Laurent Azoulay

Objective: Polyendocrine metabolic ovarian syndrome (PMOS) is a multisystemic disorder with a wide range of health consequences. We sought to describe temporal trends in the prevalence and incidence of PMOS and to characterize its clinical presentation, comorbidity burden, and pharmacological management in primary care in the United Kingdom (UK) between 1998 and 2025. Design: Population‐based matched cohort study using primary care records from the UK Clinical Practice Research Datalink. Patients: Females newly diagnosed with PMOS between the ages of 15 and 49 matched on age, general practice, calendar year of cohort entry, and duration of medical history to population controls without PMOS. Measurements: Annual prevalence and incidence of diagnosed PMOS; baseline prevalence and incidence of cardinal PMOS features and various comorbidities; and pre‐ and post‐cohort entry prevalence of hormonal contraceptive, anti‐androgen, ovulation‐inducing drug, and metformin use. Results: The prevalence of diagnosed PMOS increased from < 0.1% in 1998 to 2.0% in 2025, with incidence rising from 0.1 to 3.6 per 1,000 person‐years over the same period. Among 125,482 patients with newly diagnosed PMOS matched with an equal number of controls, patients with PMOS had higher baseline prevalence and incidence of cardinal features of PMOS, as well as diabetes, dyslipidemia, hypertension, cardiovascular and cerebrovascular disease, sleep apnea, thyroid disorders, endometriosis, uterine fibroids, chronic pain, depression, and anxiety. Use of anti‐androgens, ovulation‐inducing drugs, and metformin was substantially higher in patients with PMOS, especially in the year following diagnosis. Conclusions: PMOS is increasingly diagnosed in UK primary care and is associated with substantial multimorbidity and pharmacotherapy.

Rationale: Obstructive sleep apnea (OSA) is highly prevalent in type 2 diabetes (T2D), sharing metabolic risk factors, notably obesity, and increased cardiovascular outcomes. Glucagon-like peptide-1 receptor agonists (GLP-1RAs) have been shown to reduce OSA severity in clinical trials but their role in preventing incident OSA in T2D remains unclear. Objective: To assess whether GLP-1RA use is associated with a reduced risk of incident OSA compared with dipeptidyl peptidase-4 inhibitors (DPP-4i), a weight-neutral comparator, in patients with T2D. Methods: We conducted a population-based cohort study using the UK CPRD (2007–2023) database with an active-comparator, new-user design. Adults with T2D and BMI ≥30 kg/m² initiating a GLP-1RA or DPP-4i were included, excluding individuals with prior sleep apnea. Propensity scores with fine-stratification weighting were used within BMI strata. The primary outcome was incident clinical diagnosis of OSA. Cox proportional hazards models estimated hazard ratios (HRs) with 95% confidence intervals (CIs) using an as-treated approach, with up to 3 years of follow-up. Results: The cohort included 47,315 GLP-1RAs and 159,066 DPP-4i initiators with baseline characteristics well balanced after weighting. During follow-up, 612 and 1,197 incident OSA cases occurred among GLP-1RAs and DPP-4i users, yielding incidence rates of 5.8 and 5.4 per 1000 person-years, respectively. The HR of incident OSA with GLP-1RA use compared with DPP-4i use was 1.07 (95% CI 0.93–1.23). Results were consistent across BMI strata, sex, drug type, and multiple sensitivity analyses. Conclusions: In routine clinical care, GLP-1RAs were not associated with a decreased incidence of OSA in patients with T2D and no history of OSA, when compared with DPP-4i.

