This work is licensed under a Abstract
Objective
Type 2 diabetes increases the risk of developing breast cancer, while endocrine therapy for early breast cancer may elevate the risk of type 2 diabetes. Studies suggest screening patients with early breast cancer receiving endocrine therapy for type 2 diabetes, but the optimal strategies remain unclear. We assessed haemoglobin A1c (HbA1c) screening adherence, the prevalence of HbA1c levels above the normal range (≥ 42 mmol/mol), and an HbA1c follow-up procedure.
Methods
Patients with early (stage I–III) breast cancer without diabetes receiving endocrine therapy were offered a single HbA1c screening test after treatment initiation at Aarhus University Hospital, Denmark, from October 2022 to April 2023. HbA1c ≥ 42 mmol/mol was classified as prediabetic (42–47 mmol/mol) or diabetic (≥ 48 mmol/mol). Patients and their general practitioners received standardised digital notifications if HbA1c ≥ 42 mmol/mol, recommending follow-up. Statistical analyses included Fisher’s exact test and t-tests.
Results
Of 174 patients offered HbA1c screening, 97 (56%) underwent testing. HbA1c ≥ 42 mmol/mol was detected in 11 patients (11% (95% confidence interval (CI): 5.8–19%)), including two patients with diabetic HbA1c levels (2.1% (95% CI: 0.3–7.3%)). Overweight (body mass index ≥ 25 kg/m2) and age ≥ 55 years correlated with HbA1c ≥ 42 mmol/mol. Most patients with HbA1c ≥ 42 mmol/mol (9 of 11) underwent primary-care follow-up.
Conclusion
Adherence to HbA1c screening may be affected if the test is an extra step for patients. Targeted HbA1c screening may be beneficial for patients with early breast cancer with overweight or older age receiving endocrine therapy. Further research is needed to improve implementation and participation.
Keywords: breast, oncology, endocrine therapy, HbA1c, screening
Introduction
Type 2 diabetes (T2D) is an important risk factor for numerous diseases, including breast cancer (Martin & McGee 2018). Breast cancer is both the most commonly diagnosed cancer (excluding non-melanoma skin cancer) and the leading cause of cancer-related mortality among women (Sung et al. 2021). The relationship between T2D and breast cancer is bidirectional: not only does T2D increase the risk of developing breast cancer, but systemic adjuvant treatment for breast cancer appears to increase the risk of developing T2D (Scott et al. 2023).
Anti-oestrogen treatments, also known as endocrine therapy, may partly explain the increased risk of developing T2D among patients with breast cancer. Multiple systematic reviews have found that T2D risk is elevated among patients with early breast cancer (EBC) (stages I–III) who are receiving adjuvant endocrine therapy compared with both healthy individuals and patients with EBC who do not receive endocrine treatment (Lipscombe et al. 2013, Ye et al. 2022, Jordt et al. 2023, Kjærgaard et al. 2024). The elevated risk of developing T2D has been observed for up to 10 years after the breast cancer diagnosis (Lipscombe et al. 2012, 2013).
Early diagnosis of T2D is crucial, as up to 50% of individuals with screening-detected T2D already have complications such as coronary artery disease, retinopathy, nephropathy, or neuropathy at the time of diagnosis (Simmons et al. 2010, Duan et al. 2021). Furthermore, patients with breast cancer and pre-existing T2D experience poorer cancer outcomes than those without T2D (Kim et al. 2023). The glycated haemoglobin A1c (HbA1c) test is one of the most widely used and well-standardised tests for prediabetes and diabetes screening. Unlike fasting blood glucose testing, HbA1c testing does not require patient preparation and is unaffected by short-term dietary changes (Higgins 2013). Given the high risk of progression from prediabetes to T2D and the benefits of early intervention, screening with HbA1c is recommended in populations with an elevated risk of developing T2D (Lee et al. 2019, Duan et al. 2021).
Numerous studies on the higher prevalence of T2D among patients with EBC receiving adjuvant endocrine therapy have recommended T2D screening for this population (Lipscombe et al. 2013, Jordt et al. 2023, Scott et al. 2023). However, uncertainties remain regarding who should be screened, as well as when and how it should be done. Current general guidelines for T2D screening prioritise high-risk populations but fail to recognise patients with EBC who are receiving adjuvant endocrine therapy, despite their sustained risk for T2D for up to a decade post-diagnosis of breast cancer (Ali et al. 2023).
To address these gaps, this study evaluated HbA1c screening for patients with EBC who were receiving adjuvant endocrine therapy and did not have known diabetes. Specifically, we investigated the adherence to HbA1c screening, the prevalence of HbA1c levels above the normal range (≥ 42 mmol/mol), and the implementation of an HbA1c follow-up procedure. By focussing on HbA1c, which can be used to detect diabetes and prediabetes, this study aimed to provide insights into the feasibility of structured and targeted HbA1c screening for patients with EBC receiving adjuvant endocrine therapy.
Methods
Study population
This observational, cross-sectional, single-centre study investigated adherence to HbA1c screening among patients with EBC who were receiving adjuvant endocrine therapy, the prevalence of HbA1c levels above the normal range (≥ 42 mmol/mol), and an HbA1c follow-up procedure aimed at general practitioners. HbA1c screening was offered one time within the first year after initiation of adjuvant endocrine treatment to patients with EBC and no diagnosis of diabetes, regardless of former HbA1c measurements. All patients were seen for a follow-up consultation at the Department of Oncology, Aarhus University Hospital (AUH), Denmark, from October 1, 2022, to April 30, 2023.
