Counterfactual Framework Research Articles

Childhood Maltreatment and Revictimization in Young Adulthood: Is Problematic Substance Use a Mediator? A Linked Survey-Register Data Study.

Victims of childhood maltreatment often experience revictimization later in life. However, there is scant knowledge of the explanatory mechanisms that generate this phenomenon. Problematic substance use is a maladaptive coping strategy that may increase the risk of revictimization after maltreatment. We used linear regressions and mediation analyses to investigate the explanatory role of problematic substance use in the association between childhood maltreatment and violent revictimization in young adulthood. We analyzed linked survey-register data on a sample of senior high school graduates in Norway (n = 3,156) who were followed longitudinally until ages 32 to 33 (2021). Ordinary least squares regression analyses showed that childhood maltreatment was associated with both problematic substance use and violent revictimization, adjusted for a wide range of covariates. Moreover, problematic substance use was associated with an increased risk of violent revictimization in young adulthood. Mediation analyses showed that about one eighth (12.0%) of the association between childhood maltreatment and violent revictimization was mediated by problematic substance use. Sensitivity tests indicated that the mediation may be susceptible to unmeasured confounding. Thus, it is unclear whether the mediating role of problematic substance use can be interpreted causally within a counterfactual framework. In conclusion, reducing problematic substance use among victims of childhood maltreatment may mitigate the risk of violent revictimization later in life. Early detection and tailored treatment of problematic substance use could be one way to prevent violent revictimization following childhood maltreatment. Therefore, equitable access to primary and specialized healthcare services is essential, both for victims of maltreatment and individuals with habits of (borderline) problematic substance use. However, the majority of the maltreatment-revictimization pathway (88.0%) is not statistically explained by substance use, which implies that other mechanisms are also at play.

Income inequality, trust, and depressive symptoms among Chinese adults (CFPS): A causal mediation analysis

ObjectiveIncome inequality has been linked to depressive disorders, but the pathways behind this impact are insufficiently understood. Hence, we aimed to investigate the impact of income inequality on depressive disorders and evaluate the extent to which this impact is mediated by trust. MethodsTwo waves (2012 and 2018) of the China Family Panel Studies (CFPS) were included. Depressive symptoms were assessed using the 8-item Center for Epidemiologic Studies Depression scale (CESD8) and income inequality was measured using the Gini index calculated with household income. Based on the counterfactual framework, causal mediation analysis was applied with the difference-in-difference (DID) method. The sequential ignorability assumption, an important assumption for mediation analysis, was examined by propensity score matching (PSM) and simulation-based sensitivity analysis. ResultsCompared to the control group (Change of Gini index ≤0), CESD8 scores in the treatment group (Change of Gini index >0) increase by 0.233 (95 % CI: 0.039, 0.430), which 10.1 % (95 % CI: 3.1 %, 46.0 %) was mediated by reductions in trusts at the provincial level. At the county level, income inequality influences depressive symptoms through the indirect path (β=0.008, 95%CI: 0.001, 0.020) instead of the direct path (β= − 0.146, 95%CI: −0.287, 0.000). Dividing the trust, the significant indirect effect appeared in the trust in neighbors, foreigners, government, and doctors at the provincial level. LimitationThis study didn't deal with the impact of post-treatment confounders of the mediator-outcome relationship. ConclusionsSevere income inequality directly and indirectly exacerbated depressive symptoms. Government should carry out the implementation of decreasing income inequality and improving trust.

Look on the bright side: the relation between family values, positive aspects of care and caregiver burden

Family-centered values are important for caregiving. However, findings on their association with burden are inconsistent. We aim to analyze whether positive aspects of caregiving are mediating the effect of familism on burden among informal caregivers of older adults in Germany. Participants (n = 277) were drawn from the Attitudes Toward Informal Caregivers (ATTIC) project and include informal long-term caregivers of older relatives (aged ≥ 60) quota-sampled from Germany (December 2023). Mediation analyses (linear OLS regression) with robust standard errors were conducted with the classic and the counterfactual causal mediation framework. The classic approach indicated a significant positive direct effect of familism on burden, a significant negative direct effect of PAC on burden and a significant negative indirect effect of familism via PAC on burden; the total effect was not significant. The causal mediation approach supports this; the interaction between familism and PAC was not significant. Thus, sociocultural family-centered values seemed to worsen burden but also to reduce it through positive experiences of caregiving, which did not depend on the strength of familism values. The findings advance our understanding of the mechanisms underlying the stress appraisal of the informal care situation and emphasize the role of positive experiences of care.

