Outcomes of preexisting diabetes mellitus in breast, colorectal, and prostate cancer

Abstract

PurposePreexisting diabetes is associated with increased morbidity and mortality in cancer. We examined the impact of incident cancer on the long-term outcomes of diabetes. Methods
Using the United Kingdom Clinical Practice Research Datalink, we identified three cohorts of diabetes patients subsequently diagnosed with breast, colorectal, or prostate cancer, each matched to diabetic noncancer controls. Patients were required to have survived at least 1 year after cancer diagnosis (cases) or a matched index date (controls), and were followed up to 10 years for incident microvascular and macrovascular complications and mortality. Multivariate competing risks regression analyses were used to compare outcomes between cancer patients and controls. ResultsOverall, there were 3382 cancer patients and 11,135 controls with 59,431 person-years of follow-up. In adjusted analyses, there were no statistically significant (p ≤ 0.05) differences in diabetes complication rates between cancer patients and their controls in any of the three cancer cohorts. Combined, cancer patients were less likely (adjusted hazard ratio [HR] 0.88; 95% CI = 0.79–0.98) to develop retinopathy. Cancer patients were more likely to die of any cause (including cancer), but prostate cancer patients were less likely to die of causes associated with diabetes (HR 0.61; 95% CI = 0.43–0.88). Conclusions and implicationsThere is no evidence that incident cancer had an adverse impact on the long-term outcomes of preexisting diabetes. Implications for Cancer SurvivorsThese findings are important for cancer survivors with preexisting diabetes because they suggest that substantial improvements in the relative survival of several of the most common types of cancer are not undermined by excess diabetes morbidity and mortality.

Full text

Outcomes of preexisting diabetes mellitus in breast, colorectal,
and prostate cancer
Robert I. Griffiths
1,2
&José M. Valderas
3
&Emily C. McFadden
1
&Clare R. Bankhead
1
&
Bernadette A. Lavery
4,5
&Nada F. Khan
6
&Richard J. Stevens
1
&Nancy L. Keating
7,8
Received: 8 February 2017 /Accepted: 11 July 2017 /Published online: 22 July 2017
#The Author(s) 2017. This article is an open access publication
Abstract
Purpose Preexisting diabetes is associated with increased
morbidity and mortality in cancer. We examined the impact
of incident cancer on the long-term outcomes of diabetes.
Methods Using the United Kingdom Clinical Practice
Research Datalink, we identified three cohorts of diabetes
patients subsequently diagnosed with breast, colorectal, or
prostate cancer, each matched to diabetic noncancer controls.
Patients were required to have survived at least 1 year after
cancer diagnosis (cases) or a matched index date (controls),
andwerefollowedupto10yearsforincidentmicrovascular
and macrovascular complications and mortality. Multivariate
competing risks regression analyses were used to compare
outcomes between cancer patients and controls.
Results Overall, there were 3382 cancer patients and 11,135
controls with 59,431 person-years of follow-up. In adjusted
analyses, there were no statistically significant (p≤0.05) dif-
ferences in diabetes complication rates between cancer patients
and their controls in any of the three cancer cohorts. Combined,
cancer patients were less likely (adjusted hazard ratio [HR]
0.88; 95% CI = 0.79–0.98) to develop retinopathy. Cancer pa-
tients were more likely to die of any cause (including cancer),
but prostate cancer patients were less likely to die of causes
associated with diabetes (HR 0.61; 95% CI = 0.43–0.88).
Conclusions and implications There is no evidence that inci-
dent cancer had an adverse impact on the long-term outcomes
of preexisting diabetes.
Implications for Cancer Survivors These findings are impor-
tant for cancer survivors with preexisting diabetes because
they suggest that substantial improvements in the relative sur-
vival of several of the most common types of cancer are not
undermined by excess diabetes morbidity and mortality.
Keywords Diabetes mellitus .Breast neoplasms .Colorectal
neoplasms .Prostatic neoplasms .Outcomeassessment (health
care) .Diabetes complications .Mortality .Epidemiology
Introduction
Evidence from epidemiologic studies, as summarized in a
number of recent reviews [1–10], indicates that preexisting
diabetes is associated with worse short- and long-term out-
comes of cancer, both overall [9,10] and in specific types of
cancer including breast [3,4,10], colorectal [2–7,10], and
prostate [8]. There are several reasons why the converse also
may be true—that incident cancer adversely impacts the out-
comes of preexisting diabetes. Potential mechanisms include
adverse effects of cancer treatments on glycemic control [11,
Richard J. Stevens and Nancy L. Keating are co-senior authors.
*Robert I. Griffiths
robert.griffiths@conted.ox.ac.uk
1
Nuffield Department of Primary Care Health Sciences, University of
Oxford, Radcliffe Primary Care Building, Radcliffe Observatory
Quarter, Woodstock Road, Oxford OX2 6GG, UK
2
Division of General Internal Medicine, Johns Hopkins University
School of Medicine, Baltimore 21205, USA
3
Health Services & Policy Research Group and Exeter Collaboration
for Primary Care (APEx), University of Exeter, Exeter EX4 4SB, UK
4
Oxford University Hospitals NHS Trust, Oxford OX3 9DU, UK
5
Thames Valley Strategic Clinical Network, NHS England,
Leeds, UK
6
Royal Liverpool University Hospital, L7 8XP, Liverpool, UK
7
Department of Health Care Policy, Harvard Medical School,
Boston 02115, USA
8
Division of General Internal Medicine and Primary Care, Brigham
and Women’s Hospital, Boston 02115, USA
J Cancer Surviv (2017) 11:604–613
DOI 10.1007/s11764-017-0631-2
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

12], impact of cancer on patient self-management of diabetes
[13], and changes in the quality of diabetes primary care ser-
vices during cancer treatment and follow-up [14–19].