The HbA1c screening was based on a blood sample scheduled after the consultation. The inclusion criteria were: i) female sex and ii) having received adjuvant endocrine treatment (aromatase inhibitor or tamoxifen) for EBC for up to 1 year. The exclusion criteria included known pre-existing diabetes. Patients were included regardless of other systemic therapies received for breast cancer (e.g. chemotherapy, HER2-targeted therapy, or prednisolone), as the only treatment-related inclusion criterion was ongoing adjuvant endocrine therapy. Information was gathered through a review of electronic medical records, including pathological records, and assembled in a study-specific REDCap database hosted by Aarhus University. All time variables were calculated from their origin to the time point of the HbA1c screening test.
Patient characteristics
The patient characteristics examined included age, menopausal status, body mass index (BMI), smoking status, whether they lived with a partner, and the Charlson Comorbidity Index. Menopausal status was obtained from oncology notes. BMI was calculated as the weight (kg) divided by the height (m)2 based on the measurements recorded closest to the test date of HbA1c screening. BMI was categorised according to the definitions of the World Health Organization (WHO): underweight (BMI < 18.5 kg/m2), normal weight (18.5 ≤ BMI < 25 kg/m2), overweight (25 ≤ BMI < 30 kg/m2), and obesity (BMI ≥ 30 kg/m2) (Lim et al. 2024).
Smoking status and whether patients lived with a partner (living status) were classified according to information recorded closest to the HbA1c test date. Comorbidity was assessed using the Charlson Comorbidity Index (Charlson et al. 2022), which was obtained from the breast cancer surgery notes and grouped as 0, 1–2, 3–4, or ‘unknown’ if no score was recorded. Follow-up consultation refers to the specific follow-up consultation during which the patient was offered HbA1c screening, which was either the 3-month or 12-month follow-up after initiation of adjuvant endocrine therapy.
Breast cancer characteristics
For all breast cancer characteristics, only information regarding the currently treated breast cancer was used. The characteristics examined included the time since breast cancer diagnosis, history of breast cancer, unilateral/bilateral status, tumour size, number of positive lymph nodes, histological grade and type, and the tumour’s human epidermal growth factor receptor 2 (HER2) status. All tumours were oestrogen receptor positive. The date of the breast cancer diagnosis was defined as the date of the first pathology record identifying invasive breast cancer. Tumour size, nodal status, histological grade, and HER2-receptor status were classified according to the data recorded by the pathologist in the pathology records.
For breast cancers with multiple foci, the largest focus was used for analyses. Tumour size was classified as T1 (≤ 20 mm in the greatest dimension), T2 (> 20–≤ 50 mm in the greatest dimension), T3 (> 50 mm in the greatest dimension), T4 (extension to the chest wall or skin), or Tx (primary tumour cannot be assessed). Nodal status was classified as N0 (no positive lymph nodes), N1 (micrometastases or 1–3 positive lymph nodes), N2 (4–9 positive lymph nodes), or N3 (≥ 10 positive lymph nodes).
Breast cancer treatment modalities
The treatment modalities for breast cancer included the type and duration of the following breast cancer therapies: endocrine therapy, chemotherapy, and HER2–targeted therapy. In addition, supportive medication such as prednisolone was recorded. The first day of all treatment types was recorded as the start date of the prescription. If the patient had sequentially received an aromatase inhibitor and tamoxifen before the HbA1c screening test date, the accumulated time for endocrine therapy up to the day of the HbA1c screening test was calculated. Prednisolone use was recorded if the patient received any prednisolone during breast cancer treatment.
HbA1c analyses and categorisation
HbA1c was measured based on a venous blood sample, which the Department of Clinical Biochemistry of AUH analysed on the same day the blood test was taken. HbA1c levels were measured using capillary electrophoresis with a Capillarys 3 Tera TLA (Sebia, France) (Herpol et al. 2016). The analysis method has been accredited by the DS/EN ISO/IEC 15189 international laboratory standard.
According to the Danish diabetes guidelines and the International Expert Committee (IEC), HbA1c levels of < 42 mmol/mol were considered within the normal range. HbA1c levels above the normal range (≥ 42 mmol/mol) were classified as either pre-diabetic (42–47 mmol/mol) or diabetic (≥ 48 mmol/mol) (International Expert Committee 2009). In Denmark, a T2D diagnosis requires two HbA1c measurements of ≥ 48 mmol/mol if no relevant symptoms are present (Gillet 2009). For additional analyses, HbA1c was also classified according to the American Diabetes Association (ADA) guidelines, which define HbA1c levels above the normal range (≥ 39 mmol/mol) as pre-diabetic (39–47 mmol/mol) and diabetic (≥ 48 mmol/mol) (ElSayed et al. 2023). The results of additional analyses according to the ADA guidelines are provided as Supplementary material (see section on Supplementary materials given at the end of the article).
HbA1c follow-up procedure
The classification used in the follow-up procedure was based on the IEC criteria. An appointed research team at the Department of Oncology, AUH managed the follow-up of HbA1c measurements. If HbA1c levels were above the normal range (≥ 42 mmol/mol), the patient received a standardised letter in a secure digital mailbox, and their general practitioner was informed through standardised digital healthcare communication with a recommendation of further clinical assessment and intervention as indicated. If the HbA1c level was very high (> 86 mmol/mol) (Raz & Mosenzon 2013), immediate contact was made with the patient and their general practitioner, and an emergency discharge summary was issued, which warranted follow-up in primary care as soon as possible. Patients were not notified if their HbA1c levels were within the normal range (< 42 mmol/mol).