Obesity and brain volumes: mediation by cardiometabolic and inflammatory measures

BackgroundThis study aimed to investigate the relationship between overall obesity, central obesity and brain volumes, as well as to determine the extent to which cardiometabolic and inflammatory measures act as...

Effects of exposure on adoption of agricultural smartphone apps among smallholder farmers in Southwest, Nigeria: implications on farm-level-efficiency

BackgroundWhen considering new technologies that are not widely known such as agricultural smartphone apps, exposure plays a significant role in facilitating farmers’ decision to use the agricultural smartphone apps. In this study, we examine the role of exposure to agricultural smartphone apps on adoption of agricultural smartphone apps among smallholder farmers in Southwest, Nigeria and also evaluates the effect of use of agricultural smartphone apps on total factor productivity and technical efficiency of farmers. Through counterfactual framework evaluation approach, we used a multistage sampling procedure to select 380 farmers in Southwest, Nigeria through well-structure questionnaire.ResultsWe found that the agricultural smartphone apps user rate in 2022 for the sub-sample of the exposed farmers was higher in both Oyo and Osun State, indicating that exposure to agricultural smartphone apps played an important and significant role in increasing the adoption and use of agricultural smartphone apps in Nigeria. We also found that the mean TFPI and TE of the treated (users of agricultural smartphone apps) is higher than the non-treated and control group (non-users) implying that the users of agricultural smartphone apps have higher productivity margin than the non-users.ConclusionBased on the result of the study, it was concluded that exposure to the technology has a higher chances of increasing the use of agricultural smartphone apps across farmers populations in Southwest, Nigeria.

Factors Influencing Neuromuscular Blockade Reversal Choice in the United States Before and During the COVID-19 Pandemic: Retrospective Longitudinal Analysis.

Neuromuscular blockade (NMB) agents are a critical component of balanced anesthesia. NMB reversal methods can include spontaneous reversal, sugammadex, or neostigmine and the choice of reversal strategy can depend on various factors. Unanticipated changes to clinical practice emerged due to the COVID-19 pandemic, and a better understanding of how NMB reversal trends were affected by the pandemic may help provide insight into how providers view the tradeoffs in the choice of NMB reversal agents. We aim to analyze NMB reversal agent use patterns for US adult inpatient surgeries before and after the COVID-19 outbreak to determine whether pandemic-related practice changes affected use trends. A retrospective longitudinal analysis of a large all-payer national electronic US health care database (PINC AI Healthcare Database) was conducted to identify the use patterns of NMB reversal during early, middle, and late COVID-19 (EC, MC, and LC, respectively) time periods. Factors associated with NMB reversal choices in inpatient surgeries were assessed before and after the COVID-19 pandemic reached the United States. Multivariate logistic regression assessed the impact of the pandemic on NMB reversal, accounting for patient, clinical, procedural, and site characteristics. A counterfactual framework was used to understand if patient characteristics affected how COVID-19-era patients would have been treated before the pandemic. More than 3.2 million inpatients experiencing over 3.6 million surgical procedures across 931 sites that met all inclusion criteria were identified between March 1, 2017, and December 31, 2021. NMB reversal trends showed a steady increase in reversal with sugammadex over time, with the trend from January 2018 onwards being linear with time (R2>0.99). Multivariate analysis showed that the post-COVID-19 time periods had a small but statistically significant effect on the trend, as measured by the interaction terms of the COVID-19 time periods and the time trend in NMB reversal. A slight increase in the likelihood of sugammadex reversal was observed during EC relative to the pre-COVID-19 trend (odds ratio [OR] 1.008, 95% CI 1.003-1.014; P=.003), followed by negation of that increase during MC (OR 0.992, 95% CI 0.987-0.997; P<.001), and no significant interaction identified during LC (OR 1.001, 95% CI 0.996-1.005; P=.81). Conversely, active reversal (using either sugammadex or neostigmine) did not show a significant association relative to spontaneous reversal, or a change in trend, during EC or MC (P>.05), though a slight decrease in the active reversal trend was observed during LC (OR 0.987, 95% CI 0.983-0.992; P<.001). We observed a steady increase in NMB active reversal overall, and specifically with sugammadex compared to neostigmine, during periods before and after the COVID-19 outbreak. Small, transitory alterations in the NMB reversal trends were observed during the height of the COVID-19 pandemic, though these alterations were independent of the underlying NMB reversal time trends.