However, there is considerably less information on how can-
cer impacts the long-term outcomes of preexisting diabetes
[12,20].
This gap in our understanding of how cancer impacts
preexisting diabetes is important from several perspectives.
For instance, as early detection and advances in cancer therapy
and supportive care have substantially improved the relative
survival of many of the most common types of cancer [21],
overall morbidity and mortality in cancer depend increasingly
on the quality and outcomes of primary care for other under-
lying conditions [22]. In response, cancer organizations, such
as Cancer Research UK and Macmillan Cancer Support in the
United Kingdom (UK), have expressed concern that
overlooking other medical conditions during cancer treatment
and follow-up could result in excess morbidity and mortality,
thereby undermining gains associated with early detection and
treatment of cancer [23,24].
The objective of this study was to examine the effect of
cancer on the long-term outcomes of preexisting diabetes.
Methods
Study design and data source
We conducted a historical cohort study using the UK Clinical
Practice Research Datalink (CPRD) linked to the Office for
National Statistics (ONS) mortality data. The CPRD contains
anonymized information from general practitioner (GP) prac-
tices on demographics, symptoms, tests, diagnoses, therapies,
health-related behaviors, and referrals to secondary care for
over 11.3 million patients from 674 practices in the UK [25].
There are 4.4 million active (alive, currently registered) pa-
tients in the database, which is approximately 6.9% of the UK
population. These patients are broadly representative of the
UK general population in terms of age, sex, and ethnicity
[25]. This database (including its predecessor, the General
Practice Research Datalink) has been used extensively for
health services and epidemiologic research in cancer and dia-
betes [26–30].
Patient selection
Using CPRD, we identified three cohorts of cancer pa-
tients with preexisting diabetes, each matched to diabetic
noncancer controls. Cancer patients (cases) were included
if they met all of the following criteria: (a) diagnosed with
breast, colorectal, or prostate cancer on or after January
1st, 2000; (b) diagnosed with type I or type II diabetes at
least 2 years before their date of cancer diagnosis (index
date); (c) had no other cancer diagnosis, except
nonmelanoma skin cancer, before their index date; (d)
were age ≥50 years at their index date; (e) had at least
2 years of eligible CPRD data before their index date; (f)
had an index date before the end of the eligible CPRD
data; and (g) survived and were otherwise eligible for
follow-up 1 year after cancer diagnosis. Patients were re-
quired to have been age ≥50 years at their index date
because diabetes in older adults is linked to higher mor-
tality [31] and because older adults with diabetes are at
substantial risk for both acute and chronic microvascular
and cardiovascular complications [32].Menwithbreast
cancer were excluded. Read codes from the Department
of Health, Data and Business (QOF) Rules, Cancer and
Diabetes Indicator Sets, version 25.0 [33], were used to
identify diabetes and cancer in CPRD.
Control selection: matching
Each cancer patient (case) was matched to up to four
noncancer patients (controls) with preexisting diabetes on
GP practice number, sex (colorectal only), and age (±1 year)
at cancer diagnosis. Matched controls were also required to
have met inclusion criteria d–g above. In addition to the three
cohorts of breast, colorectal, and prostate cancer patients plus
noncancer controls, a fourth cohort was constructed by com-
bining the three individual cohorts.
Patients were followed up to 10 years after their index date
for new microvascular and macrovascular complications (as
described below). They were followed from 1 year after their
index date up to 10 years for all-cause and diabetes mortality.
Variables
Diabetes complications
Complications of diabetes consisted of incident microvascular
and macrovascular conditions [34] first identified in patients’
electronic health records up to 10 years after their index date.
Microvascular conditions consisted of retinopathy, neuropa-
thy, nephropathy, chronic kidney disease (stage 4 or 5),
and the composite outcome of any of the four above.
Macrovascular complications consisted of peripheral arterial
disease, acute myocardial infarction or coronary syndrome,
cerebrovascular accident, lower limb amputation, and the
composite outcome of any of the four above. Incident compli-
cations were identified using published lists of Read codes
[35] present in patients’records up to 10 years after their index
date. Those patients identified with a specific complication
prior to their index date were excluded from the population
at risk for that complication during follow-up.
J Cancer Surviv (2017) 11:604–613 605
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Mortality
Variables were constructed for all-cause and diabetes mortal-
ity, which was defined as the presence of an International
Classification of Diseases, 10th Revision, Clinical
Modification (ICD-10-CM) code in the ONS data indicating
that the underlying cause of death was due to diabetes (ICD-
10-CM E10–14), hyperglycemia (R73), hypoglycemia
(E16.1, E16.2), myocardial infarction (I21–I22), ischemic
heart disease (I20, I24, I25), stroke/sequelae (I60–I64,
I69.0–I69.4), heart failure (I50), sudden death due to cardiac
arrest (I46), peripheral vascular disease (I70–I74), or kidney
disease (N00–N28) [36,37].
