Disparities in adherence to diabetes screening guidelines among males and females in a universal care setting: A population-based study of 1,380,697 adults

Value is Gendered: The Need for Sex and Gender Considerations in Health Economic Evaluations

Economic evaluations play a crucial role in health resource allocation by assessing the costs and effects of various interventions. However, existing methodologies often overlook significant differences related to sex and gender, leading to a 'blind spot' in understanding patient heterogeneity. This paper highlights how biological and social factors influence costs and health outcomes differently for women, emphasising the need for a more explicit consideration of these differences in economic evaluations to ensure efficient and equitable resource allocation. The paper is structured to first outline how sex and gender factors impact costs and outcomes. It then identifies biases in current economic evaluation methods and practices, using real-world examples to illustrate the implications of these biases on policymaking and health equity. Notably, we argue that neglecting gender considerations can lead to inefficiencies and inequities in healthcare resource distribution. Key areas of gender bias include the estimation of productivity losses, quality of life variations and the secondary household effects of interventions. The analysis reveals that women often face higher healthcare costs and experience different health outcomes due to systemic biases in treatment and care. The paper concludes with practical recommendations for analysts, decision makers and research funders, advocating for the integration of sex and gender-responsive methodologies in health economic evaluations. Ultimately, this work calls for a paradigm shift in health economics to better reflect the complexities of sex and gender and improve health outcomes for all.

Utility of photoplethysmography in detecting elevated blood glucose among non-diabetics

INTRODUCTION

This study aimed to evaluate a technique of using photoplethysmography (PPG) for detecting elevated blood glucose in individuals.

METHOD

This is a prospective, cross-sectional study in which 500 healthy volunteers were recruited at a tertiary hospital in Singapore from October 2021 to February 2023. Capillary glucose was measured concurrently with PPG signals acquired using the wrist-worn Actxa Tracker (Spark + Series 2) and the In-Ear Prototype model SVT, which were worn for a duration of 8 min. Participants with a capillary blood test reading ≤11.1 mmol/dL had to consume a standard glucose tolerance drink and return 1 h later for a second capillary blood test. Two hundred and forty-four features were subsequently extracted from the PPG signals.

RESULTS

Of the 500 volunteers, 17 were excluded because of incomplete records. This led to a total of 483 participants' records being included in the final analysis. For predicting elevated capillary blood glucose level, demographics alone achieved an area under the curve (AUC) of 0.75. When wearable features derived from PPG were combined with demographics, AUC improved significantly to 0.82 (P = 0.0001).

CONCLUSION

This study shows that a non-invasive method of assessing diabetes mellitus risk using PPG combined with demographics is a viable option to provide a cheaper and more accessible modality for population-wide diabetes mellitus risk assessment.

Association Between Hemoglobin A1c and Development of Cardiovascular Disease in Canadian Men and Women Without Diabetes at Baseline: A Population-Based Study of 608 474 Adults

BACKGROUND

We examined the association between hemoglobin A1c (HbA1c) and the development of cardiovascular disease (CVD) in men and women, without diabetes or CVD at baseline.

METHODS AND RESULTS

This retrospective cohort study included adults aged 40 to <80 years in Alberta, Canada. Men and women were divided into categories based on a random HbA1c during a 3-year enrollment period. The primary outcome of CVD hospitalization and secondary outcome of combined CVD hospitalization/mortality were examined during a 5-year follow-up period until March 31, 2021. A total of 608 474 individuals (55.2% women) were included. Compared with HbA1c 5.0% to 5.4%, men with HbA1c of 5.5% to 5.9% had an increased risk of CVD hospitalization (adjusted hazard ratio [aHR], 1.12 [95% CI, 1.07-1.19]) whereas women did not (aHR, 1.01 [95% CI, 0.95-1.08]). Men and women with HbA1c of 6.0% to 6.4% had a 38% and 17% higher risk and men and women with HbA1c ≥6.5% had a 79% and 51% higher risk of CVD hospitalization, respectively. In addition, HbA1c of 6.0% to 6.4% and HbA1c ≥6.5% were associated with a higher risk (14% and 41%, respectively) of CVD hospitalization/death in men, but HbA1c ≥6.5% was associated with a 24% higher risk only among women.

CONCLUSIONS

In both men and women, HbA1c ≥6.0% was associated with an increased risk of CVD and mortality outcomes. The association between CVD and HbA1c levels of 5.5% to 5.9%, considered to be in the "normal" range, highlights the importance of optimizing cardiovascular risk profiles at all levels of glycemia, especially in men.