Statistical analysis and ethical considerations
Only patients with complete data for all analysed variables were included in the statistical analyses. The statistical software Stata version 18.5 (USA) was used for statistical analysis, and a statistical significance level of 5% was used for hypothesis testing. Fisher’s exact test was applied to compare categorical data for patient characteristics, breast cancer characteristics, and treatment modalities between subgroups, and a two-sample comparison t-test was used for continuous normally distributed data. Fisher’s exact test was applied due to categories of patient characteristics containing a small number of patients. Data are presented with 95% confidence intervals (CI).
This study did not require approval from an ethics committee or patient consent, as it does not fall under the definition of health science research according to the Danish Committee Act §2, no. 1. The Scientific Ethics Committees for the Central Denmark Region reviewed the project and confirmed that it did not require ethical notification or approval under §14, section 1 of the same Act. Access to data was approved by the Hospital Management at AUH.
Results
Study population
Between October 2022 and April 2023, 230 patients with EBC had follow-up consultations at the Department of Oncology of AUH. All patients were offered HbA1c screening and scheduled for a post-consultation blood sample. Of the 230 patients, 56 were excluded from the study cohort due to missing BMI measurements, choosing not to receive the offered endocrine therapy, or having known pre-existing diabetes (Fig. 1). Among the 174 eligible patients offered HbA1c screening, 97 took the screening test. Patients who took the HbA1c screening test had more often been treated with chemotherapy and received supportive prednisolone treatment than those who did not take the HbA1c test (Table 1). Patients were equally likely to undergo HbA1c screening after the 3- and 12-month follow-up consultations. Furthermore, those who took the test tended to be younger, more likely to live with partners, and less likely to be active smokers than those who did not take the test, although these differences did not reach significance (P-values between 0.06 and 0.14).
Figure 1.
Flow diagram of patient inclusion. Abbreviations: HbA1c: haemoglobin A1c; BMI: body mass index; N: number of patients.
Table 1.
Patient characteristics and treatment modalities of patients with EBC receiving adjuvant endocrine therapy who were scheduled for an HbA1c screening test.
| Variables | Did not take the HbA1c screening test | Took the HbA1c screening test | P-value |
|---|---|---|---|
| (n = 77) | (n = 97) | ||
| Patient characteristics | |||
| Age | *0.06 | ||
| In years (SD) | 66.0 (2.9) | 62.6 (2.2) | |
| Age, n (%) | †0.85 | ||
| Age <55 years | 16 (21%) | 18 (19%) | |
| Age ≥55 years | 61 (79%) | 79 (81%) | |
| Menopausal status, n (%) | †1.00 | ||
| Pre-menopausal | 14 (18%) | 18 (19%) | |
| Post-menopausal | 63 (82%) | 79 (81%) | |
| BMI | *0.81 | ||
| In kg/m2 (SD) | 26.7 (1.1) | 26.5 (1.0) | |
| BMI group, n (%) | †0.77 | ||
| Underweight: <18.5 kg/m2 | 1 (1.3%) | 1 (1.0%) | |
| Normal weight: 18.5–<25 kg/m2 | 31 (40%) | 38 (39%) | |
| Overweight: 25–<30 kg/m2 | 25 (33%) | 38 (39%) | |
| Obesity: ≥30 kg/m2 | 20 (26%) | 20 (21%) | |
| BMI group, n (%) | †0.88 | ||
| BMI <25 kg/m2 | 32 (42%) | 39 (40%) | |
| BMI ≥25 kg/m2 | 45 (58%) | 58 (60%) | |
| Smoking, n (%) | †0.14 | ||
| Never smoker | 50 (65.0%) | 66 (68%) | |
| Former smoker | 12 (16%) | 21 (22%) | |
| Active smoker | 12 (16%) | 10 (10%) | |
| Unknown | 3 (3.9%) | 0 (0%) | |
| Living status, n (%) | †0.14 | ||
| Living alone | 21 (27%) | 17 (17%) | |
| Living together‡ | 56 (73%) | 80 (83%) | |
| Charlson comorbidity index, n (%) | †0.67 | ||
| 0 | 37 (48%) | 47 (49%) | |
| 1–2 | 21 (27%) | 20 (21%) | |
| 3–4 | 1 (1.3%) | 3 (3.1%) | |
| Unknown | 18 (23%) | 27 (28%) | |
| Follow-up consultation§, n (%) | †0.64 | ||
| 3 months | 45 (58%) | 61 (63%) | |
| 12 months | 32 (42%) | 36 (37%) | |
| Treatment modalities | |||
| Endocrine therapy duration | *0.24 | ||
| Duration in days (SD) | 238.1 (52.2) | 204.5 (28.6) | |
| Endocrine therapy type, n (%) | †0.88 | ||
| Tamoxifen | 13 (17%) | 19 (20%) | |
| Aromatase inhibitors | 56 (73%) | 69 (71%) | |
| Both sequentially | 8 (10%) | 9 (9%) | |
| Chemotherapy, n (%) | †0.011 | ||
| No | 57 (74%) | 53 (55%) | |
| Yes | 20 (26%) | 44 (45%) | |
| Chemotherapy duration | *0.69 | ||
| Duration in days (SD) | 112.4 (14.8) | 109.0 (9.2) | |
| HER2 treatment, n (%) | †0.21 | ||
| No | 72 (94%) | 85 (88%) | |
| Yes | 5 (6.0%) | 12 (12%) | |
| HER2 therapy duration | *0.92 | ||
| Duration in days (SD) | 273.8 (106.2) | 278.8 (55.6) | |
| Prednisolone use║, n (%) | †0.027 | ||
| No | 57 (74%) | 56 (58%) | |
| Yes | 20 (26%) | 41 (42%) | |
| Last prednisolone dosage, n (%) | †0.28 | ||
| Within 6 months | 14 (70%) | 22 (54%) | |
| More than 6 months | 6 (30%) | 19 (46%) |
Data are shown as: i) number (%), where % is calculated by the number in the cell divided by the total number in the group; or ii) mean (SD), where SD is the standard deviation of the mean value.