Covariates
Baseline characteristics consisted of age at index date, sex,
calendar year of index date, most recent (before the index
date) smoking status, most recent drinking status, and index
of multiple deprivation quintile, from least deprived (=1) to
most deprived (=5); body mass index (kg/m
2
); and Charlson
comorbidity index [38,39]. Baseline clinical and laboratory
values consisted of blood pressure (mm Hg), total cholesterol
(mmol/L), and glycosylated hemoglobin (HbA1c:
mmol/mol), identified with the use of the most recent value
within 1 year before the index date. Categorical variables for
laboratory values were constructed by using cutoffs that
corresponded to the thresholds for meeting the laboratory-
based performance indicators in the QOF Rules, Diabetes
Indicator Set, version 25.0 [33]: blood pressure ≤140/80 mm
Hg, total cholesterol ≤5 mmol/L (193 mg/dL), and HbA1c
≤59 (7.5%), 59–≤64 (8.0%), 64–≤75 (9.0%), and >75 mmol/
mol. Baseline antidiabetic agents were identified with the use
of the British National Formulary codes in the CPRD therapy
file [40] within 1 year before the index date.
Statistical methods
Since there was a reasonable chance overall mortality would
be higher in cancer patients than controls, we used competing
risks regression according to the approach proposed by Fine
and Gray [41] to estimate the cumulative incidence function
and unadjusted and adjusted hazard ratios (HR) for each mi-
crovascular and macrovascular complication, comparing can-
cer patients to noncancer controls. Patients were censored at
10 years after their index date or at the end of their eligibility
for follow-up in the data, whichever came first. The compet-
ing risk was death prior to the date of censoring. Adjusted
analyses included baseline demographic and clinical covari-
ates as described above. Patients with a specific diabetes com-
plication prior to their index date were excluded from that
particular analysis.
Competing risks regression [41] also was used to estimate
unadjusted and adjusted HRs for all-cause and diabetes mor-
tality. Only patients eligible for linkage to the ONS data were
included in the mortality analyses. Also, since study inclusion
criteria required patients to have survived at least 1 year after
their index date, survival analyses included only years 2–10 of
follow-up.
Results
Overall, there were 14,517 patients in the combined cohort:
3382 (23.3%) cancer patients and 11,135 (76.7%) controls
(Table 1). There were no statistically significant differences
in age, sex, and year of diagnosis between cancer patients and
controls, because age and sex were used as matching criteria
in constructing the cohorts, and controls received the same
index date as their corresponding cases. There were small
but statistically significant (p≤0.05) differences between can-
cer patients and controls in the distributions of smoking status,
BMI, and Charlson comorbidity index (Table 1). There were
no differences between cancer patients and controls in mean
baseline blood pressure or HbA1c. Cancer patients had statis-
tically significantly lower mean baseline total cholesterol than
controls (Table 1). However, the absolute difference was only
0.1 mmol/L (3.9 mg/dL).
The total number of years of follow-up in the combined
cohort was 59,431, 13,372 (22.5%) of which were for cancer
patients. Overall, the median length of follow-up was
1495 days (4.1 years), and follow-up was statistically signifi-
cantly shorter in cancer patients (median 1444 days) than in
controls (median 1511 days: log-rank test for equality of sur-
vivor functions, p< 0.0001).
Diabetes complications
There were no differences between cancer patients and con-
trols in the adjusted rate of any microvascular or
macrovascular complication (Fig. 1), either in the combined
cohort or in any of the three cancer cohorts (total of 80 unad-
justed and adjusted comparisons), except that in the combined
cohort and only in the colorectal cancer cohort, cancer patients
were less likely to develop retinopathy (adjusted HRs 0.88;
95% CI = 0.79–0.98 and 0.81; 95% CI = 0.66–0.91, respec-
tively). The results of sensitivity analyses in which a variable
specifying the type of diabetes (type I or type II) was added to
the list of predictors in the adjusted models were virtually
identical tothose reported above (results not shown) (Table 2).
