While controlled interrupted time series (CITS) are commonly used to evaluate public health policies, how to incorporate control(s) into their statistical modelling has received limited attention. We aimed to compare the statistical performance of different model formulations for including control groups in various segmented regression model specifications (with a particular focus on CITS and Difference-in-Difference [DiD] designs) under conditions where their assumptions are met, as well as when they are violated. Based on a real-world dataset, we simulated and compared the statistical performance of four model formulations grounded on segmented regressions for including control groups in a pre- and post-evaluation. The compared model formulations were: (1) CITS segmented regression, (2) DiD segmented regression, (3) single ITS of the difference between control and intervention series, and (4) incorporating the control as a covariate in a single ITS. Models were tested across scenarios challenging assumptions around the control group (e.g., non-parallel trends -challenging DiD assumptions-, or inconsistent trend difference over time between groups -challenging CITS assumption-) or regression errors (e.g., heteroscedasticity or autocorrelation). We also included models, including restricted cubic splines of time, which may mitigate distortions from assumption violations. Additionally, we tested for detecting non-parallel trends. Standard DiD, CITS, and the ITS of the difference between series yielded the lowest bias whenever their design assumptions were satisfied. Overall, including time splines as covariates into ITS of the difference between series achieved the lowest bias and highest coverage also when design assumptions were violated. This makes it a valuable tool for causal inference in settings with parallel, non-parallel or inconsistent trend patterns between groups. Since violations of the trends assumption are often undetectable, methods robust to such violations are extremely valuable. Modelling CITS as an ITS of the difference between series is among the most robust methods to embed control series into model specifications. Incorporating time splines as model covariates within an ITS of the difference has the potential of reducing bias from assumption violations (including parallel trends) without negative impacts when assumptions hold.

Age-specific melanoma incidence trends in Canada.

Burden of cancer attributable to occupational asbestos exposure in the Americas, 1990-2023: an analysis using the Global Burden of Disease Study 2023.

In 2019, British Columbia (BC) became the first jurisdiction in North America to introduce a mandatory non-medical switching (NMS) policy requiring patients with Crohn's disease (CD) and ulcerative colitis (UC) to be switched from originator biologics to biosimilar equivalents. This study examines the influence of the BC Biosimilars Initiative NMS policy on the longer-term infliximab and adalimumab biosimilar utilization and patterns of continuation/discontinuation of biosimilars. We used health administrative data from BC to study drug utilization trends between 2015 and 2022. We used an interrupted time series design with segmented regression analysis to evaluate the impact of the Biosimilars Initiative on the utilization of biosimilars for infliximab and adalimumab and marginal Cox regression models to explore predictors of discontinuation of biosimilars post NMS. The Biosimilars Initiative was associated with an increase in the proportion of biosimilars within the anti-TNF market of 29.3% (95% CI: 28.6%, 29.9%) in Phase 1 (infliximab) and 49.5% (95% CI: 48.4%, 50.7%) in the adalimumab phase; thereafter, biosimilar use increased by 1.5% per year (95% CI: 0.3%, 2.7%). After the NMS, by the end of 2022, 13.5% (n = 4648) had switched from a biosimilar to an originator or a biologic in a different class. Female patients and biologic-naive patients starting on a biosimilar were more likely to switch from a biosimilar to another drug. Our results suggest that the Biosimilars Initiative NMS in BC was successful in switching patients to biosimilars and that most patients remained on biosimilar versions.

2D: 4D digit ratio, bilateral asymmetry, and insulin resistance in women with obesity.