OBJECTIVE: To estimate the effect of initiating glucagon-like peptide 1 receptor agonists (GLP-1 RAs) versus dipeptidyl peptidase 4 (DPP-4) inhibitors on incident nonarteritic anterior ischemic optic neuropathy (NAION) among adults with type 2 diabetes. RESEARCH DESIGN AND METHODS: This active-comparator, new-user cohort emulated a pragmatic target trial using the U.K. Clinical Practice Research Datalink. Patients aged ≥18 years with a physician diagnosis of type 2 diabetes who newly initiated a GLP-1 RA or a DPP-4 inhibitor were included. DPP-4 inhibitors were selected as the comparator because they may be used as second-line treatment, like GLP-1 RAs, and have no established association with NAION. We estimated risks, risk differences (RDs), and risk ratios (RRs) of incident NAION, adjusted using propensity-score fine-stratification weighting. RESULTS: At 1 year, there were 14 NAION events among 106,858 GLP-1 RA initiators (18.5 per 100,000) and 53 among 416,369 DPP-4 inhibitor initiators (7.2 per 100,000). GLP-1 RAs were associated with an increased risk of NAION compared with DPP-4 inhibitors (RR 2.56; 95% CI 1.44–4.86; RD 11.3 per 100,000). The risk of NAION was higher during the first 6 months of use, diminishing with longer duration of use, and was higher in patients aged <50 years, men, ever-smokers, and those with ≥1% hemoglobin A1c reduction. CONCLUSIONS: GLP-1 RAs were associated with an increased 1-year risk of NAION compared with DPP-4 inhibitors among adults with type 2 diabetes, particularly in younger patients, men, ever-smokers, and patients with a marked hemoglobin A1c reduction.

Purpose The objective of this study was to determine whether pre-existing type 2 diabetes is associated with an increased risk of breast cancer-related and all-cause mortality, compared with non-diabetes, and if the risk varies across glycated hemoglobin A1c (HbA1c) categories. Methods Using the Clinical Practice Research Datalink, we assembled a cohort of patients at least 18 years old, newly diagnosed with invasive breast cancer between 1998 and 2020, with follow-up until March 2021. Patients were followed from 3 months after breast cancer diagnosis until one of the study outcomes, end of registration with the general practice, or end of study. Multivariable Cox proportional hazards models, adjusted for 33 confounders, were fitted to estimate the hazard ratios (HRs) and 95% confidence intervals (CIs) of breast cancer-related and all-cause mortality, comparing patients with and without type 2 diabetes among patients with and without metastatic breast cancer. In a secondary analysis, HRs for the outcomes were estimated across glycated HbA1c categories. Results The cohort included 157,298 patients with incident breast cancer. Among patients with non-metastatic breast cancer (n = 125,268), type 2 diabetes was associated with a 12% increased risk of breast cancer mortality and 21% increased risk of all-cause mortality. Among patients with metastatic breast cancer (n = 32,030), type 2 diabetes was associated with a 22% increased risk of breast cancer mortality and a 24% increased risk of all-cause mortality. The secondary analysis showed that higher glycated HbA1c was associated with an increased risk, in both the non-metastatic and metastatic cohorts. Conclusion In this large population-based cohort study, type 2 diabetes was associated with a greater risk of breast cancer-related and all-cause mortality, with an increased risk among patients with type 2 diabetes in the highest glycated HbA1c categories, compared to non-diabetes.

Baseline Characteristics with standardised differences for propensity Score Analysis using standard mortality ratio weights for bladder cancer cohort

Crude and adjusted hazard ratios for association between 5-alpha reductase inhibitor use and kidney and bladder cancer compared to non-use

Baseline Characteristics with standardised differences for propensity Score Analysis using standard mortality ratio weights for kidney cancer cohort

Background:Preclinical evidence suggests that 5α-reductase inhibitors (5ARi), commonly used to treat benign prostatic hyperplasia (BPH), are associated with reduced incidence of certain urologic cancers, yet epidemiologic studies are conflicting. This study aimed to determine whether 5ARi's are associated with a reduced risk of kidney and bladder cancers.Methods:We conducted a new-user active-comparator cohort study in the United Kingdom Clinical Practice Research Datalink. From a base cohort of patients with incident BPH, new users of 5ARi's and α-blockers were identified. Patients were followed up until a first ever diagnosis of kidney or bladder cancer, death from any cause, end of registration, or December 31, 2017. Cox proportional hazards models were used to calculate HRs and 95% confidence intervals (CI) for incident kidney and bladder cancer.Results:There were 5,414 and 37,681 new users of 5ARi's and α-blockers, respectively. During a mean follow-up of 6.3 years, we found no association between the use of 5ARi's and kidney (adjusted HR, 1.26; 95% CI, 0.74–2.12; n = 23) or bladder (adjusted HR, 0.89; 95% CI, 0.64–1.23; n = 57) cancer risk compared with α-blockers. Similar results were observed across sensitivity analyses.Conclusions:In this study, we found no association between the use of 5ARi's and kidney or bladder cancer incidence in men with BPH when compared with α-blocker use.Impact:The findings of this study indicate that 5ARi's are unlikely to reduce kidney or bladder cancer risk.