Robust inference for causal mediation analysis of recurrent event data.

Recurrent events, including cardiovascular events, are commonly observed in biomedical studies. Understanding the effects of various treatments on recurrent events and investigating the underlying mediation mechanisms by which treatments may reduce the frequency of recurrent events are crucial tasks for researchers. Although causal inference methods for recurrent event data have been proposed, they cannot be used to assess mediation. This study proposed a novel methodology of causal mediation analysis that accommodates recurrent outcomes of interest in a given individual. A formal definition of causal estimands (direct and indirect effects) within a counterfactual framework is given, and empirical expressions for these effects are identified. To estimate these effects, a semiparametric estimator with triple robustness against model misspecification was developed. The proposed methodology was demonstrated in a real-world application. The method was applied to measure the effects of two diabetes drugs on the recurrence of cardiovascular disease and to examine the mediating role of kidney function in this process.

Associations of methylmalonic acid and depressive symptoms with mortality: a population-based study

Methylmalonic acid (MMA), a biomarker of mitochondrial dysfunction, has been reported to be associated with depression in specific populations (i.e., older adults and postpartum women). Our study aimed to investigate to what extent MMA was associated with depressive symptoms and mortality in the general population, and assess whether depressive symptoms mediate the relationship between MMA and mortality. We analyzed cross-sectional data from 8343 participants from the US National Health and Nutrition Examination Survey. MMA was measured by liquid chromatography-tandem mass spectrometry, while depressive symptoms were measured by the Patient Health Questionnaire-9. Mortality data were obtained through linkage with National Death Index records. Linear regression models were performed to assess the association between MMA and depressive symptoms. The Cox proportional hazard regression model was utilized to assess the association of MMA and depressive symptoms with mortality. Mediation analysis was conducted within the counterfactual framework. In this general population, each SD (around 0.49 μmol/L) increase in MMA was associated with a 0.03 SD (approximately 0.15 score) increase in depressive symptoms (β = 0.033, 95% CI: 0.010, 0.055, p = 0.005). Notably, this association was more pronounced in men and participants over 60 years old. Higher levels of MMA and having more depressive symptoms were associated with a higher risk of mortality. However, depressive symptoms do not mediate the relationship between MMA and mortality. Elevated MMA levels were associated with depressive symptoms and an increased risk of mortality. These findings suggest that mitochondrial dysfunction may contribute to the multifactorial etiology of depression.

Causality-based counterfactual explanation for classification models

Counterfactual explanation is one branch of interpretable machine learning that produces a perturbation sample to change the model’s original decision. The generated samples can act as a recommendation for end-users to achieve their desired outputs. Most of the current counterfactual explanation approaches are the gradient-based method, which can only optimize the differentiable loss functions with continuous variables. Accordingly, the gradient-free methods are proposed to handle the categorical variables, which however have several major limitations: (1) causal relationships among features are typically ignored when generating the counterfactuals, possibly resulting in impractical guidelines for decision-makers; (2) the counterfactual explanation algorithm requires a great deal of effort into parameter tuning for determining the optimal weight for each loss functions which must be conducted repeatedly for different datasets and settings. In this work, to address the above limitations, we propose a prototype-based counterfactual explanation framework (ProCE). ProCE is capable of preserving the causal relationship underlying the features of the counterfactual data. In addition, we design a novel gradient-free optimization based on the multi-objective genetic algorithm that generates the counterfactual explanations for the mixed-type of continuous and categorical features. Numerical experiments demonstrate that our method compares favorably with state-of-the-art methods and therefore is applicable to existing prediction models. All the source codes and data are available at https://github.com/tridungduong16/multiobj-scm-cf.