The burden of pediatric and adolescent thyroid cancer is increasing. This research aimed to study the long-term incidence trends among United States (U.S.) individuals aged 0-19 years from 1999 to 2022, assess heterogeneity by sex, age, race/ethnicity, and state, evaluate COVID-19 pandemic-associated changes, and forecast incidence through 2027 METHODS: This population-based longitudinal ecological analysis used the U.S. Cancer Statistics Incidence Database queried via Centers for Disease Control and Prevention (CDC) Wide-ranging Online Data for Epidemiologic Research (WONDER). Cases were defined using ICD-O-3 site code C73.9. Age-adjusted incidence rates per 100,000 were calculated by demographic strata. Joinpoint-type segmented regression and Poisson regression were used to assess temporal trends and estimate annual percent change and incidence rate ratios (IRR). State-level geographic variation was assessed using mapped age-adjusted rates and global spatial autocorrelation (Moran's I). Forecasts (2023-2027) were generated using Autoregressive Integrated Moving Average (ARIMA) model, Exponential Smoothing State Space (ETS) model, and Poisson regression. From 1999-2022, 17,538 pediatric thyroid cancer cases were diagnosed. Incidence increased steeply with age, peaking at ages 15-19 (crude rate 2.58/100,000). Females had higher incidence than males (age-adjusted 1.4 vs 0.3/100,000). Rates were highest among White (1.0/100,000) and Asian/Pacific Islander (0.9/100,000) children and lowest among Black/African American children (0.3/100,000). Significant spatial clustering was observed (Moran's I=0.39, p < 0.001), with higher rates in Northeastern states and lower rates in the South. A marked decline occurred in 2020, consistent with disrupted diagnosis during the COVID-19 pandemic. Poisson models indicated a sustained upward temporal trend (IRR≈1.032 per year, p < 0.001). Forecasts suggested continued increases through 2027 (≈17% rise over 5 years; ≈37% over 10 years). Pediatric and adolescent thyroid cancer incidence rose steadily in the U.S. from 1999 to 2022, with pronounced socio-demographic heterogeneity and a notable diagnostic disruption in 2020. Continued surveillance and investigation of etiologic drivers and diagnostic intensity are mandatory, alongside efforts to reduce inequities in access to evaluation and care.

Temporal trends in macrovascular complications in young-onset diabetes in Korea: A nationwide population-based study.

The paradox of strong institutions: when better governance intensifies public debt’s drag on economic growth in Heavily Indebted Poor Countries (HIPCs) in Africa

Time-trend analysis of 5,113 consecutive free flaps (1981-2023) from a single surgeon at Chang Gung Memorial Hospital, using segmented regression and joinpoint modelling. 5,113 free flaps included 1,927 ALT, 946 fibula, 1,010 toe transfers, 469 radial, 232 latissimus dorsi and 67 gracilis. ALT introduction was associated with explosive growth (+ 21 flaps/year; p=.003, 1996 to 1999) and rapid substitution of competing soft tissue flaps, with declines in LD, gracilis, and radial forearm. Fibula showed a much slower uptake, with joinpoint models showing moderate growth (+ 2.13 flaps/year; P<.001) until repurposing for mandible reconstruction in 1989. This subsequently correlated with increased site-specific use and tracking of oral cancer incidence (incidence rate ratio = 1.05, p <0.001). Toe transfers declined after national Safety Legislation in 1991, with a 4-year lag (-33.7, P<.001). Microsurgical practice evolves through both technical advances and external forces such as health policy. The contrasting trajectories of the ALT and fibula flaps highlight how innovations reshape reconstructive practice differently. External drivers such as legislation can impact on microsurgical need such as toe transfers. Together these underscore the need for long-range horizons in surgical planning, and the need for intepreting time-trends alongside epidemiological shifts and health policy.

The rapid expansion of healthcare infrastructure may exert increasing pressure on the sustainability of social health insurance. China's concurrent pursuit of universal coverage, coupled with extensive hospital construction, offers a valuable context for examining whether the patterns observed in China align with Roemer's Law (1961), which is often summarized as 'a hospital bed built is a bed filled' within an insured population. This study investigates the relationship between hospital bed density and insurance expenditure to deepen our understanding of the factors associated with the growth of healthcare costs. We compiled a panel dataset encompassing 31 Chinese provinces covering the period from 2011 to 2024. To strengthen identification and address potential endogeneity concerns, we employed two-way fixed effects models alongside instrumental variable (IV-2SLS) estimation. Mediation analysis was employed to investigate potential pathways, while panel threshold regression was utilized to examine nonlinear patterns in the relationship between supply and expenditure. The baseline estimates indicated a positive association, suggesting that higher provincial bed density correlates with increased province-level insurance spending. Mediation analysis revealed that the Average Length of Stay may serve as a potential aggregate pathway, accounting for 17.8% of the estimated relationship. Furthermore, threshold regression analysis indicated a possible nonlinear pattern, with an estimated threshold of 7.271 beds per 1,000 population. Below this threshold, hospital bed density is positively related to insurance spending (β = 0.217); however, above this threshold, the association loses statistical significance. Regional analysis demonstrated that the positive association was most pronounced in the western region, while no statistically significant association was observed in the northeastern region. Our findings support a conditional, province-level interpretation of Roemer's Law within the Chinese context, indicating that the association between bed density and insurance expenditure varies across institutional and capacity settings. Specifically, higher provincial bed density is more strongly associated with a longer average length of stay at the provincial level than with increased admission volumes. Beyond the exploratory threshold estimate, the marginal association between additional bed supply and insurance expenditure appears to weaken. These results suggest that payment reform, length-of-stay management, and regulatory oversight should be prioritized alongside careful capacity planning.