Two-sample comparison t-test.
Fisher’s exact test.
In a relationship or married.
The treatment follow-up timespan, after which the patient was booked for an HbA1c screening test.
Prednisolone use during cancer treatment.
Abbreviations: EBC, early breast cancer; HbA1c, haemoglobin A1c; HER2, human epidermal growth factor receptor 2; N, number of patients; BMI, body mass index.
HbA1c levels
Of the 97 patients who took the HbA1c screening test, 86 had HbA1c levels within the normal range, and 11 had HbA1c levels above the normal range (≥ 42 mmol/mol). This corresponds to a prevalence of HbA1c levels above the normal range of 11% (95% CI: 5.8–19%). Among these 11 patients, nine had HbA1c levels categorised as pre-diabetic, and two had HbA1c levels categorised as diabetic (Fig. 2). This corresponds to a 2.1% (95% CI: 0.3–7.3%) prevalence of diabetic HbA1c levels.
Figure 2.
HbA1c screening test results. Abbreviations: HbA1c: haemoglobin A1c; N: number of patients.
Target group analysis
The analyses revealed significant associations of HbA1c levels above the normal range with age, BMI, and comorbidity score (Table 2). HbA1c levels above the normal range were associated with older age (69 years (95% CI: 63–75 years) vs 62 years (95% CI: 59–64 years)), higher BMI (31.0 kg/m2 (95% CI: 26.8–35.2 kg/m2) vs 25.9 kg/m2 (95% CI: 24.9–26.9 kg/m2)), and increased comorbidity (Charlson Comorbidity Index 0 at 27 vs 51%, and Charlson Comorbidity Index 3–4 at 18 vs 1.16%). The analyses of age and BMI categories showed a higher prevalence of HbA1c levels above the normal range among patients with EBC aged 55 years or older and those with a BMI of 25 kg/m2 or higher.
Table 2.
Patient characteristics of patients with EBC receiving adjuvant endocrine therapy who took the HbA1c screening test.
| Variables | HbA1c‡ < 42 mmol/mol | HbA1c ≥ 42 mmol/mol | P-value |
|---|---|---|---|
| (n = 86) | (n = 11) | ||
| Age | *0.04 | ||
| In years (SD) | 61.8 (2.4) | 68.9 (5.6) | |
| Age, n (%) | †0.09 | ||
| Age <55 years | 18 (21%) | 0 (0%) | |
| Age ≥55 years | 68 (79%) | 11 (100%) | |
| Menopausal status, n (%) | †0.35 | ||
| Pre-menopausal | 17 (20%) | 1 (9.09%) | |
| Post-menopausal | 69 (80%) | 10 (91%) | |
| BMI | *0.001 | ||
| In kg/m2 (SD) | 25.9 (1.0) | 31.0 (4.2) | |
| BMI group, n (%) | †0.05 | ||
| Underweight <18.5 kg/m2 | 1 (1.16%) | 0 (0%) | |
| Average weight 18.5–<25 kg/m2 | 37 (43%) | 1 (9.09%) | |
| Overweight 25–<30 kg/m2 | 33 (38%) | 5 (46%) | |
| Obesity ≥30 kg/m2 | 15 (17%) | 5 (46%) | |
| BMI group, n (%) | †0.05 | ||
| BMI <25 kg/m2 | 38 (44%) | 1 (9.09%) | |
| BMI ≥25 kg/m2 | 48 (56%) | 10 (91%) | |
| Smoking, n (%) | †1.00 | ||
| Never smoker | 58 (67%) | 8 (73%) | |
| Previous smoker | 19 (22%) | 2 (18%) | |
| Active smoker | 9 (11%) | 1 (9.09%) | |
| Living status, n (%) | †0.40 | ||
| Living alone | 14 (16%) | 3 (27%) | |
| Living together§ | 72 (84%) | 8 (73%) | |
| Comorbidity, n (%) | †0.04 | ||
| 0 | 44 (51%) | 3 (27%) | |
| 1–2 | 17 (20%) | 3 (27%) | |
| 3–4 | 1 (1.16%) | 2 (18%) | |
| Unknown | 24 (28%) | 3 (27%) | |
| Follow-up consultation║, n (%) | †0.74 | ||
| 3 months | 53 (62%) | 8 (73%) | |
| 12 months | 33 (38%) | 3 (27%) |
Data are shown as: i) number (%), where % is calculated by the number in the cell divided by the total number in the group; or ii) mean (SD), where SD is the standard deviation of the mean value.
Two-sample comparison t-test.
Fisher’s exact test.
Classified according to International Expert Committee Guidelines.
In a relationship or married.
The treatment follow-up timespan, after which the patient was booked for an HbA1c screening test.
Abbreviations: EBC, early breast cancer; HbA1c, haemoglobin A1c; N, number of patients; BMI, body mass index.
Breast cancer characteristics and HbA1c levels
Breast cancer characteristics showed no significant associations with HbA1c levels (Table 3). Patients with HbA1c levels above the normal range showed a non-significant tendency towards a higher N-stage, as 36% had N0 and 9.09% had N3, compared to patients with HbA1c levels within the normal range, where 57% had N0 and only 1.16% had N3.
Table 3.
