Table 2.
OR (95 CI) for the association between diabetes, diabetes management and breast cancer in MCC-Spain, overall and by tumor subtype
| Exposure variables | Controls | Overall | HR +/HER2− | HER2+ | TN | p-het | ||||
|---|---|---|---|---|---|---|---|---|---|---|
| Cases | OR (95 %CI) | Cases | OR (95 %CI) | Cases | OR (95 %CI) | Cases | OR (95 %CI) | |||
| Overall population | ||||||||||
| Diabetes | ||||||||||
| Non-diabetic | 995 | 836 | 1.00 | 530 | 1.00 | 121 | 1.00 | 74 | 1.00 | |
| Diabetic | 99 | 80 | 1.09 (0.82–1.45) | 51 | 0.94 (0.68–1.29) | 10 | 0.83 (0.38–1.85) | 17 | 2.25 (1.22–4.15) | <0.01 |
| Diabetes management | ||||||||||
| Non-diabetic | 995 | 724 | 1.00 | 529 | 1.00 | 121 | 1.00 | 74 | 1.00 | |
| Conservative management | 21 | 14 | 0.92 (0.45–1.88) | 6 | 0.49 (0.18–1.32) | 5 | 2.00 (0.99–4.71) | 1 | * | <0.01 |
| Treatment with oral hypoglycemic agents | 70 | 46 | 0.83 (0.55–1.25) | 32 | 0.85 (0.58–1.25) | 2 | * | 11 | 2.13 (1.08–4.23) | <0.01 |
| Treatment with insulin (± oral hypoglycemic agents) | 8 | 20 | 2.98 (1.26–7.01) | 13 | 2.94 (0.74–11.62) | 2 | * | 5 | 4.81 (1.31–17.63) | 0.17 |
| Diabetic population | ||||||||||
| Metformin use (years)** | 43 | 35 | 0.95 (0.86–1.04) | 24 | 0.89 (0.81–0.99) | 3 | 1.11 (0.95–1.28) | 8 | 1.02 (0.92–1.13) | 0.01 |
| Sulfonylurea use (years)** | 24 | 18 | 1.05 (0.99–1.13) | 13 | 1.03 (0.93–1.13) | 1 | * | 4 | 1.10 (1.00–1.20) | 0.40 |
| Insulin use (years)** | 6 | 20 | 1.10 (0.98–1.23) | 13 | 1.11 (0.93–1.33) | 2 | * | 5 | 1.07 (0.83–1.38) | 0.91 |
Numbers may differ due to lack of information on tumor receptors in some participants
Models for diabetes and diabetes management are adjusted for age, study level (no studies/<high school/high school/>high school), BMI (continuous), age at menarche, age at first birth, existence of previous biopsies and family history of the studied cancer (none/one first degree/more than one first degree/second degree)
Models for metformin and sulfonylurea further adjusted for insulin treatment (yes/no) and for treatment with other hypoglycemic agent (yes/no), while models for insulin are adjusted for treatment with metformin (yes/no) or sulfonylurea (yes/no)
Associations that reach statistical significance are in bold
* Values are not presented due to the small number of cases in these subgroups
** Based on diabetic participants who had received drug treatment for at least 1 year