Direct Causal Effects Research Articles

Interpretation of Mendelian randomization using a single measure of an exposure that varies over time.

Mendelian randomization (MR) is a powerful tool through which the causal effects of modifiable exposures on outcomes can be estimated from observational data. Most exposures vary throughout the life course, but MR is commonly applied to one measurement of an exposure (e.g. weight measured once between ages 40 and 60 years). It has been argued that MR provides biased causal effect estimates when applied to one measure of an exposure that varies over time. We propose an approach that emphasizes the liability that causes the entire exposure trajectory. We demonstrate this approach using simulations and an applied example. We show that rather than estimating the direct or total causal effect of changing the exposure value at a given time, MR estimates the causal effect of changing the underlying liability for the exposure, scaled to the effect of the liability on the exposure at that time. As such, results from MR conducted at different time points are expected to differ (unless the effect of the liability on exposure is constant over time), as we illustrate by estimating the effect of body mass index measured at different ages on systolic blood pressure. Univariable MR results should not be interpreted as time-point-specific direct or total causal effects, but as the effect of changing the liability for the exposure. Estimates of how the effects of a genetic variant on an exposure vary over time, together with biological knowledge that provides evidence regarding likely effective exposure periods, are required to interpret time-point-specific causal effects.

Evaluating causes of effects by posterior effects of causes

SummaryFor the case with a single causal variable, Dawid et al. (2014) defined the probability of causation, and Pearl (2000) defined the probability of necessity to assess the causes of effects. For a case with multiple causes that could affect each other, this paper defines the posterior total and direct causal effects based on the evidence observed for post-treatment variables, which could be viewed as measurements of causes of effects. Posterior causal effects involve the probabilities of counterfactual variables. Thus, as with the probability of causation, the probability of necessity and direct causal effects, the identifiability of posterior total and direct causal effects requires more assumptions than the identifiability of traditional causal effects conditional on pre-treatment variables. We present assumptions required for the identifiability of posterior causal effects and provide identification equations. Further, when the causal relationships between multiple causes and an endpoint can be depicted by causal networks, we can simplify both the required assumptions and the identification equations of the posterior total and direct causal effects. Finally, using numerical examples, we compare the posterior total and direct causal effects with other measures for evaluating the causes of effects and the population attributable risks.

Causality of anthropometric markers associated with polycystic ovarian syndrome: Findings of a Mendelian randomization study.

Using body mass index (BMI) as a proxy, previous Mendelian randomization (MR) studies found total causal effects of general obesity on polycystic ovarian syndrome (PCOS). Hitherto, total and direct causal effects of general- and central obesity on PCOS have not been comprehensively analyzed. To investigate the causality of central- and general obesity on PCOS using surrogate anthropometric markers. Summary GWAS data of female-only, large-sample cohorts of European ancestry were retrieved for anthropometric markers of central obesity (waist circumference (WC), hip circumference (HC), waist-to-hip ratio (WHR)) and general obesity (BMI and its constituent variables-weight and height), from the IEU Open GWAS Project. As the outcome, we acquired summary data from a large-sample GWAS (118870 samples; 642 cases and 118228 controls) within the FinnGen cohort. Total causal effects were assessed via univariable two-sample Mendelian randomization (2SMR). Genetic architectures underlying causal associations were explored. Direct causal effects were analyzed by multivariable MR modelling. Instrumental variables demonstrated no weak instrument bias (F > 10). Four anthropometric exposures, namely, weight (2.69-77.05), BMI (OR: 2.90-4.06), WC (OR: 6.22-20.27), and HC (OR: 6.22-20.27) demonstrated total causal effects as per univariable 2SMR models. We uncovered shared and non-shared genetic architectures underlying causal associations. Direct causal effects of WC and HC on PCOS were revealed by two multivariable MR models containing exclusively the anthropometric markers of central obesity. Other multivariable MR models containing anthropometric markers of both central- and general obesity showed no direct causal effects on PCOS. Both and general- and central obesity yield total causal effects on PCOS. Findings also indicated potential direct causal effects of normal weight-central obesity and more complex causal mechanisms when both central- and general obesity are present. Results underscore the importance of addressing both central- and general obesity for optimizing PCOS care.