Breast cancer characteristics of patients with EBC receiving adjuvant endocrine therapy who took the HbA1c screening test.
| Variables | HbA1c‡ < 42 mmol/mol | HbA1c ≥ 42 mmol/mol | P-value |
|---|---|---|---|
| (n = 86) | (n = 11) | ||
| Time since diagnosis§ | *0.72 | ||
| Duration in days (SD) | 326.2 (35.7) | 307.4 (111.2) | |
| Previous breast cancer, n (%) | †1.00 | ||
| No | 83 (97%) | 11 (100%) | |
| Yes | 3 (3.49%) | 0 (0%) | |
| Unilateral/bilateral, n (%) | †0.58 | ||
| Unilateral | 80 (93%) | 10 (91%) | |
| Bilateral | 6 (6.98%) | 1 (9.09%) | |
| T-stadium, n (%) | †0.14 | ||
| T1 | 52 (61%) | 4 (36%) | |
| T2 | 30 (35%) | 6 (55%) | |
| T3 | 3 (3.49%) | 0 (0%) | |
| T4 | 1 (1.16%) | 0 (0%) | |
| Tx | 0 (0%) | 1 (9.09%) | |
| N-stadium, n (%) | †0.19 | ||
| N0 | 49 (57%) | 4 (36%) | |
| N1 | 30 (35%) | 5 (46%) | |
| N2 | 6 (6.98%) | 1 (9.09%) | |
| N3 | 1 (1.16%) | 1 (9.09%) | |
| Histological type, n (%) | †1.00 | ||
| No special type | 69 (80%) | 9 (82%) | |
| Lobular | 13 (15%) | 2 (18%) | |
| Other║ | 4 (4.7%) | 0 (0%) | |
| Histological grade, n (%) | †0.95 | ||
| Unclassified | 8 (9.30%) | 1 (9.09%) | |
| Grade I | 17 (20%) | 2 (18%) | |
| Grade II | 50 (58%) | 6 (55%) | |
| Grade III | 11 (13%) | 2 (18%) | |
| HER2 status, n (%) | †0.70 | ||
| Normal | 73 (85%) | 9 (82%) | |
| Overexpression | 12 (14%) | 2 (18%) | |
| Not applicable | 1 (1.16%) | 0 (0%) |
Data are shown as: i) number (%), where % is calculated by the number in the cell divided by the total number in the group; or ii) mean (SD), where SD is the standard deviation of the mean value.
Two-sample comparison t-test.
Fisher’s exact test.
Classified according to International Expert Committee Guidelines.
Time from breast cancer diagnosis to the HbA1c screening test.
Including ductal/lobular combined.
Abbreviations: EBC, early breast cancer; HbA1c, haemoglobin A1c; HER2, human epidermal growth factor receptor 2; N, number of patients.
Breast cancer treatment modalities and HbA1c levels
Some breast cancer treatment modalities showed non-significant tendencies towards associations with HbA1c levels above the normal range (Table 4). Treatment with aromatase inhibitors was more frequent among patients with HbA1c levels above the normal range (91 vs 69%). A shorter duration of HER2-targeted treatment was also observed in patients with HbA1c levels above the normal range.
Table 4.
Breast cancer treatment modalities of patients with EBC receiving adjuvant endocrine therapy who took the HbA1c screening test.
| Variables | HbA1c‡ < 42 mmol/mol | HbA1c ≥ 42 mmol/mol | P-value |
|---|---|---|---|
| (n = 86) | (n = 11) | ||
| Endocrine therapy duration | *0.99 | ||
| Duration in days (SD) | 204.5 (30.0) | 204.1 (109.8) | |
| Endocrine therapy type, n (%) | †0.43 | ||
| Tamoxifen | 18 (21%) | 1 (9.09%) | |
| Aromatase inhibitors | 59 (69%) | 10 (91%) | |
| Both sequentially | 9 (10%) | 0 (0%) | |
| Chemotherapy, n (%) | †1.00 | ||
| No | 47 (55%) | 6 (55%) | |
| Yes | 39 (45%) | 5 (45%) | |
| Chemotherapy duration | *0.89 | ||
| Duration in days (SD) | 109.2 (9.2) | 107.2 (57.7) | |
| HER2 treatment, n (%) | †0.62 | ||
| No | 76 (88%) | 9 (82%) | |
| Yes | 10 (12%) | 2 (18%) | |
| HER2 therapy duration | *0.05 | ||
| Duration in days (SD) | 300 (56.5) | 172.5 (235.1) | |
| Prednisolone use§, n (%) | †0.76 | ||
| No | 49 (57%) | 7 (64%) | |
| Yes | 37 (43%) | 4 (36%) | |
| Last prednisolone dosage, n (%) | †0.32 | ||
| Within 3 months | 0 (0%) | 0 (0%) | |
| Within 6 months | 21 (57%) | 1 (25%) | |
| More than 6 months | 16 (43%) | 3 (75%) |
Data are shown as: i) number (%), where % is calculated by the number in the cell divided by the total number in the group; or ii) mean (SD), where SD is the standard deviation of the mean value.
Two-sample comparison t-test.
Fisher’s exact test.
Classified according to International Expert Committee Guidelines.
Prednisolone use during cancer treatment.
Abbreviations: EBC, early breast cancer; HbA1c, haemoglobin A1c; HER2, human epidermal growth factor receptor 2; N, number of patients.
HbA1c screening and follow-up procedure
Of the 174 patients who were offered HbA1c screening, 97 (56%) took the test, indicating moderate adherence. The HbA1c follow-up procedure was aimed at general practitioners to ensure patient safety and was considered successful. All 11 patients with HbA1c levels above the normal range and their general practitioners received standardised electronic letters recommending further clinical assessment. No patients had very high HbA1c levels.