Mediated effects of a randomised control trial for a text messaging smoking cessation intervention for online help-seekers and primary care visitors

Background and AimsDigital smoking cessation interventions have been shown to be effective in helping individuals achieve prolonged smoking abstinence. Nonetheless, the mechanisms that drive such effects are unclear. The current study aimed to estimate a digital smoking cessation intervention's natural direct and indirect effects.MethodsThis secondary analysis of mediated effects uses data from a randomised controlled trial which included participants who smoked at least one cigarette a week, had access to a mobile phone, and were 18 years or older. The comparator was existing smoking cessation support available to all members of the Swedish public. Primary outcomes were prolonged smoking abstinence and point prevalence of smoking abstinence, measured at 3- and 6-months post-randomisation. A counterfactual framework was used to estimate three hypothesised mediators of the intervention's effects: importance, knowledge of how to change (know-how), and confidence.ResultsBetween 18/09/20 and 16/06/22, 1012 participants were randomised. The intervention led to improved confidence and know-how, which both partially mediated the effects of the digital intervention on smoking abstinence at 3- and 6 months post-randomisation.ConclusionsA digital smoking cessation intervention was found to partially affect smoking abstinence by improving individuals’ confidence in their ability to quit smoking and developing knowledge on how to quit. Face-value single-item mediator measures, lack of blinding, and attrition limit the study. Future studies should address these limitations and assess additional mechanisms mediating intervention effects.Trial registrationISRCTN13455271.

Economic performance of the pairing-off poverty alleviation between China’ cities

This study evaluates the economic performance of China's East-West pairing-off poverty alleviation, a policy enacted in 2016 aiming to eradicate absolute poverty by 2020 through substantial fiscal transfers from the assisting cities to the assisted impoverished counties. Employing a counterfactual framework and analyzing data from 2013 to 2018, this research assesses the economic performance of China's East-West pairing-off poverty alleviation. Our findings reveal a negative impact of the fiscal pressure burdened by the assisting cities on their economic performance. Meanwhile, the allocation of anti-poverty funds to the assisted counties does not exhibit a significant impact on their economic outcomes. The mechanism analysis suggests that the fiscal pressure experienced by the assisting cities shows negative impacts on fixed assets investment, innovation activities, and economic growth expectations. Further analysis indicates a negative relationship between the policy and the growth of secondary industry and rural income in the eastern regions, while simultaneously identifying a positive impact on the enterprises' financial burden. These results underscore the complex economic trade-offs inherent in poverty alleviation strategy transferring fiscal funds from developed regions to less developed regions and call for a global understanding of pure cash transfer programmes' impacts.

PSF-C-Net: A Counterfactual Deep Learning Model for Person Re-Identification Based on Random Cropping Patch and Shuffling Filling

In the task of person re-identification (re-ID), capturing the long-range dependency of instances is crucial for accurate identification. The existing methods excel at extracting local features but often overlook the global information of instance images. To address this limitation, we propose a convolution-based counterfactual learning framework, called PSF-C-Net, to focus on global information rather than local detailed features. PSF-C-Net adopts a parameter-sharing dual-path structure to perform counterfactual operations in the prediction space. It takes both the actual instance image and a counterfactual instance image that disrupts the contextual relationship as the input. The counterfactual framework enables the interpretable modeling of global features without introducing additional parameters. Additionally, we propose a novel method for generating counterfactual instance images, which effectively constructs an explicit counterfactual space, to reliably implement counterfactual strategies. We have conducted extensive experiments to evaluate the performance of PSF-C-Net on the Market-1501 and Duke-MTMC-reID datasets. The results demonstrate that PSF-C-Net achieves state-of-the-art performance.

Renal Hyperfiltration as a New Mechanism of Smoking-Related Mortality.