Evidence on legislative reforms to reduce coercive measures remains limited. We evaluated short-to-medium-term effects of Sweden's July 2020 stricter legal regulation of coercive measures in youth < 18 years. In a controlled interrupted time-series study, we used 2017-2021 registry and electronic medical record data from inpatient child and adolescent mental health services (CAMHS) in Region Stockholm, Sweden. Controls included Finnish CAMHS and Region Stockholm young adult psychiatric patients (18-24 years). Monthly outcomes were rates of seclusion, mechanical restraint, involuntary medication and a composite measure, mean durations, and sedative medication use. Segmented regression estimated level- and trend changes, adjusting for pre-intervention trends. Among 7,075 admissions (3,840 individuals) to Stockholm inpatient CAMHS, involuntary medication increased immediately in minors relative to young adults (incidence rate ratio 2.49, 95% CI 1.63-3.81), with no other changes in patient-day-standardised models. Comparisons with Finnish CAHMS provided no consistent evidence of reduced coercion. Mean durations also increased in minors versus young adults (12% for seclusion; 6% for mechanical restraint), while sedative medication use remained stable. Findings were robust in sensitivity analyses. Contrary to the reform's aims, stricter legislation alone was not associated with reduced coercive measure use.

Evaluating the success of policies aiming to reduce early term birth (37+0-38+6 weeks' gestation) is analytically challenging. Determine the best method of reporting temporal changes of early term birth rates. We analysed 54,646 singleton births at Western Australia's main tertiary hospital (2009-2019). Using interrupted time series, segmented and nominal regression, we assessed how varying gestational age outcomes, denominator, or statistical method affected conclusions regarding temporal changes (pre and post-Initiative implementation) in the timing of birth. Interrupted time series showed yearly rates of birth at 37+0-38+6 weeks increased pre-Initiative, with no instant level change or change in the slope post-Initiative, regardless of denominator: all births (slope change -0.0, 95% CI -0.5, 0.5); births > 36+6 weeks (slope change: -0.4, 95% CI -1.0, 0.2), and no change to the number of early term births averted. Conversely, when examining births < 39+0 weeks, pre-Initiative rates increased by 0.9% (95% CI 0.5, 1.3) per year, then stabilised post-Initiative (slope change: -0.8, 95% CI -1.4, -0.3), with no change to the number of births < 39+0 weeks averted. Nominal logistic regression showed that pre-Initiative, the rise in the early term birth rate was driven by a reduction in births at ≥ 39+0 weeks, and post-Initiative by a reduction in preterm birth. Using segmented quantile regression, lengthening of gestational age at birth in days was observed more so at the lower percentiles, compared to the estimated counterfactual. Examining early term birth rates in isolation can mask changes in birth timing in the presence of a changing preterm or ≥ 39+0 birth rate. Researchers should examine all births < 39+0 weeks rather than just early term births, or use nominal logistic regression. For gestational age in days as the outcome, segmented quantile regression can be used.

Cardiometabolic multimorbidity (CMM) is one of the most prevalent patterns of multimorbidity worldwide and presents a growing challenge to public health. Metabolic dysregulation and visceral adipose play central roles in the development of CMM. The atherogenic index of plasma (AIP) has been proposed as a comprehensive indicator of lipid metabolic abnormalities, whereas the body roundness index (BRI) is a novel anthropometric measure reflecting central obesity and visceral adipose tissue (VAT). However, evidence regarding the associations of AIP and BRI with CMM remains limited, particularly in southern Chinese populations and young adults. This study examined the separate and joint associations of AIP and BRI with CMM, aiming to provide preliminary scientific evidence to identify individuals more likely to have prevalent CMM. This cross-sectional study included 2505 adults from the 2024 Guangzhou Residents' Nutrition Survey. Multivariable logistic regression models and segmented logistic regression analyses were employed to examine the association patterns of AIP and BRI with CMM, as well as to assess potential threshold effects. For joint analysis, participants were categorized into four groups by AIP and BRI levels to evaluate the joint association and interaction between these indices and CMM. Among the 2505 participants, 213 (8.50%) were diagnosed with CMM. Compared with the lowest tertile, the highest tertile of AIP (OR = 4.025, 95% CI 2.591-6.455) and BRI (OR = 10.461, 95% CI 5.523-22.496) were associated with a higher likelihood of CMM. AIP was linearly associated with CMM. In contrast, BRI demonstrated a nonlinear association with CMM, with an inflection point at 4.52, below which the odds of CMM increased more rapidly. Joint analyses revealed that participants in the "high AIP+high BRI" group had the strongest association with CMM (OR = 5.081, 95% CI 3.354-7.828). Subgroup analysis revealed that the association between the "high AIP+high BRI" group and CMM was stronger in participants < 60 years. Individuals with elevated levels of AIP and BRI are more likely to have CMM. AIP is linearly associated with CMM, whereas a threshold effect is observed for BRI. The joint assessment of AIP and BRI demonstrates a stronger association with CMM compared to either indicator alone. These findings suggest that the joint assessment of AIP and BRI may be a useful tool for identifying individuals at a higher likelihood of prevalent CMM, particularly in young and middle-aged adults.