The Hyperintense study: Assessing the effects of induced blood pressure increase and decrease on MRI markers of cerebral small vessel disease: Study rationale and protocol.

Background:Neuroimaging markers of cerebral small vessel disease (SVD) are common in older individuals, but the pathophysiological mechanisms causing these lesions remain poorly understood. Although hypertension is a major risk factor for SVD, the direct causal effects of increased blood pressure are unknown. The Hyperintense study is designed to examine cerebrovascular and structural abnormalities, possibly preceding SVD, in young adults with hypertension. These patients undergo a diagnostic work-up that requires patients to temporarily discontinue their antihypertensive agents, often leading to an increase in blood pressure followed by a decrease once effective medication is restarted. This allows examination of the effects of blood pressure increase and decrease on the cerebral small vessels.Methods:Hyperintense is a prospective observational cohort study in 50 hypertensive adults (18–55 years) who will temporarily discontinue antihypertensive medication for diagnostic purposes. MRI and clinical data is collected at four timepoints: before medication withdrawal (baseline), once antihypertensives are largely or completely withdrawn (T = 1), when patients have restarted medication (T = 2) and reached target blood pressure and 1 year later (T = 3). The 3T MRI protocol includes conventional structural sequences and advanced techniques to assess various aspects of microvascular integrity, including blood-brain barrier function using Dynamic Contrast Enhanced MRI, white matter integrity, and microperfusion. Clinical assessments include motor and cognitive examinations and blood sampling.Discussion:The Hyperintense study will improve the understanding of the pathophysiological mechanisms following hypertension that may cause SVD. This knowledge can ultimately help to identify new targets for treatment of SVD, aimed at prevention or limiting disease progression.

Trends in dispensing of individual prescription opioid formulations, Canada 2005\u20132020

BackgroundCanada has experienced a distinctly bifurcated pattern of (strong) opioid utilization post-2000, with multifold increases rendering it one of the world’s highest opioid consumption rates, followed by subsequent substantive declines since 2011/2012. Several interventions to control especially high-risk opioid use have been implemented post-2010 at different levels, yet with their effects assessed mostly for overall opioid utilization. Little knowledge exists for over-time patterns of individual opioid formulations.MethodsRaw information on community-based prescription opioid dispensing for years 2005–2020 were obtained from a large national database based on a stratified sample of 6500 retail pharmacies across Canada (IQVIA/Compuscript), These data were converted into Defined-Daily-Doses/1000 population/day (DDD/1000/day) for individual (strong and weak) opioid formulations—specifically: fentanyl, hydromorphone, hydrocodone, morphine, oxycodone, codeine—per standard methods. Descriptive data on individual opioid dispensing were computed, and segmented regression (or ‘broken-stick’) analysis was applied to the overtime dispensing towards assessing potentially significant ‘breakpoints’ interrupting linear utilization trends. Akaike information criterion (AIC) values were computed to assess the resulting models’ quality-of-fit.ResultsFive of the six opioid formulations featured a lower dispensing level in 2020 compared with 2005, but mostly with peak values in years between, contributing to the overall inversion pattern. For five of the six opioid formulations, a three-segmented model emerged as the best fit for the dispensing observed; only hydrocodone presented a linear (downward) dispensing trend. Among the five interrupted trend models for individual formulations, four (fentanyl, morphine, oxycodone, codeine but not hydromorphone) indicated their initial breakpoint during 2011–2014 introducing a downward dispensing trend. Inconsistently, morphine also featured a recent breakpoint (2018) towards a dispensing increase.ConclusionsWhile all opioids showed marked declines, we found heterogeneous patterns of dispensing for individual opioid formulations. While we cannot estimate direct causal effects, opioid control interventions appear to have had differential impacts on dispensing of individual formulations. The earliest breakpoint occurred towards substantive decreases for oxycodone dispensing in 2011; subsequently, there were increases in dispensing of hydromorphone and fentanyl likely due to substitution effects, followed by across-the-board declines post-2015/2016. Recent ‘safer opioid’ distribution programs to reduce illicit/toxic opioid exposure linked with high levels of poisoning fatalities seem to fuel resurgences in select opioid (e.g., morphine) dispensing.