Renal hyperfiltration (RHF), an established risk factor for mortality, is prevalent among tobacco smokers. The aim of this study was to assess the mediating role of RHF in the association between smoking and mortality. Data of this study were retrieved from the cohort of the Kuopio Ischemic Heart Disease Risk Factor Study (KIHD), including 2064 males from Finland. Study participants were followed over a 35-year period. Using classic and counterfactual mediation analysis approaches, we estimated the mediative effect of RHF in the association between smoking and each of the following outcomes: All-cause mortality, cardiovascular disease (CVD) mortality, and non-CVD mortality. The risk of all-cause mortality in smokers was twice that in nonsmokers (hazard ratio [HR], 2.06; 95% confidence interval [CI]: 1.84 to 2.31). Under the counterfactual framework the direct effect of smoking on all-cause mortality, controlled for RHF, corresponded to an HR of 2.00 (95% CI: 1.78 to 2.30). Of the effect of smoking on mortality, 5% (p-value = .016) was mediated by RHF. This finding concerned particularly non-CVD mortality. RHF mediated the effect of smoking on non-CVD and all-cause mortality, but not on CVD mortality. The generalizability of our study results is however limited by its focus on a Finnish male cohort, underscoring the need for further investigation into RHF's broader implications across diverse populations. This study elucidates the complex interplay between smoking, renal hyperfiltration (RHF), and mortality, offering novel insights into the mediating role of RHF. Our findings demonstrate that RHF significantly mediates the relationship between smoking and non-cardiovascular disease (non-CVD), but not CVD mortality. This distinction underscores the multifaceted role of RHF beyond its established association with cardiovascular events. By highlighting the specific pathways through which RHF mediates some of the smoking-attributed mortality, this research contributes to our understanding of the mechanisms linking smoking to mortality.

Dual-Tower Counterfactual Session-Aware Recommender System.

In the complex dynamics of modern information systems such as e-commerce and streaming services, managing uncertainty and leveraging information theory are crucial in enhancing session-aware recommender systems (SARSs). This paper presents an innovative approach to SARSs that combines static long-term and dynamic short-term preferences within a counterfactual causal framework. Our method addresses the shortcomings of current prediction models that tend to capture spurious correlations, leading to biased recommendations. By incorporating a counterfactual viewpoint, we aim to elucidate the causal influences of static long-term preferences on next-item selections and enhance the overall robustness of predictive models. We introduce a dual-tower architecture with a novel data augmentation process and a self-supervised training strategy, tailored to tackle inherent biases and unreliable correlations. Extensive experiments demonstrate the effectiveness of our approach, outperforming existing benchmarks and paving the way for more accurate and reliable session-based recommendations.

Can the Water Resource Fee-to-Tax Reform Promote the “Three-Wheel Drive” of Corporate Green Energy-Saving Innovations? Quasi-Natural Experimental Evidence from China

The long-standing, unrestrained utilization of energy resources by China’s manufacturing sector has created irreversible obstacles to regional sustainable development. Consequently, the Chinese government has implemented a water resource tax policy in certain regions, with the aim of compelling manufacturing enterprises to adopt green and energy-saving innovations. This study used panel data from Chinese manufacturing companies listed on the A-share market from 2009 to 2020 and employed a double machine learning model to explore whether the water resource fee-to-tax reform can compel enterprises to enhance their tripartite green energy-saving innovation drive. These innovations consist of vision-driven and mission-driven green energy-saving technological innovations and green management energy-saving innovations. Following a quasi-natural experiment, our findings revealed the following: (1) The water resource fee-to-tax policy promoted the internal coupling coordination of the triple-driven system. (2) The policy compelled progress in mission-driven green energy-saving technological innovations and green energy-saving management innovations but hindered vision-driven green energy-saving technological innovations. (3) Within the internal systems of manufacturing enterprises, green energy-saving management innovations play a positive mediating role between the water resource fee-to-tax policy and the mission-driven green energy-saving technology innovation subsystem, but they lack a similar positive mediating mechanism for the vision-driven green energy-saving technology innovation subsystem. (4) The counterfactual framework verified that the mechanistic pathway “water resource fee-to-tax → green energy-saving management innovation → mission-driven/vision-driven green energy-saving technological innovation” could be further extended to other manufacturing enterprises not currently under policy compulsion. (5) In the interaction system between manufacturing enterprises and external markets, the development of marketization and financial technology positively regulated the promoting effect of the water resource fee-to-tax policy on mission-driven green energy-saving technological innovations and green energy-saving management innovations, but it did not have a similar effect on vision-driven green energy-saving technological innovations.