Mortality
All-cause mortality was statistically significantly higher in
cancer patients than controls (Table 3). However, there
606 J Cancer Surviv (2017) 11:604–613
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Tabl e 1 Patient characteristics—
combined cohorts Combined cohorts pvalue
Cancer (n= 3382) Control (n= 11,135) All (n= 14,517)
n%n%n%
Age
50–<60 248 7.3 800 7.2 1048 7.2 0.740
60–<70 946 28.0 3098 27.8 4044 27.9
70–<80 1498 44.3 5041 45.3 6539 45.0
≥80 690 20.4 2196 19.7 2886 19.9
Sex
Male 1974 58.4 6551 58.8 8525 58.7 0.630
Female 1408 41.6 4584 41.2 5992 41.3
Year of diagnosis
2000–2004 546 16.1 1649 14.8 2195 15.1 0.087
2005–2009 1500 44.4 4902 44.0 6402 44.1
≥2010 1336 39.5 4584 41.2 5920 40.8
Smoking status
Nonsmoker 991 29.3 3364 30.2 4355 30 <0.001
Ex-smoker 1716 50.7 5633 50.6 7349 50.6
Current smoker 324 9.6 1240 11.1 1564 10.8
Not reported 351 10.4 898 8.1 1249 8.6
Body mass index
<25 579 17.1 1932 17.4 2511 17.3 <0.001
25–<30 1272 37.6 4183 37.6 5455 37.6
≥30 1344 39.7 4656 41.8 6000 41.3
Not reported 187 5.5 364 3.3 551 3.8
Charlson comorbidity index
1–2 1878 55.5 5898 53.0 7776 53.6 0.024
3–4 1044 30.9 3578 32.1 4622 31.8
>4 460 13.6 1659 14.9 2119 14.6
Type of cancer (or control)
Breast 1036 30.6 3194 28.7 4230 29.1 <0.001
Colorectal 1069 31.6 4047 36.3 5116 35.2
Prostate 1277 37.8 3894 35.0 5171 35.6
Type of diabetes
Type I 141 4.2 554 5.0 695 4.8 0.054
Type II 3241 95.8 10,581 95.0 13,822 95.2
Any microvascular complication
No 2412 71.3 7750 69.6 10,162 70.0 0.056
Yes 970 28.7 3385 30.4 4355 30.0
Any macrovascular complication
No 2681 79.3 8684 78.0 11,365 78.3 0.113
Yes 701 20.7 2451 22.0 3152 21.7
Any antidiabetic agent
No 769 22.7 2363 21.2 3132 21.6 0.060
Yes 2613 77.3 8772 78.8 11,385 78.4
Blood pressure ≤140/80 mm Hg
Yes 2026 59.9 6770 60.8 8796 60.6 <0.001
No 1250 37.0 4252 38.2 5502 37.9
Not reported 106 3.1 113 1.0 219 1.5
Systolic mean (SD) 3276 136 (15) 11,022 136 (16) 14,298 136 (16) 0.205
Diastolic mean (SD) 3276 74 (9) 11,022 74 (9) 14,298 74 (9) 0.090
Total cholesterol ≤5 mmol/L (193 mg/dL)
Yes 2706 80.0 8971 80.6 11,677 80.4 <0.001
No 475 14.0 1826 16.4 2301 15.9
Not reported 201 5.9 338 3.0 539 3.7
Mean (SD) in mmol/L 3181 4.2 (0.94) 10,797 4.3 (1.57) 13,978 4.2 (1.45) <0.001
Mean in mg/dL 162.4 166.3 162.4
HbA1c (mmol/mol)
≤59 (7.5%) 2179 64.4 6995 62.8 9174 63.2 0.362
59–≤64 (8.0%) 310 9.2 1054 9.5 1364 9.4
65–≤75 (9.0%) 322 9.5 1074 9.6 1396 9.6
>75 185 5.5 702 6.3 887 6.1
Not reported 386 11.4 1310 11.8 1696 11.7
Mean (SD) 2996 54.0 (12.5) 9825 54.5 (13.1) 12,821 54.3 (13.0) 0.060
Mean (%) 7.1 7.1 7.1
SD standard deviation, HbA1c glycosylated hemoglobin
J Cancer Surviv (2017) 11:604–613 607
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

0.1 .2 .3 .4 .5 .6 .7 .8 .9 1
0 1 2 3 4 5 6 7 8 9 10
Combined Cohorts
0.1 .2 .3 .4 .5 .6 .7 .8 .9 1
Cumulative Incidence
0 1 2 3 4 5 6 7 8 9 10
Breast Cancer
0.1 .2 .3 .4 .5 .6 .7 .8 .9 1
0 1 2 3 4 5 6 7 8 9 10
Year (Index date = 0)
Colorectal Cancer
0.1 .2 .3 .4 .5 .6 .7 .8 .9 1
Cumulative Incidence
0 1 2 3 4 5 6 7 8 9 10
Year (Index date = 0)
Prostate Cancer
Any Microvascular Complication*
0.1 .2 .3 .4 .5 .6 .7 .8 .9 1
0 1 2 3 4 5 6 7 8 9 10
Combined Cohorts
0.1 .2 .3 .4 .5 .6 .7 .8 .9 1
Cumulat ive Incide nce
0 1 2 3 4 5 6 7 8 9 10
Breast Cancer
0.1 .2 .3 .4 .5 .6 .7 .8 .9 1
0 1 2 3 4 5 6 7 8 9 10
Year (Index date = 0)
Colorectal Cancer
0.1 .2 .3 .4 .5 .6 .7 .8 .9 1
Cumulative Incidence
0 1 2 3 4 5 6 7 8 9 10
Year (Index date = 0)
Prostate Cancer
Any Macrovascular Complication*
608 J Cancer Surviv (2017) 11:604–613
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

was no evidence cancer adversely impacted diabetes-
related mortality. Diabetes-related mortality was signifi-
cantly lower among cancer patients in the combined co-
hort (adjusted HR 0.76; 95% CI = 0.61–0.94) and in the
prostate cancer cohort (adjusted HR 0.61; 95% CI = 0.43–
0.88) (Table 3). The results of sensitivity analyses in
which a variable specifying the type of diabetes (type I
or type II) was added to the list of predictors in the
models were virtually identical to those reported above
(results not shown).