Bayesian identification of structural coefficients in causal models and the causal false-positive risk of confounders and colliders in linear Markovian models

BackgroundCausal inference has seen an increasing popularity in medical research. Estimation of causal effects from observational data allows to draw conclusions from data when randomized controlled trials cannot be conducted. Although the identification of structural causal models (SCM) and the calculation of structural coefficients has received much attention, a key requirement for valid causal inference is that conclusions are drawn based on the true data-generating model.MethodsIt remains widely unknown how large the probability is to reject the true structural causal model when observational data from it is sampled. The latter probability – the causal false-positive risk – is crucial, as rejection of the true causal model can induce bias in the estimation of causal effects. In this paper, the widely used causal models of confounders and colliders are studied regarding their causal false-positive risk in linear Markovian models. A simulation study is carried out which investigates the causal false-positive risk in Gaussian linear Markovian models. Therefore, the testable implications of the DAG corresponding to confounders and colliders are analyzed from a Bayesian perspective. Furthermore, the induced bias in estimating the structural coefficients and causal effects is studied.ResultsResults show that the false-positive risk of rejecting a true SCM of even simple building blocks like confounders and colliders is substantial. Importantly, estimation of average, direct and indirect causal effects can become strongly biased if a true model is rejected. The causal false-positive risk may thus serve as an indicator or proxy for the induced bias.ConclusionWhile the identification of structural coefficients and testable implications of causal models have been studied rigorously in the literature, this paper shows that causal inference also must develop new concepts for controlling the causal false-positive risk. Although a high risk cannot be equated with a substantial bias, it is indicative of the induced bias. The latter fact calls for the development of more advanced risk measures for committing a causal type I error in causal inference.

Average direct and indirect causal effects under interference

Summary We propose a definition for the average indirect effect of a binary treatment in the potential outcomes model for causal inference under cross-unit interference. Our definition is analogous to the standard definition of the average direct effect and can be expressed without needing to compare outcomes across multiple randomized experiments. We show that the proposed indirect effect satisfies a decomposition theorem stating that in a Bernoulli trial, the sum of the average direct and indirect effects always corresponds to the effect of a policy intervention that infinitesimally increases treatment probabilities. We also consider a number of parametric models for interference and find that our nonparametric indirect effect remains a natural estimand when re-expressed in the context of these models.

Effectiveness of Localized Lockdowns in the COVID-19 Pandemic.

Non-pharmaceutical interventions, such as social distancing and lockdowns, have been essential to control the COVID-19 pandemic. In particular, localized lockdowns in small geographic areas have become an important policy intervention to prevent viral spread in cases of resurgence. These localized lockdowns can result in lower social and economic costs compared to larger-scale suppression strategies. Using an integrated dataset from Chile (March 3 through June 15, 2020) and a novel synthetic control approach, in this paper we estimate the effect of localized lockdowns, disentangling its direct and indirect causal effects on SARS-CoV-2 transmission. Our results show that the effects of localized lockdowns are strongly modulated by their duration and are influenced by indirect effects from neighboring geographic areas. Our estimates suggest that extending localized lockdowns can slow down the pandemic; however, localized lockdowns on their own are insufficient to control pandemic growth in the presence of indirect effects from contiguous neighboring areas that do not have lockdowns. These results provide critical empirical evidence about the effectiveness of localized lockdowns in interconnected geographic areas.

Do Patient Experiences Have Mediating Roles on Patient Loyalty?