Introduction:The yearly observance of Ramadan, which includes prolonged daily fasting, presents a challenging scenario for patients with chronic illnesses, with existing studies presenting inconsistent findings on its metabolic implications.Objective:This study aimed to objectively measure the immediate and sustained impacts of Ramadan fasting on monthly mean Hemoglobin A1c (HbA1c), Systolic Blood Pressure (SBP), and Diastolic Blood Pressure (DBP) in a large, real-world cohort of Saudi Arabian patients with concomitant T2DM and HTN.Methods:A retrospective longitudinal study was carried out using electronic health record (EHR) data from a Saudi multi-facility healthcare system from December 2023 to May 2024. The analysis used a segmented regression for Interrupted Time Series (ITS) to estimate changes in outcomes, with Ramadan (March 2024) serving as the intervention point.Results:The final cohort comprised 15 289 patients. The ITS analysis showed that Ramadan resulted in a substantial reduction in mean HbA1c by −0.25% (P < .001) and mean SBP by −4.12 mmHg (P < .001). However, the post-Ramadan trend for HbA1c demonstrated a significant monthly increase of 0.15% (P = .002), indicating a rebound impact. The reduction in SBP was not sustained, as it returned to pre-Ramadan levels in the months that followed.Conclusions:In this large Saudi cohort, Ramadan fasting was linked to transitory improvements in glycemic and blood pressure control. The ensuing rebound implies that, while fasting can be metabolically advantageous, the effects do not last without ongoing intervention. These results highlight the important necessity for rigorous post-Ramadan clinical follow-up to consolidate any gains made during the fasting period. These findings position Ramadan as a novel window for metabolic reset and a potential preventive opportunity against cardiorenal metabolic syndrome progression in high-risk populations.

Sensory perception in autism is strikingly heterogeneous, with individuals showing both hypo- and hypersensitivity across different sensory domains. While sensory differences are widely recognized as a core feature of autism, the structure and underlying patterns of this variability remain poorly understood. Previous studies have yielded mixed findings, often examining sensory processing in isolation within single domains, thereby limiting a comprehensive understanding of sensory sensitivity in autism. We compiled psychophysical data from 107 autistic and 408 age- and IQ-matched non-autistic individuals across 32 experimental conditions spanning multiple perceptual domains, including size, brightness, orientation, pitch, and face processing. Two complementary statistical approaches were used: segmented regression and a Bayesian hierarchical model. Despite substantial inter- and intra-individual variability, both models revealed a consistent domain-specific pattern: on average, autistic individuals showed reduced sensitivity to faces and speech, while performance on basic non-social tasks was comparable to or exceeded that of the comparison group. Bayesian modelling further indicated that social relevance, rather than domain alone, accounted for the primary source of divergence between groups. This study focused on sensory sensitivity thresholds and did not assess perceptual biases or changes in subjective appearance of the stimuli. A full account of perception in autism requires considering these broader alterations. The current findings suggest that sensory differences in autism reflect a structured perceptual profile shaped by social relevance, stimulus complexity, and individual variability. The results highlight the importance of individualized sensory profiling and may inform both theoretical models and personalized approaches to intervention in autism.