Discussion
Evidence indicates that preexisting diabetes is associated
with worse outcomes in several types of cancer. However,
there is less information on whether incident cancer is
associated with worse diabetes outcomes. This as an im-
portant gap because as the relative survival for many
types of cancer continues to improve, overall survival de-
pends increasingly on the quality and outcomes of care
for other underlying conditions. Overall, this study sug-
geststhatintheUK,thereisnoreasontosuspectthatthe
outcomes of diabetes in the presence of cancer are ad-
versely affected by the competing needs of cancer care.
Among 80 unadjusted and adjusted comparisons of di-
abetes complications, we found no instance in which can-
cer was associated with higher incidence of a complica-
tion. Cancer was associated with lower adjusted incidence
of retinopathy in the colorectal cancer and the combined
cohorts, but the reason for this is not immediately clear,
and chance finding due to multiple testing cannot be ruled
out. As might be expected, cancer was associated with
higher all-cause mortality even though we excluded pa-
tients who survived less than 1 year. However, there was
no evidence of an adverse impact on diabetes-related mor-
tality. In fact, our findings suggest that prostate cancer
was associated with lower diabetes mortality in competing
risks regression that accounted for death due to other
causes.
There are several possible reasons that we found no
adverse associations between cancer and diabetes compli-
cations or mortality. First, we were able to adjust for dif-
ferences in baseline characteristics between diabetic
cancer patients and diabetic noncancer controls that oth-
erwise could have confounded associations between can-
cer and diabetes outcomes. These included baseline BMI,
smoking status, HbA1c, cholesterol, and blood pressure.
Second, our study was conducted in the UK, which has a
robust primary care system in which, under the quality
and outcomes framework [33], there are financial incen-
tives for GPs to provide high-quality diabetes care
throughout the continuum of care for other conditions
such as cancer.
Our study has several strengths, which are attributable
largely to the high quality of the database we used.
Notably, because the study was based on CPRD, we were
able to adjust for additional clinical factors typically not
available in US health insurance claims databases, which
otherwise may have confounded associations between can-
cer and the outcomes of diabetes. These included baseline
BMI, smoking status, blood pressure, cholesterol, and
HbA1c. In addition, we had rich data about clinical out-
comes over a relatively long period of time. Our study also
has several limitations. First, at the time it was conducted,
we were unable to link the CPRD data to information from
the National Cancer Intelligence Network (NCIN), which
would have given us details on cancer stage and initial
treatment. Although primary care data have a high sensi-
tivity and specificity for identifying cancer [42], registry
data would have allowed us to exclude cancer patients
diagnosed with metastatic disease. We considered using
Read codes in the primary care data files or ICD-10 codes
in the Hospital Episode Statistics (HES) inpatient data to
stage patients. However, we are not aware of any studies in
the UK that validate the use of ICD-10 codes for this pur-
pose, and because only two thirds of the patients in our
study were linked to HES, doing so would have limited
our sample sizes for all the analyses. Instead, we excluded
patients who died within the first year after their index
date. Second, since we did not have access to cancer treat-
ment data, we were unable to identify subgroups of cancer
patients who may have been at higher risk of diabetes
complications or related mortality due to the cancer treat-
ment they received. This should be the subject of further
research once linkage to NCIN and chemotherapy data-
bases becomes more readily available. Third, although
we required all patients to have survived at least 1 year
after their index date, requiring longer survival might have
enabled us to better assess the longer-term effects of cancer
treatment among the cancer patients. Finally, our findings
may not be generalizable to long-term survivors of breast,
colorectal, and prostate cancer in an era of full implemen-
tation of QOF, which was designed to improve the quality
of diabetes primary care, to other countries that do not have
primary care performance measures for diabetes care in
place, or to other types of cancers.
Fig. 1 Cumulative incidence of microvascular and macrovascular
complications. Asterisk, microvascular complications consisted of
retinopathy, neuropathy, nephropathy, or chronic kidney disease (stage
4 or 5). Macrovascular complications consisted of peripheral arterial
disease, acute myocardial infarction or coronary syndrome,
cerebrovascular accident, or lower limb amputation. Cancer patients
(red) and controls (blue)
J Cancer Surviv (2017) 11:604–613 609
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Tab l e 2 Multivariate analyses of complications
