)
(<25kg/[m.sup.2])
Overweight (>25 to 85 (33) 194 (34)
<30kg/[m.sup.2])
Obese ([greater than 128 (50) 163 (29)
or equal to]
30kg/[m.sup.2])
Unknown 30 103
Comorbidity
No 180 (64) 534 (82)
Yes 102 (36) 118 (18)
Unknown 5 19
Smoking status
Never smoker 81 (32) 296 (51)
Ever smoker 172 (68) 284 (49)
Unknown 34 91
Diagnostic Factors Adjusted Adjusted (1)
([parallel]) OR (95% CI)
OR (95% CI)
Stage at diagnosis
Stage I 1.00 1.00
Stage II+ 1.55 (1.16-2.08) 1.21 (0.81-1.81)
([double dagger])
Unstageable
Method of detection
Screened 1.00 1.00
Non-screened 1.51 (1.08-2.13) 1.51 (0.96-2.37)
([section])
Unknown
Distance to Regional
Cancer Centre
0-15 km 1.00 1.00
>15-100 km 2.11 (1.41-3.14) 2.12 (1.27-3.54)
([dagger]) ([dagger])
>100 km 5.18 (3.39-7.90) * 5.09 (3.03-8.56) *
Body mass index (BMI)
Normal weight 1.00 1.00
(<25kg/[m.sup.2])
Overweight (>25 to 2.22 (1.46-3.39) 2.46 (1.49-4.07)
([dagger]) ([dagger])
<30kg/[m.sup.2])
Obese ([greater than 3.98 (2.64-6.00) * 3.80 (2.32-6.23) *
or equal to]
30kg/[m.sup.2])
Unknown
Comorbidity
No 1.00 1.00
Yes 2.80 (2.01-3.90) * 2.48 (1.60-3.86) *
Unknown
Smoking status
Never smoker 1.00 1.00
Ever smoker 2.21 (1.62-3.02) * 2.34 (1.60-3.43) *
Unknown
Unknown and unstageable data are not included in the statistical
tests.
([parallel]) Models were adjusted for age at diagnosis, period at
diagnosis, and RCC
([paragraph]) Model was adjusted for age at diagnosis, period at
diagnosis, stage at diagnosis, method of detection, istance to RCC,
BMI, comorbidity, and smoking status
* p-value <0.0001
([dagger]) p-value <0.001
([double dagger]) p-value <0.01
([section]) p-value <0.05
Table 3. Odds Ratios (OR) and 95% Confidence Intervals (CI) for Risk
of Stage of Breast Cancer Diagnosis (Stage II+ vs. Stage I) in First
Nations Women (n=212) Diagnosed between 1995-2004 in Ontario, Canada
Diagnostic Factors Unadjusted Adjusted
OR (95% CI) ([parallel])
OR (95% CI)
Age at diagnosis (years)
[greater than or 1.00 1.00
equal to] 50
<50 1.88 (1.10-3.22) 1.05 (0.53-2.10)
([section])
Method of detection
Screened 1.00 1.00
Non-screened 4.02 (2.24-7.22) * 5.03 (2.48-10.21) *
Distance to Regional
Cancer Centre
0-15 km 1.00 1.00
>15-100 km 0.95 (0.47-1.91) 0.85 (0.37-1.96)
>100 km 1.39(0.71-2.73) 1.56 (0.70-3.47)
Body mass index (BMI)
Normal weight 1.00 1.00
(<25kg/[m.sup.2])
Overweight (>25 to 2.24 (1.05-4.79) 2.98 (1.27-6.98)
<30kg/[m.sup.2]) ([section]) ([section])
Obese (>30kg/m2) 3.03 (1.47-6.26) 4.46 (1.95-10.21)
([dagger]) ([dagger])
Comorbidity
No 1.00 1.00
Yes 0.59 (0.36-0.99) 0.51 (0.27-0.96)
([section]) ([section])
([parallel]) Model was adjusted for age at diagnosis, method of
detection, distance to RCC, BMI, and comorbidity
* p-value <0.0001
([dagger]) p-value <0.001
([double dagger]) p-value <0.01
([section]) p-value <0.05