Impact of Rural E-Commerce Participation on Farmers’ Household Development Resilience: Evidence from 1229 Farmers in China

This paper investigates the impact of e-commerce participation on household development resilience using a sample of 1229 households in the Shandong and Shaanxi provinces of China in 2022. It constructs the developmental resilience index of farm households from three dimensions of economy, society and culture using the entropy method, and establishes a counterfactual framework using the Propensity Score Matching (PSM) method. The results suggest that participation in e-commerce has a significant and positive impact on farming household development resilience. The PSM method estimates that participation in e-commerce increases the developmental resilience of farming households by 9.63% compared to non-participation, with economic, social, and cultural resilience increasing by 9.29%, 9.84%, and 9.92%, respectively. The robustness test results confirm the findings. Further analysis reveals that participation in e-commerce enhances farm household development resilience through three mechanisms: improving economic efficiency, network relationship linkage, and risk appetite. Heterogeneity analysis shows that the impact of e-commerce participation on household development resilience varies among farmers with different endowment constraints. In particular, farmers with more years of education and cooperative members benefit more from e-commerce participation, especially live and platform e-commerce.

Modeling the causal mechanism in process safety management (PSM) systems from historical accidents

Process safety management (PSM) underpins reducing major accidents in the process industries. For decades, PSM systems have been implemented globally. However, few studies have modeled the complexities within the PSM system, leading to some underlying chronic problems within organizations that go unnoticed. Here, we investigate causal mechanisms among PSM elements from historical accidents to deliver a practical tool for safety intervention formulation. Specifically, based on the Center for Chemical Process Safety's (CCPS) PSM framework, 100 chemical accident reports derived from the U.S. Chemical Safety and Hazard Investigation Board (CSB) were analyzed and converted into structured information stored in 22 PSM elements. Then, employing two causal structure learning algorithms, the causal structures among 22 elements were learned from the processed accident information. Finally, the causal effects among PSM elements were estimated by extending Rubin's counterfactual causal framework to an approach applicable to multi-treatment scenarios. We demonstrate that the causal structures and effects learned from accident reports reveal the causal mechanism within the PSM system. A total of 20 significant causal relationships among single elements, nine significant causal relationships among element combinations, and several interesting PSM failure paths were identified. According to these causal patterns, the PSM effort will shift from the traditional "what to do" to "how to do it."

Exploring a Potential Interaction Between the Effect of Specific Maternal Smoking Patterns and Comorbid Antenatal Depression in Causing Postpartum Depression.

To explore a potential interaction between the effect of specific maternal smoking patterns and the presence of antenatal depression, as independent exposures, in causing postpartum depression (PPD). This case-control study of participants with singleton term births (N = 51220) was based on data from the 2017-2018 Pregnancy Risk Assessment Monitoring System. Multivariable log-binomial regression models examined the main effects of smoking patterns and self-reported symptoms of antenatal depression on the risk of PPD on the adjusted risk ratio (aRR) scale and tested a two-way interaction adjusting for covariates selected in a directed acyclic graph (DAG). The interaction effects were measured on the additive scale using relative excess risk due to interaction (RERI), the attributable proportion of interaction (AP), and the synergy index (SI). Causal effects were defined in a counterfactual framework. The E-value quantified the potential impact of unobserved/unknown covariates, conditional on observed covariates. Among 6841 women in the sample who self-reported PPD, 35.7% also reported symptoms of antenatal depression. Out of 3921 (7.7%) women who reported smoking during pregnancy, 32.6% smoked at high intensity (≥10 cigarettes/day) in all three trimesters and 36.6% had symptoms of antenatal depression. The main effect of PPD was the strongest for women who smoked at high intensity throughout pregnancy (aRR 1.65; 95% CI: 1.63, 1.68). A synergistic interaction was detected, and the effect of all maternal smoking patterns was augmented, particularly in late pregnancy for Increasers and Reducers. Strong associations and interaction effects between maternal smoking patterns and co-occurring antenatal depression support smoking prevention and cessation interventions during pregnancy to lower the likelihood of PPD.