Type of complication Cohort
Combined Breast Colorectal Prostate
Unadjusted hazard
ratio [95% CI]
Adjusted hazard
ratio [95% CI]
Unadjusted hazard
ratio [95% CI]
Adjusted hazard
ratio [95% CI]
Unadjusted hazard
ratio [95% CI]
Adjusted hazard
ratio [95% CI]
Unadjusted hazard
ratio [95% CI]
Adjusted hazard
ratio [95% CI]
Any microvascular 0.96 [0.87–1.06] 0.99 [0.89–1.09] 1.03 [0.86–1.23] 1.04 [0.86–1.25] 1.01 [0.85–1.20] 1.02 [0.85–1.22] 0.89 [0.76–1.05] 0.94 [0.80–1.11]
Chronic kidney disease 1.08 [0.85–1.36] 1.13 [0.89–1.43] 1.23 [0.83–1.82] 1.22 [0.81–1.82] 0.94 [0.62–1.44] 1.05 [0.68–1.64] 1.01 [0.68–1.51] 1.11 [0.74–1.68]
Nephropathy 1.12 [0.66–1.88] 1.19 [0.70–2.02] 1.16 [0.41–3.28] 1.19 [0.41–3.45] 1.71 [0.78–3.73] 1.82 [0.83–3.98] 0.65 [0.25–1.70]
Neuropathy 1.07 [0.88–1.31] 1.10 [0.90–1.35] 1.22 [0.85–1.76] 1.23 [0.85–1.78] 1.31 [0.96–1.81] 1.30 [0.94–1.80] 0.83 [0.59–1.17] 0.84 [0.59–1.18]
Retinopathy 0.86** [0.77–0.96] 0.88* [0.79–0.98] 0.96 [0.79–1.17] 0.97 [0.79–1.18] 0.82 [0.67–1.00] 0.81* [0.66–0.99] 0.83* [0.69–0.99] 0.89 [0.74–1.06]
Any macrovascular 0.94 [0.80–1.11] 0.96 [0.82–1.13] 1.04 [0.77–1.40] 1.00 [0.74–1.37] 0.92 [0.69–1.21] 0.94 [0.71–1.24] 0.93 [0.72–1.21] 0.98 [0.75–1.26]
Acute myocardial infarction 1.09 [0.87–1.37] 1.11 [0.88–1.39] 0.82 [0.50–1.32] 0.79 [0.48–1.29] 1.42 [0.99–2.04] 1.37 [0.94–1.99] 1.03 [0.72–1.49] 1.12 [0.77–1.62]
Cerebrovascular accident 0.99 [0.78–1.25] 0.99 [0.78–1.26] 1.33 [0.87–2.03] 1.42 [0.92–2.17] 0.90 [0.59–1.37] 0.97 [0.64–1.47] 0.87 [0.59–1.28]
Lower limb amputation 0.76 [0.45–1.29] 0.83 [0.49–1.42] 0.51 [0.11–2.28] 0.45 [0.10–2.09] 1.18 [0.56–2.49] 1.21 [0.53–2.74] 0.55 [0.23–1.31] 0.62 [0.25–1.52]
Peripheral arterial disease 0.74* [0.56–0.97] 0.78 [0.59–1.03] 0.99 [0.56–1.74] 0.94 [0.52–1.72] 0.60* [0.36–1.00] 0.66 [0.39–1.10] 0.79 [0.52–1.18] 0.86 [0.57–1.30]
CI confidence interval
*p< 0.05; **p<0.01
610 J Cancer Surviv (2017) 11:604–613
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Tab l e 3 Multivariate survival analyses
Cohort
Combined Breast Colorectal Prostate
Unadjusted hazard
ratio [95% CI]
Adjusted hazard
ratio [95% CI]
Unadjusted hazard
ratio [95% CI]
Adjusted hazard
ratio [95% CI]
Unadjusted hazard
ratio [95% CI]
Adjusted hazard
ratio [95% CI]
Unadjusted hazard
ratio [95% CI]
Adjusted hazard
ratio [95% CI]
Mortality
All
cause
1.47*** [1.33–1.63] 1.57*** [1.41–1.74] 1.47** [1.21–1.80] 1.52*** [1.24–1.85] 1.64*** [1.38–1.95] 1.71*** [1.43–2.04] 1.44*** [1.22–1.72] 1.60*** [1.35–1.91]
Diabetes 0.73** [0.59–0.90] 0.76*** [0.61–0.94] 0.87 [0.58–1.31] 0.91 [0.59–1.40] 0.88 [0.63–1.23] 0.87 [0.62–1.23] 0.57** [0.40–0.82] 0.61*** [0.43–0.88]
**p< 0.01; ***p< 0.001
J Cancer Surviv (2017) 11:604–613 611
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Conclusions
Overall, incident cancer appears to have had little adverse
impact on the long-term outcomes of preexisting diabetes dur-
ing 10 years after a diagnosis of breast, colorectal, or prostate
cancer. These findings are important for cancer survivors with
preexisting diabetes because they suggest that the advances in
cancer therapy and supportive care, which have substantially
improved the relative survival of several of the most common
types of cancer, are not undermined by excess diabetes mor-
bidity and mortality.
Authors’contributions All authors contributed to the conception and
design of the work. RIG and ECM acquired the data. RIG and ECM
performed the analyses. All authors contributed to the interpretation of
the data for the work. RIG drafted the manuscript. All authors contributed
to revising it critically for important intellectual content. RIG takes full
responsibility for the work as a whole, including the study design, access
to data, and the decision to submit and publish the manuscript.
Compliance with ethical standards
Funding This study was funded by the Population Research
Committee, Cancer Research UK. Quality and Outcomes of Care for
Chronic Conditions in Older Patients Diagnosed with Breast,
Colorectal, or Prostate Cancer Compared to Non-Cancer Controls: An
Observational Study Using the Clinical Practice Research Datalink
(CPRD). Reference # 16609. 1 July 2013–29 February, 2016. In addition,
Dr. Keating is supported by K24CA181510 from the US National Cancer
Institute.
Conflict of interest The authors declare that they have no conflicts of
interest.
Ethical approval This article does not contain any studies with human
participants or animals performed by any of the authors. The protocol for
this study was approved by the Independent Scientific Advisory
Committee (ISAC) to the CPRD on 1 August 2013 (ISAC reference
number 13_124) with subsequent amendments approved by ISAC on
22 May 2014. Changes relevant to the analyses presented here include
identification of diabetes complications through the primary care record,
as described above, rather than from hospital data; this change was made
because hospital data linkage was not available in all patients. The ISAC
protocol was made available to the reviewers and editors during the peer-
review process.