Most patients (9 of 11) with HbA1c levels above the normal range had a follow-up HbA1c test with their general practitioner. The two patients with HbA1c levels within the diagnostic range for T2D had their condition confirmed, leading to a diagnosis and initiation of T2D treatment. Medical records from general practitioners are inaccessible at the regional hospital level in Denmark, so this study could not determine whether patients initiated lifestyle changes or whether other non-medical interventions were implemented.
HbA1c levels according to ADA guidelines
According to ADA criteria (normal HbA1c < 39 mmol/mol; see Supplementary materials for full results), 27 patients (27% (95% CI: 19–38%)) had HbA1c ≥ 39 mmol/mol, including 16 with levels between 39 and 41 mmol/mol (Fig. 2). Elevated HbA1c by ADA definition was associated with older age, higher BMI, and shorter duration of endocrine therapy, and was more common among patients who had received chemotherapy. These patients also underwent HbA1c screening closer to their breast cancer diagnosis and showed a non-significant tendency towards larger tumour size.
Discussion
The results indicated moderate patient adherence to HbA1c screening, as about half of the patients took the test. A notable proportion of patients with EBC receiving adjuvant endocrine therapy (11 out of 97) had HbA1c levels above the normal range, and key risk factors were older age or having overweight. These findings suggest that targeted HbA1c screening may benefit these subgroups the most. The HbA1c follow-up procedure was considered effective, as most patients (9 out of 11) with HbA1c levels above the normal range had follow-up in primary care.
The moderate adherence to HbA1c screening indicates that tests beyond standard care may be a participation barrier. In the adjuvant breast cancer setting, routine blood tests are only performed when clinically indicated, which makes HbA1c screening an additional step for patients. Physician adherence may also influence participation, as physicians may forget to remind patients about HbA1c screening during busy clinical routines.
Our results align with existing evidence that older age and overweight are significant risk factors for T2D (Wu et al. 2014). However, patients who took the HbA1c screening test were younger, more likely to live with a partner, and less likely to be active smokers than those who did not take it. Therefore, patients with potentially greater risk of metabolic dysregulation due to advanced age or smoking may be underrepresented in this study. The associations between HbA1c levels and older age, higher BMI, and more comorbidity may also be affected by this selection effect. The prevalence of HbA1c levels above the normal range (11%) is therefore potentially an underestimation of the true burden in the fully eligible population.
Another important consideration is the discrepancy between ADA and IEC HbA1c classification, as 16 patients had HbA1c levels between 39 and 41 mmol/mol, which are considered prediabetic by ADA standards but normal under IEC criteria. This difference nearly tripled the prevalence of HbA1c levels above the normal range when applying ADA thresholds (27 vs 11%). While these borderline values are considered normal by the IEC, research suggests that even modest HbA1c elevations are associated with poorer cancer prognosis (Holm et al. 2025). These findings highlight the relevance of recognising patients with mild HbA1c elevations, as early intervention could mitigate both metabolic deterioration and potentially adverse cancer outcomes.
While our study found an association between treatment with aromatase inhibitors and HbA1c levels above the normal range, these findings should be interpreted cautiously. In Denmark, aromatase inhibitors are only prescribed to postmenopausal patients, while tamoxifen is used for premenopausal patients. Since ageing itself is a well-established risk factor for T2D, the association may reflect confounding by age rather than a direct drug effect, considering the short exposure to endocrine therapy in our cohort. Nonetheless, the literature remains inconclusive on the contributions of different endocrine therapies to glycaemic alterations. Some studies found increased T2D risk with both tamoxifen and aromatase inhibitors, while others reported a stronger association with tamoxifen (Ng et al. 2018, Gupta et al. 2022, Kim et al. 2022, Kwan et al. 2022, Rillamas-Sun et al. 2023).
These therapies appear to disrupt glucose metabolism through different pathways. By antagonising oestrogen receptors in mammary tissue, tamoxifen may induce pancreatic β-cell apoptosis and impair insulin secretion (Le May et al. 2006). In contrast, aromatase inhibitors reduce systemic oestrogen levels, which may diminish the protective effects of oestrogen against oxidative stress in β-cells, thus impairing insulin production (Carpenter & Miller 2005). While our results cannot establish causality, they support the plausibility that endocrine therapies may contribute to dysglycaemia via different mechanisms.
The optimal timing for T2D screening during adjuvant breast cancer treatment is an important consideration, as endocrine therapy is administered for 5–10 years, and the T2D risk has been observed for up to 10 years after the breast cancer diagnosis. In our study, the short exposure to endocrine therapy means that the T2D risk is potentially more driven by age and BMI. This is an important limitation of our study, as HbA1c reflects the average blood glucose levels in the past 3 months, and the HbA1c screening test was performed after an average of 204 days of endocrine therapy. Future research should explore the timeframe of T2D screening during endocrine therapy. A one-time screening early in the treatment period, such as this study, may miss patients who develop metabolic dysfunction later, and repeated screenings may increase healthcare costs and patient burden. Targeted T2D screening based on risk factors such as age and BMI could be an efficient approach.
Effective follow-up is crucial for patients with high HbA1c levels, as up to 50% of individuals have vascular complications upon diagnosis with T2D (Simmons et al. 2010, Duan et al. 2021). In our study, a structured follow-up procedure ensured that most patients with high HbA1c levels received appropriate primary care follow-up, and two cases of T2D were confirmed and treated. While this study highlights the potential need for long-term health monitoring among patients with EBC receiving adjuvant endocrine therapy and a follow-up procedure with primary care, the organisation of healthcare is different across countries, which thus defines whether long-term metabolic monitoring is the responsibility of oncology clinics or primary care.