The study aimed to evaluate the mediating roles of patient experiences on patient loyalty. The data were collected through an electronic questionnaire regarding feedback from 5732 patients received outpatient clinics. Patient loyalty was evaluated using the Net Promoter Score (NPS11) that patients were asked whether they would like to recommend the hospital to their relatives or friends. Patient experiences with physicians, nurses, and waiting times were also asked in the questionnaire. After preliminary analysis, mediation analyses were performed to evaluate direct and indirect causal effects among variables for NPS11. While patient experiences are used as possible mediators, Branch Groups in the first and Admission Time in the second model are independent variables. In the analyses, Surgical Medical Science (p = 0.019) and Day Shift (p = 0.000) have a direct mediating effect on NPS11. Nursing care experiences were found to be a mediator variable for NPS11 in both models (p = 0.000 for both). Patient loyalty was associated with Surgical Medical Science and Day Shift primarily whereas Nursing care experience had a mediating role.

Obesity and Kidney Function: A Two-Sample Mendelian Randomization Study.

Obesity and type 2 diabetes (T2D) are correlated risk factors for chronic kidney disease (CKD). Using summary data from GIANT (Genetic Investigation of Anthropometric Traits), DIAGRAM (DIAbetes Genetics Replication And Meta-analysis), and CKDGen (CKD Genetics), we examined causality and directionality of the association between obesity and kidney function. Bidirectional 2-sample Mendelian randomization (MR) estimated the total causal effects of body mass index (BMI) and waist-to-hip ratio (WHR) on kidney function, and vice versa. Effects of adverse obesity and T2D were examined by stratifying BMI variants by their association with WHR and T2D. Multivariable MR estimated the direct causal effects of BMI and WHR on kidney function. The inverse variance weighted random-effects MR for Europeans was the main analysis, accompanied by several sensitivity MR analyses. One standard deviation (SD ≈ 4.8 kg/m2) genetically higher BMI was associated with decreased estimated glomerular filtration rate (eGFR) [β=-0.032 (95% confidence intervals: -0.036, -0.027) log[eGFR], P = 1 × 10-43], increased blood urea nitrogen (BUN) [β = 0.010 (0.005, 0.015) log[BUN], P = 3 × 10-6], increased urinary albumin-to-creatinine ratio [β = 0.199 (0.067, 0.332) log[urinary albumin-to-creatinine ratio (UACR)], P = 0.003] in individuals with diabetes, and increased risk of microalbuminuria [odds ratios (OR) = 1.15 [1.04-1.28], P = 0.009] and CKD [1.13 (1.07-1.19), P = 3 × 10-6]. Corresponding estimates for WHR and for trans-ethnic populations were overall similar. The associations were driven by adverse obesity, and for microalbuminuria additionally by T2D. While genetically high BMI, unlike WHR, was directly associated with eGFR, BUN, and CKD, the pathway to albuminuria was likely through T2D. Genetically predicted kidney function was not associated with BMI or WHR. Genetically high BMI is associated with impaired kidney function, driven by adverse obesity, and for albuminuria additionally by T2D.

Type 2 diabetes mellitus increases risk of erectile dysfunction independent of obesity and dyslipidemia: A Mendelian randomization study.

The causal effects of individual risk factors of metabolic syndrome on erectile dysfunction (ED) are still unclear. To evaluate the causal effect of risk factors of metabolic syndrome on ED through Mendelian randomization (MR). Data for risk factors were obtained from multiple databases with 173,082-757,601 individuals, and that for ED were collected from a genome-wide association study including 223,805 Europeans. We performed univariate MR analysis using inverse-variance weighted, MR-Egger, weighted-median, weighted mode methods and multivariable MR analysis to evaluate the total and direct causal effects. The univariable MR supported that type 2 diabetes mellitus (odds ratios [OR]=1.14, 95% confidence intervals [CI]: 1.08-1.21, p<0.001) and body mass index (BMI) (OR=1.27, 95% CI: 1.12-1.44, p<0.001) were associated with ED. After excluding the SNPs associated with BMI and other risk factors, the results of multivariable MR for T2D (OR=1.15, 95% CI: 1.05-1.25, p=0.001) remained consistent. However, the results of multivariable MR provided limited evidence for the causality between BMI and ED (OR=1.06, 95% CI: 0.88-1.29, p=0.532). For systolic blood pressure and lipid components (low-density lipoprotein, high-density lipoprotein and triglycerides), both univariable and multivariable MR failed to offer sufficient evidence to confirm their causal effect on ED. T2D showed a direct causal effect on ED independent of obesity and dyslipidemia.