Open Access This article is distributed under the terms of the Creative
Commons Attribution 4.0 International License (http://
creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided you give appro-
priate credit to the original author(s) and the source, provide a link to the
Creative Commons license, and indicate if changes were made.
References
1. Giovannuci E, Harlan DM, Archer MC, Bergenstal RM, Gapstur
SM, Habel LA, et al. Diabetes and cancer: a consensus report.
Diabetes Care. 2010;33(7):1674–85.
2. Mills KT, Bellows CF, Hoffman AE, Kelly TN, Gagliardi G.
Diabetes and colorectal cancer prognosis: a meta-analysis. Dis
Colon Rectum. 2013;56(11):1304–19.
3. De Bruijn KMJ, Arends LR, Hansen BE, Leeflang S, Ruiter R, van
Eijck CHJ. Systematic review and meta-analysis of the association
between diabetes mellitus and incidence and mortality in breast and
colorectal cancer. Br J Surg. 2013;100:1421–9.
4. Peairs KS, Barone BB, Snyder CF, Yeh H-C, Stein KB, Derr RL,
et al. Diabetes mellitus and breast cancer outcomes: a systematic
review and meta-analysis. J Clin Oncol. 2011;29:40–6.
5. Jiang Y, Ben Q, Shen H, Lu W, Zhang Y, Zhu J. Diabetes mellitus
and incidence and mortality of colorectal cancer: a systematic re-
view and meta-analysis of cohort studies. Eur J Epidemiol.
2011;26:863–76.
6. Luo W, Cao Y, Liao C, Gao F. Diabetes mellitus and the incidence
and mortality of colorectal cancer: a meta-analysis of 24 cohort
studies. Color Dis. 2011;14:1307–12.
7. Stein KB, Snyder CF, Barone BB, Yeh H-C, Peairs KS, Derr RL,
et al. Colorectal cancer outcomes, recurrence, and complications in
persons with and withoutdiabetes mellitus: a systematic review and
meta-analysis. Dig Dis Sci. 2010;55:1839–51.
8. Snyder CF, Stein KB, Barone BB, Peairs KS, Yeh H-C, Derr RL,
et al. Does pre-existing diabetes affect prostate cancer prognosis? A
systematic review. Prostate Cancer Prostatic Dis. 2010;13:58–64.
9. Barone BB, Yeh H-C, Snyder CF, Peairs KS, Stein KB, Derr RL,
et al. Postoperative mortality in cancer patients with preexisting
diabetes: systematic review and meta-analysis. Diabetes Care.
2010;33(4):931–9.
10. Barone BB, Yeh H-C, Snyder CF, Pearis KS, Stein KB, Derr RL,
et al. Long-term all-cause mortality in cancer patients with
preexisting diabetes mellitus: a systematic review and meta-analy-
sis. JAMA. 2008;300(23):2754–64.
11. Psarakis HM. Clinical challenges in caring for patients with diabe-
tes and cancer. Diabetes Spectrum. 2006;19(3):157–62.
12. Keating NL, Liu P-H, O’Malley AJ, Freedland SJ, Smith MR.
Androgen deprivation therapy and diabetes control among diabetic
men with prostate cancer. Eur Urol. 2014;65(4):816–24.
13. Hershey DS, Tipton J, Given B, Davis E. Perceived impact of
cancer treatment on diabetes self-management. The Diabetes
Educator. 2012;38(6):779–90.
14. Snyder CF, Frick KD, Herbert RJ, Blackford AL, Neville BA,
Wolff AC, et al. Quality of care for comorbid conditions during
the transition to survivorship: differences between cancer survivors
and noncancer controls. J Clin Oncol. 2013;31(9):1140–8.
15. Khan NF, Mant D, Rose PW. Qualityof care for chronic diseases in
a British cohort of long-term cancer survivors. Ann Fam Med.
2011;8(5):418–24.
16. Chiao EY, Nambi PV, Naik AD. The impact of diabetes process and
outcome quality measures on overall survival in patients with co-
morbid colorectal cancer. J Cancer Surviv. 2010;4(4):381–7.
17. Hanchate AD, Clough-Gorr KM, Ash AS, Thwin SS, Silliman RA.
Longitudinal patterns in survival, comorbidity, healthcare utiliza-
tion and quality of care among older women following breast can-
cer diagnosis. J Gen Intern Med. 2010;25(10):1045–50.
18. Keating NL, Zaslavsky AM, Herrinton LJ, Selby JV, Wolf RE,
Ayanian JZ. Quality of diabetes care among cancer survivors with
diabetes. Med Care. 2007;45(9):869–75.
19. Earle CC, Neville BA. Under use of necessary care among cancer
survivors. Cancer. 2004;101(8):1712–9.
20. Bayliss EA, Blatchford PJ, Newcomer SR, Steiner JF, Fairclough
DL. The effect of incident cancer, depression and pulmonary dis-
ease exacerbations on type 2 diabetes control. J Gen Intern Med.
2011;26(6):575–81.
21. Cancer Research UK. Cancer survival for common cancers. http://
www.cancerresearchuk.org/health-professional/cancer-statistics/
612 J Cancer Surviv (2017) 11:604–613
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

survival/common-cancers-compared#undefined. Accessed
Feb 2016.