The clinical significance of our findings is the potential for early detection and intervention to reduce complications associated with hyperglycaemia (Duan et al. 2021). Emerging evidence suggests that glycaemic control may influence breast cancer outcomes (Phillips et al. 2023, Holm et al. 2025). Incorporating HbA1c screening into cancer follow-up could help identify under-recognised side effects of endocrine therapy and take proactive steps. To ensure cost-effectiveness, screening could target high-risk groups, aligning with recommendations for screening high-risk populations for T2D (Ekoe et al. 2018, Duan et al. 2021). Our findings suggest that patients with EBC aged ≥ 55 years or with a BMI ≥ 25 kg/m2 could be a target group for HbA1c screening, as they have the highest risk for HbA1c levels above the normal range.
While our study provides important insights, it has limitations. The relatively small sample size – and particularly the very limited number of patients with HbA1c levels above the normal range – restricts the robustness and generalisability of our findings. Although we observed trends suggesting associations between elevated HbA1c and higher N-stage, use of aromatase inhibitors, and shorter HER2-targeted treatment, these results are based on few cases and lack statistical significance. Therefore, no reliable conclusions can be drawn, and the observed tendencies should be interpreted with caution. Future studies with larger cohorts are needed to evaluate cost-effectiveness and the optimal integration of HbA1c screening strategies into cancer care. Furthermore, the benefits of early detection and intervention on both metabolic and cancer-specific outcomes should be assessed to guide healthcare policies.
Conclusion
Adherence to HbA1c screening may be compromised if the screening test is an extra step for patients. Patients with EBC who have overweight, are older, and are receiving endocrine therapy may primarily benefit from targeted HbA1c screening. The optimal timing for screening remains uncertain and additional studies are necessary to determine whether screening should be conducted once, repeatedly, or according to risk factors. Our findings underscore the need for further research into structured HbA1c screening to optimise implementation, enhance participation, and evaluate its impact on patient outcomes.
Supplementary materials
Declaration of interest
The authors declare that no relevant conflict of interest could be perceived as prejudicing the impartiality of this work.
Funding
Department of Oncology Research Fund (Grant reference nr: 09-9638) and, to LN, the Danish Cancer Society (Grant nr: R382-A22956).
Author contribution statement
SBS, MC, OD, TL, and SB designed the study. SBS set up the HbA1c follow-up protocol with general practitioners under the guidance of SB, MC, OD, and TL. SB had the overall responsibility for the study and supervised SBS in running the study. LN collected and analysed the data and prepared the manuscript under the supervision of JH and SB. All authors read and approved the final manuscript.
Acknowledgements
We would like to thank the Department of Oncology Research Fund and the Danish Cancer Society for financial support, and the patients, doctors, and nurses at the Department of Oncology, Aarhus University Hospital, for study participation.
References
- Ali MK, Imperatore G, Benoit SR, et al. 2023. Impact of changes in diabetes screening guidelines on testing eligibility and potential yield among adults without diagnosed diabetes in the United States. Diabetes Res Clin Pract 197 110572. ( 10.1016/j.diabres.2023.110572) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carpenter R & Miller WR. 2005. Role of aromatase inhibitors in breast cancer. Br J Cancer 93(Supp. 1) S1–S5. ( 10.1038/sj.bjc.6602688) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Charlson ME, Carrozzino D, Guidi J, et al. 2022. Charlson comorbidity index: a critical review of clinimetric properties. Psychother Psychosom 91 8–35. ( 10.1159/000521288) [DOI] [PubMed] [Google Scholar]
- Duan D, Kengne AP & Echouffo-Tcheugui JB. 2021. Screening for diabetes and prediabetes. Endocrinol Metab Clin North Am 50 369–385. ( 10.1016/j.ecl.2021.05.002) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ekoe JM, Goldenberg R & Katz P. 2018. Screening for diabetes in adults. Can J Diabetes 42(Supp. 1) S16–S19. ( 10.1016/j.jcjd.2017.10.004) [DOI] [PubMed] [Google Scholar]
- Elsayed NA, Aleppo G, Aroda VR, et al. 2023. 2. Classification and diagnosis of diabetes: standards of care in diabetes-2023. Diabetes Care 46 S19–S40. ( 10.2337/dc23-s002) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gillet MJ 2009. International Expert Committee report on the role of the A1c assay in the diagnosis of diabetes. Clin Biochem Rev 30 197–200. [PMC free article] [PubMed] [Google Scholar]
- Gupta T, Purington N, Liu M, et al. 2022. Incident comorbidities after tamoxifen or aromatase inhibitor therapy in a racially and ethnically diverse cohort of women with breast cancer. Breast Cancer Res Treat 196 175–183. ( 10.1007/s10549-022-06716-y) [DOI] [PubMed] [Google Scholar]
- Herpol M, Lanckmans K, Van Neyghem S, et al. 2016. Evaluation of the Sebia Capillarys 3 tera and the bio-rad D-100 systems for the measurement of hemoglobin A1c. Am J Clin Pathol 146 67–77. ( 10.1093/ajcp/aqw081) [DOI] [PubMed] [Google Scholar]