Heterogeneity and endogeneity become increasingly common in statistical modeling and econometric practice. Threshold regression provides a simple yet flexible modeling strategy to account for heterogeneity. This paper studies estimation and inference for multi-threshold regression with endogenous regressors. We exploit a novel two-stage estimation procedure based on instrumental variables, which first identifies a set of possible threshold locations using the group LASSO estimation, and then refines the candidate set by a predetermined selection criterion. Given that the performance of the conventional information criteria is sensitive to the choice of the penalization factor, we develop a data-adaptive threshold-based cross-validation criterion incorporating an order-preserved sample-splitting strategy to determine the number of thresholds. Regarding inference, we develop test statistics to test for the presence of threshold effects and the existence of endogeneity, respectively. Numerical simulations and an application analyzing the threshold effects of 401(k) plans on wealth demonstrate the excellent finite sample performance of our methods. Finally, we develop a user-friendly R package MultiThreshold to implement the methodologies.

Heterogeneous Rebound Patterns of Foodborne Diseases after COVID-19 Non-Pharmaceutical Interventions, South Korea, 2005-2024.

Colon cancer is a major subtype of colorectal cancer and one of the leading causes of cancer-related morbidity and mortality worldwide. Age is a well-established risk factor, yet the precise nature of its association with colon cancer risk – particularly potential nonlinear effects – remains underexplored in large population-based datasets. We conducted a cross-sectional analysis using data from 16,982 adults aged ≥ 18 years who participated in the 2011 to 2018 National Health and Nutrition Examination Survey. Colon cancer status was determined through self-reported medical history. Multivariable logistic regression models were used to assess the independent association between age and colon cancer prevalence, adjusting for sociodemographic, anthropometric, and behavioral covariates. Segmented regression was applied to evaluate potential nonlinear threshold effects. Among 16,982 participants, 117 (0.69%) reported a diagnosis of colon cancer. Age was significantly associated with colon cancer prevalence across all models (adjusted OR = 1.047 per year, 95% CI: 1.032–1.063, P < .001). Subgroup analyses confirmed consistent associations across strata of smoking, diabetes, and hypertension. Segmented regression identified a nonlinear relationship with a threshold at 59 years: below this age, risk increased modestly (OR = 1.030), but rose more steeply thereafter (OR = 1.080). This nationally representative study highlights a strong and nonlinear association between aging and colon cancer prevalence, with a significant risk acceleration after age 59. These findings support the need for age-informed screening strategies and underscore the importance of early intervention and prevention efforts targeted toward older adults.

Diabetes mellitus (DM) significantly elevates cerebrovascular risk in stroke survivors, yet metabolic predictors specific to this population remain underexplored. The uric acid-to-high-density lipoprotein cholesterol ratio (UHR) encapsulates pro-oxidant and anti-atherogenic pathways relevant to DM pathogenesis, but its clinical utility in stroke cohorts is unestablished. This cross-sectional study analyzed 1999 to 2018 National Health and Nutrition Examination Survey data from 964 adults with stroke history (representing 3,481,079 U.S. adults). UHR was standardized into sex-stratified z-scores. DM was defined by American Diabetes Association criteria. Generalized variance inflation factors (GVIF) confirmed absence of multicollinearity (all GVIF(1/(2*Df)) < 5). Weighted multivariable logistic regression assessed UHR-DM associations across 3 models: unadjusted (model 1), sex/age-adjusted (model 2), and fully adjusted for clinical covariates (model 3). Restricted cubic splines evaluated nonlinearity, piecewise regression identified effect thresholds, and time-dependent receiver operating characteristic analysis assessed predictive performance with inverse probability weighting. Mediation analysis quantified body mass index (BMI)’s intermediary role. Elevated UHR z-scores demonstrated dose-dependent DM associations, with quartile 4 (Q4) exhibiting 3.12-fold higher odds (95% confidence interval [CI]: 1.86–5.25, P < .001) in model 3. Restricted cubic splines revealed significant nonlinearity (P < .05), identifying UHR z-score = −0.419 (raw value = 0.088) as an inflection point: below this threshold, each 1-unit increase conferred 3.39-fold higher DM risk (95% CI: 1.56–7.40, P = .002); above it, risk increased 1.20-fold (95% CI: 1.08–1.45, P = .041). BMI mediated 38% of the total effect (β = 0.04, 95% CI: 0.03–0.06; P < .001). A rea under the curve (AUC) for UHR z-scores predicting DM improved from 0.61 (unadjusted) to 0.68 (model 3-adjusted). Subgroup analyses showed effect modification by sex (P for interaction < .05). In stroke survivors, UHR elevation independently suggests DM risk through threshold-dependent mechanisms, with BMI mediating over one-third of this relationship. UHR assessment could enhance DM risk stratification in poststroke care.

Impact of leg length discrepancy on gait kinematics and biomechanics.