Increased Prevalence of Adverse Health Outcomes Across the Lifespan in Those Affected by Polycystic Ovary Syndrome: A Canadian Population Cohort

Background

Polycystic ovary syndrome (PCOS) is the most common metabolic-endocrine disorder impacting the health and quality of life of women over the lifespan. Evidence-based data on the scope of adverse health outcomes in those affected by PCOS is critical to improve healthcare and quality of life in this population. The aim of this study was to determine the prevalence of adverse health outcomes in those with PCOS compared to age-matched controls.

Methods

We conducted a retrospective observational case-control study in those diagnosed with PCOS and age-matched controls using the Alberta Health Services Health Analytics database and the International Classification of Diseases, for the period from 2002-2018 in Alberta, Canada.

Results

The cohort consisted of n = 16,531 exposed PCOS cases and n = 49,335 age-matched un-exposed controls. The prevalences of hypertension, renal disease, gastrointestinal disease, eating disorders, mental illness, depression-anxiety, rheumatoid arthritis, respiratory infections, and all malignancies were 20%-40% (P < 0.0001) higher in those with PCOS, compared to controls. The prevalence of obesity, dyslipidemia, nonalcoholic fatty liver disease, and type 2 diabetes was 2-3 fold higher in those with PCOS (P < 0.001). Cardiovascular, cerebrovascular, and peripheral vascular disease were 30%-50% higher, and they occurred 3-4 years earlier in those with PCOS (P < 0.0001); a 2-fold higher prevalence of dementia occurred in those with PCOS, compared to controls.

Conclusion

These findings provide evidence that PCOS is associated with a higher prevalence of morbidities over the lifespan, and the potential scope of the healthcare burden in women affected by PCOS.

Failure to control conventional cardiovascular risk factors in women with type 2 diabetes might explain worse mortality

INTRODUCTION

The standardised mortality rate (SMR) for people with diabetes in England is 1.5-1.7, with differences in outcomes between sexes. There has been little work examining the factors that could have an impact on this or on what may determine sex differences in outcome.

METHODS

Data were extracted for patients with type 2 diabetes (T2D) in Salford (England) in 2010 for the years up to 2020, including any deaths recorded. Expected deaths were calculated from annual Office of National Statistics mortality rate and life expectancy by age and gender, adjusted for the local Index of Multiple Deprivation (IMD). This provided the SMR deprivation (SMRd), and life expectancy years lost per death (LEYLD). The effects of treatment type, and clinical features on SMRd relative to sex were examined by univariable and multivariable analysis.

RESULTS

Data from n = 11,806 (F = 5184; M = 6622) patients were included. Of these, n = 5540 were newly diagnosed and n = 3921 died (F = 1841; M = 2080). In total, n = 78,930 patient years. The expected deaths numbered n = 2596 (adjusted for age, sex, and IMD). Excess deaths were n = 1325 (F = 689; M = 636). Life expectancy years lost (LEYL) 18,989 (F = 9714; M = 9275). SMRd 1.51 (F = 1.60; M = 1.44) and LEYLD 4.84 years (F = 5.28; M = 4.46). The impact of risk factors was not different by sex. However, women had higher prevalence of % diagnosed >65 years of age; % last eGFR <60 mLs/min/1.73 m2 , and lower prevalence of % prescribed ACE-inhibitor/ARB, DPP4-inhibitor and SGLT2-inhibitor. Applying the male prevalence rate to the female population and expected mortality suggested n = 437 (55%) of excess T2D female deaths were attributed to sex difference in the prevalence of these risk and protective factors.

CONCLUSIONS

Outcomes in women with T2DM are worse than in men, contributed to by greater prevalence of adverse factors and less prescribing of cardioprotective medication.

SynthA1c: Towards Clinically Interpretable Patient Representations for Diabetes Risk Stratification

Early diagnosis of Type 2 Diabetes Mellitus (T2DM) is crucial to enable timely therapeutic interventions and lifestyle modifications. As the time available for clinical office visits shortens and medical imaging data become more widely available, patient image data could be used to opportunistically identify patients for additional T2DM diagnostic workup by physicians. We investigated whether image-derived phenotypic data could be leveraged in tabular learning classifier models to predict T2DM risk in an automated fashion to flag high-risk patients without the need for additional blood laboratory measurements. In contrast to traditional binary classifiers, we leverage neural networks and decision tree models to represent patient data as 'SynthA1c' latent variables, which mimic blood hemoglobin A1c empirical lab measurements, that achieve sensitivities as high as 87.6%. To evaluate how SynthA1c models may generalize to other patient populations, we introduce a novel generalizable metric that uses vanilla data augmentation techniques to predict model performance on input out-of-domain covariates. We show that image-derived phenotypes and physical examination data together can accurately predict diabetes risk as a means of opportunistic risk stratification enabled by artificial intelligence and medical imaging. Our code is available at https://github.com/allisonjchae/DMT2RiskAssessment.