- Higgins T 2013. HbA1c for screening and diagnosis of diabetes mellitus. Endocrine 43 266–273. ( 10.1007/s12020-012-9768-y) [DOI] [PubMed] [Google Scholar]
- Holm JB, Bruun JM, Christiansen P, et al. 2025. HbA(1c) levels and breast cancer prognosis in women without diabetes. BMC Cancer 25 790. ( 10.1186/s12885-025-14121-z) [DOI] [PMC free article] [PubMed] [Google Scholar]
- International Expert Committee 2009. International Expert Committee report on the role of the A1C assay in the diagnosis of diabetes. Diabetes Care 32 1327–1334. ( 10.2337/dc09-9033) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jordt N, Kjærgaard KA, Thomsen RW, et al. 2023. Breast cancer and incidence of type 2 diabetes mellitus: a systematic review and meta-analysis. Breast Cancer Res Treat 202 11–22. ( 10.1007/s10549-023-07043-6) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kim JE, Choi J, Park J, et al. 2022. Effects of endocrine therapy on cardiovascular diseases and type 2 diabetes among breast cancer survivors: the national health insurance service database of Korea. J Am Heart Assoc 11 e026743. ( 10.1161/jaha.122.026743) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kim YJ, Goak I, Jin HY, et al. 2023. 1422-P: The impact of type 2 diabetes mellitus on the clinical outcome of breast cancer. Diabetes 72(Supp. 1) 1422-P. ( 10.2337/db23-1422-p) [DOI] [Google Scholar]
- Kjærgaard KA, Kousholt A, Thomsen RW, et al. 2024. Risk of type-2-diabetes after breast cancer treatment: a population-based cohort study in Denmark. J Natl Cancer Inst 117 537–544. ( 10.1093/jnci/djae261) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kwan ML, Cheng RK, Iribarren C, et al. 2022. Risk of cardiometabolic risk factors in women with and without a history of breast cancer: the pathways heart study. J Clin Oncol 40 1635–1646. ( 10.1200/jco.21.01738) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Le May C, Chu K, Hu M, et al. 2006. Estrogens protect pancreatic beta-cells from apoptosis and prevent insulin-deficient diabetes mellitus in mice. Proc Natl Acad Sci U S A 103 9232–9237. ( 10.1073/pnas.0602956103) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee K, Tripathy D, Demark-Wahnefried W, et al. 2019. Effect of aerobic and resistance exercise intervention on cardiovascular disease risk in women with early-stage breast cancer: a randomized clinical trial. JAMA Oncol 5 710–714. ( 10.1001/jamaoncol.2019.0038) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lipscombe LL, Fischer HD, Yun L, et al. 2012. Association between tamoxifen treatment and diabetes: a population-based study. Cancer 118 2615–2622. ( 10.1002/cncr.26559) [DOI] [PubMed] [Google Scholar]
- Lim Y, Haq N & Boster J. 2024. Obesity and comorbid conditions. In StatPearls. Treasure Island, FL, USA: StatPearls Publishing. (https://www.ncbi.nlm.nih.gov/books/NBK574535/) [PubMed] [Google Scholar]
- Lipscombe LL, Chan WW, Yun L, et al. 2013. Incidence of diabetes among postmenopausal breast cancer survivors. Diabetologia 56 476–483. ( 10.1007/s00125-012-2793-9) [DOI] [PubMed] [Google Scholar]
- Martin SD & McGee SL. 2018. Metabolic reprogramming in type 2 diabetes and the development of breast cancer. J Endocrinol 237 R35–R46. ( 10.1530/joe-18-0037) [DOI] [PubMed] [Google Scholar]
- Ng HS, Koczwara B, Roder D, et al. 2018. Incidence of comorbidities in women with breast cancer treated with tamoxifen or an aromatase inhibitor: an Australian population-based cohort study. J Comorb 8 16–24. ( 10.15256/joc.2018.8.125) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Phillips AL, Reeves DJ & Storey S. 2023. Impact of diabetes (type 2) and glycemic control on health-related outcomes of patients receiving chemotherapy for non-metastatic breast cancer: a retrospective analysis. Support Care Cancer 31 114. ( 10.1007/s00520-022-07563-9) [DOI] [PubMed] [Google Scholar]
- Raz I & Mosenzon O. 2013. Early insulinization to prevent diabetes progression. Diabetes Care 36(Supp. 2) S190–S197. ( 10.2337/dcs13-2014) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rillamas-Sun E, Kwan ML, Iribarren C, et al. 2023. Development of cardiometabolic risk factors following endocrine therapy in women with breast cancer. Breast Cancer Res Treat 201 117–126. ( 10.1007/s10549-023-06997-x) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Scott L, Truong LL, Houlden RL, et al. 2023. Screening and management recommendations for type 2 diabetes in women with breast cancer. Can J Diabetes 48 66–72. ( 10.1016/j.jcjd.2023.07.008) [DOI] [PubMed] [Google Scholar]
- Simmons RK, Echouffo-Tcheugui JB & Griffin SJ. 2010. Screening for type 2 diabetes: an update of the evidence. Diabetes Obes Metab 12 838–844. ( 10.1111/j.1463-1326.2010.01244.x) [DOI] [PubMed] [Google Scholar]
- Sung H, Ferlay J, Siegel RL, et al. 2021. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 71 209–249. ( 10.3322/caac.21660) [DOI] [PubMed] [Google Scholar]
- Wu Y, Ding Y, Tanaka Y, et al. 2014. Risk factors contributing to type 2 diabetes and recent advances in the treatment and prevention. Int J Med Sci 11 1185–1200. ( 10.7150/ijms.10001) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ye F, Wen J, Yang A, et al. 2022. The influence of hormone therapy on secondary diabetes mellitus in breast cancer: a meta-analysis. Clin Breast Cancer 22 e48–e58. ( 10.1016/j.clbc.2021.06.014) [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.