22. Ning Yi, Shen Qin, Herrick K, Mikkelsen R, Anscher M, Houlihan
R, Lapane K. Cause of death in cancer survivors. Cancer Res
2012;72(8 Suppl):Abstract nr LB-339.
23. Cancer Research UK. Longer cancer survival means nearly half of
cancer patients die from other diseases. http://www.
cancerresearchuk.org/about-us/cancer-news/news-report/2012-04-
03-longer-cancer-survival-means-nearly-half-of-cancer-patients-
die-from-other-diseases?view=rss. Accessed Feb 2016.
24. Macmillan Cancer Support. Throwing light on the consequences of
cancer and its treatment. July 2013. http://www.ncsi.org.uk/wp-
content/uploads/MAC14312_CoT_Throwing-light_report_
FINAL.pdf. Accessed Feb 2016
25. Herrett E, Gallagher AM, Bhaskaran K, Forbes H, Mathur R, van
Staa T, et al. Data resource profile: Clinical Practice Research
Datalink (CPRD). Int J Epidemiol. 2015;44(3):827–36.
26. Khan NF, Mant D, Carpenter L, Forman D, Rose PW. Long-term
health outcomes in a British cohort ofbreast, colorectal and prostate
cancer survivors: a database study. Br J Cancer. 2011;105:S29–37.
27. Taylor KS, Heneghan CJ, Farmer AJ, Fuller AM, Adler AI,
Aronson JK, et al. All-cause and cardiovascular mortality in
middle-aged people with type 2 diabetes compared with people
without diabetes in a large U.K. primary care database. Diabetes
Care. 2013;36(8):2366–71.
28. Tzoulaki I, Molokhia M, Curcin V, Little MP, Millett CJ, Ng A,
et al. Risk of cardiovascular disease and all cause mortality among
patients with type 2 diabetes prescribed oral antidiabetes drugs:
retrospective cohort study using UK general practice research data-
base. BMJ. 2009;339:b4731.
29. Redaniel MTM, Jeffreys M, May MT, Ben-Shlomo Y, Martin RM.
Associations of type 2 diabetes and diabetes treatment with breast
cancer risk and mortality: a population-based cohort study among
British women. Cancer Causes Control. 2012;23:1785–95.
30. Salisbury C, Johnson L, Purdy S, Valderas JM, Montgomery AA.
Epidemiology and impact of multimorbidity in primary care: a ret-
rospective cohort study. Br J Gen Pract. 2011;61(582):e12–21.
31. Brown AF, Mangione CM, Saliba D, Sarkisian CA. Healthcare
Foundation/American Geriatrics Society panel on improving care
for elders with diabetes. Guidelines for improving the care of the
older person with diabetes mellitus. J Am Geriatr Soc.
2003;51(Suppl. Guidelines):S265–80.
32. Kirkman MS, Jones Briscoe V, Clark N, Florez H, Haas LB, Halter
JB, et al. Diabetes in older adults. Diabetes Care. 2012;35:2650–64.
33. Primary Care Commissioning (PCC). QOF business rules v25.
http://www.pcc-cic.org.uk/article/qof-business-rules-v25.0.
Accessed Feb 2016.
34. Adler AI, Stratton IM, Neil HAW, Yudkin JS, Matthews DR, Cull
CA, et al., on behalf of the UK Prospective Diabetes Study Group.
Association of systolic blood pressure with macrovascular and mi-
crovascular complications of type 2 diabetes (UKPDS 36): prospec-
tive observational study. BMJ. 2000;321:412–9.
35. Kontopantelis E, Springate DA, Reeves D, Ashcroft DM, Rutter M,
Buchan I, et al. Glucose, blood pressure and cholesterol levels and
their relationships to clinical outcomes in type 2 diabetes: a retro-
spective cohort study. Diabetologia. 2015;58(3):505–18.
36. Australian Institute of Health and Welfare. http://www.aihw.gov.au/
diabetes-indicators/deaths/. Accessed 19 Jan 2016.
37. Ricci-Cabello I, Stevens S, Kontopantelis E, Dalton ARH, Griffiths
RI, Campbell JL, et al. Impact of the prevalence of concordant and
discordant conditions on the quality of diabetes care in family prac-
tices in England. Ann Fam Med. 2015;13:514–22.
38. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method
of classifying prognostic comorbidity in longitudinal studies: devel-
opment and validation. J Chronic Dis. 1987;40(5):373–83.
39. Khan NF, Perera R, Harper S, Rose PW. Adaptation and validation
of the Charlson Index for Read/OXMIS coded databases. BMC
Fam Pract. 2010:11(1).
40. Dave S, Petersen I. Creating medical and drug code lists to identify
cases in primary care databases. Pharmacoepidemiol Drug Saf.
2009;18(8):704–7.
41. Fine JP, Gray RJ. A proportional hazards model for the
subdistribution of a competing risk. J Am Stat Assoc.
1999;94(446):496–509.
42. Boggon R, van Staa TP, Chapman M, Gallagher AM, Hammad TA,
Richards MA. Cancer recording and mortality in the General
Practice Research Database and linked cancer registries.
Pharmacoepidemiol Drug Saf. 2013;22:168–75.
J Cancer Surviv (2017) 11:604–613 613
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com