Risk of Breast Cancer by Individual Insulin Use: An International Multicenter Study

Abstract

OBJECTIVESeveral studies have been published in 2009 suggesting a possible association between insulin glargine and increased risk of malignancies, including breast cancer. The objective of this study was to assess the relation between the individual insulins (glargine, aspart, lispro, and human insulin) and development of breast cancer. RESEARCH DESIGN AND METHODSSeven hundred seventy-five incident cases of primary invasive or in situ carcinoma breast cancer occurring in women with diabetes from 92 centers in the U.K., Canada, and France were matched to a mean of 3.9 diabetic community control subjects (n = 3,050; recruited from 580 general practices) by country, age, recruitment date, and diabetes type and management. The main risk model was a multivariate conditional logistic regression model with case/control status as the dependent variable and individual insulin use, 8 years preceding the index date, as the independent variable, controlling for past use of any insulin, oral antidiabetes drugs, reproductive factors, lifestyle, education, hormone replacement therapy and history of contraceptive use, BMI, comorbidities, diabetes duration, and annual number of physician visits. Glargine was also compared with every other insulin by computing all ratios using the variance-covariance matrix of logistic model parameters. RESULTSAdjusted odds ratios of breast cancer for each type of insulin versus no use of that insulin were 1.04 (95% CI 0.76-1.44) for glargine, 1.23 (0.79-1.92) for lispro, 0.95 (0.64-1.40) for aspart, and 0.81 (0.55-1.20) for human insulin. Two-by-two comparisons found no difference between glargine and the different types of insulins. Insulin dosage or duration of use and tumor stage did not change the results. CONCLUSIONSThis international study found no difference in the risk of developing breast cancer in patients with diabetes among the different types of insulin with short-to midterm duration of use. Longer-term studies would be of interest.

Effect of Long-Acting Insulin Analogs on the Risk of Cancer: A Systematic Review of Observational Studies

“…Time-window bias was observed in two studies (18,22). In one of the two studies, despite matching on calendar time, time-window bias was potentially present because case and control subjects were not matched on diabetes duration (as depicted in Supplementary Fig.…”

“…In one of the two studies, despite matching on calendar time, time-window bias was potentially present because case and control subjects were not matched on diabetes duration (as depicted in Supplementary Fig. 5) (18). Consequently, the opportunity for exposure differed between the case and control subjects due to the varying diabetes durations (a mean of 14.5 years among case subjects and 13.2 among control subjects).…”

“…The mean or median durations of follow-up and age ranged from 0.9 to 7.0 years and from 52.3 to 77.4 years, respectively. Thirteen studies examined ever use of long-acting insulin analogs, which was defined as at least one prescription, compared with nonuse, other, human, or NPH insulin (5)(6)(7)(8)14,16,18,19,21,23,(25)(26)(27). One study examined duration of time since starting long-acting insulin analogs, and one examined mean daily dose (22,28).…”

“…1). Following our inclusion criteria, 16 cohort and 3 case-control studies were included in this systematic review (5)(6)(7)(8)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28). All studies evaluated insulin glargine, with four studies also investigating insulin detemir (15,17,25,28).…”

“…These studies raised important concerns but were also criticized for important methodological shortcomings (9)(10)(11)(12)(13). Since then, several observational studies assessing the association between longacting insulin analogs and cancer have been published but yielded inconsistent findings (14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28). Such discrepancies may be due to methodological limitations, including inadequate durations of followup between insulin initiation and cancer incidence, protopathic bias, detection bias, the inclusion of prevalent users, and time-related biases such as immortal time bias, time-window bias, and time-lag bias (29).…”