Long-term ambient ozone exposure and incident cardiovascular diseases: National cohort evidence in China

BackgroundLong-term ozone (O3) exposure has been associated with cardiovascular disease (CVD) mortality in mounting cohort evidence, yet its relationship with incident CVD was poorly understood, especially in low- and middle-income countries (LMICs) experiencing high ambient air pollution. MethodsWe carried out a nationwide perspective cohort study from 2010 through 2018 by dynamically enrolling 36948 participants across Chinese mainland. Warm-season (April–September) O3 concentrations were estimated using satellite-based machine-learning models with national coverage. Cox proportional hazards model with time-varying exposures was employed to evaluate the association of long-term O3 exposure with incident CVD (overall CVD, hypertension, stroke, and coronary heart disease [CHD]). Assuming causality, a counterfactual framework was employed to estimate O3-attributable CVD burden based on the exposure-response (E-R) relationship obtained from this study. Decomposition analysis was utilized to quantify the contributions of four key direct driving factors (O3 exposure, population size, age structure, and incidence rate) to the net change of O3-related CVD cases between 2010 and 2018. ResultsA total of 4428 CVD, 2600 hypertension, 1174 stroke, and 337 CHD events were reported during 9-year follow-up. Each 10-μg/m³ increase in warm-season O3 was associated with an incident risk of 1.078 (95% confidence interval [CI]: 1.050–1.106) for overall CVD, 1.098 (95% CI: 1.062–1.135) for hypertension, 1.073 (95% CI: 1.019–1.131) for stroke, and 1.150 (95% CI: 1.038–1.274) for CHD, respectively. We observed no departure from linear E-R relationships of O3 exposure with overall CVD (Pnonlinear= 0.22), hypertension (Pnonlinear= 0.19), stroke (Pnonlinear= 0.70), and CHD (Pnonlinear= 0.44) at a broad concentration range of 60–160 µg/m3. Compared with rural dwellers, those residing in urban areas were at significantly greater O3-associated incident risks of overall CVD, hypertension, and stroke. We estimated 1.22 million (10.6% of overall CVD in 2018) incident CVD cases could be attributable to ambient O3 pollution in 2018, representing an overall 40.9% growth (0.36 million) compared to 2010 (0.87 million, 9.7% of overall CVD in 2010). This remarkable rise in O3-attributable CVD cases was primary driven by population aging (+24.0%), followed by increase in O3 concentration (+10.5%) and population size (+6.7%). ConclusionsLong-term O3 exposure was associated with an elevated risk and burden of incident CVD in Chinese adults, especially among urban dwellers. Our findings underscored policy priorities of implementing joint control measures for fine particulate matter and O3 in the context of accelerated urbanization and population aging in China.

CoreRec: A Counterfactual Correlation Inference for Next Set Recommendation

Next set recommendation aims to predict the items that are likely to be bought in the next purchase. Central to this endeavor is the task of capturing intra-set and cross-set correlations among items. However, the modeling of cross-set correlations poses challenges due to specific issues. Primarily, these correlations are often implicit, and the prevailing approach of establishing an indiscriminate link across the entire set of objects neglects factors like purchase frequency and correlations between purchased items. Such hastily formed connections across sets introduce substantial noise. Additionally, the preeminence of high-frequency items in numerous sets could potentially overshadow and distort correlation modeling with respect to low-frequency items. Thus, we devoted to mitigating misleading inter-set correlations. With a fresh perspective rooted in causality, we delve into the question of whether correlations between a particular item and items from other sets should be relied upon for item representation learning and set prediction. Technically, we introduce the Counterfactual Correlation Inference framework for next set recommendation, denoted as CoreRec. This framework establishes a counterfactual scenario in which the recommendation model impedes cross-set correlations to generate intervened predictions. By contrasting these intervened predictions with the original ones, we gauge the causal impact of inter-set neighbors on set prediction—essentially assessing whether they contribute to spurious correlations. During testing, we introduce a post-trained switch module that selects between set-aware item representations derived from either the original or the counterfactual scenarios. To validate our approach, we extensively experiment using three real-world datasets, affirming both the effectiveness of CoreRec and the cogency of our analytical approach.