Racial Disparities Affect Timing of Alveolar Bone Grafting for Patients With Cleft Lip and Palate: An Analysis of the 2012 Kids’ Inpatient Database

Objective: The timing of alveolar bone graft (ABG) for patients with cleft lip and palate is critical to the success of the operation. Sociodemographic factors have been implicated in delaying surgery for other diagnoses but their impact on patients undergoing ABG is unknown. The purpose of this study is to examine the association of sociodemographic factors on the age of patients undergoing ABG. Methods: The Kids’ Inpatient Database (KID) from 2012 was used to define a cohort of patients with cleft lip and/or palate between the ages of 5 and 14. Patients were selected using ICD-9 procedure codes for repair of oronasal fistula and bone graft. Gender, race, income quartile, hospital region, and type of insurance were used as independent variables. Time to ABG was then calculated as hazard ratios (HR) of having surgery using Cox regression models with 95% confidence intervals. Results: A cohort of 395 patients was identified. Median age at ABG was 9 ± 0.15 years. On univariate analysis, Non-Hispanic Black (NHB) patients underwent surgery later than Non-Hispanic White (NHW) patients (11.13 ± 0.44 years vs 9.67 ± 0.19 years, P = .0026). Patients in the higher 2 income quartiles were more likely to have an early surgery as compared to those in the lowest quartile (9.16 ± 0.43 and 9.32 ± 0.43 years vs 10.20 ± 0.33, P = .016 and 0.043 respectively). No factor was significant on multivariate analysis. Conclusions: Patients who are NHB are associated with delayed time to ABG as compared to NHW. Conversely, those in the upper income quartile by zip code underwent surgery at an earlier age. While this study identifies demographic factors associated with delay of ABG, more research is needed to understand the direct causal effects for this delay in surgery and its implication for future growth and need for orthognathic surgery.

Raising doubt about the anticipated consequences of suicidal behavior: Evidence for a new approach from laboratory and real-world experiments

ObjectiveConverging evidence from basic science and experimental suicide research suggest that the anticipated consequences of suicide may have direct causal effects on suicidal behavior and accordingly represent a promising intervention target. Raising doubt about individuals’ desirable anticipated consequences of suicide may be one means of disrupting this target. We tested this possibility across two complementary experimental studies. MethodStudy 1 tested the effects of raising doubt about desirable anticipated consequences on virtual reality (VR) suicide in the lab, randomizing 413 participants across four conditions. In Study 2, 226 suicidal adults were randomized to an anticipated consequence manipulation or control condition then re-assessed at 2- and 8-weeks post-baseline. ResultsIn Study 1, anticipating that engaging in VR suicide would guarantee a desirable outcome significantly increased the VR suicide rate; conversely, raising doubt about the desirable anticipated consequences significantly reduced the VR suicide rate. In Study 2, raising doubt about the anticipated consequences of attempting suicide by firearm significantly reduced the perceived lethality of firearms as well as self-predicted likelihood of future suicide attempts, with effects sustained at 2-week follow-up. ConclusionsFindings suggest that raising doubt about desirable anticipated consequences of suicide merits further research as one potential approach to inhibit suicidal behavior.

The assessment of presenteeism and activity impairment in Behçet's syndrome and recurrent aphthous stomatitis: a multicentre study.

To evaluate key factors for presenteeism and activity impairment in multinational patients with Behçet's syndrome (BS) and recurrent aphthous stomatitis (RAS). In this cross-sectional study, 364 BS patients from Jordan, Brazil, the United Kingdom and Turkey and 143 RAS patients from the United Kingdom and Turkey were included. The Work Productivity Activity Impairment (WPAI) scale was used for presenteeism and activity impairment. Mediation analyses were performed to evaluate both direct and indirect causal effects. Presenteeism score was higher in active patients with genital ulcers and eye involvement as well as patients with comorbidities and current smokers than the others in BS (P < 0.05). In RAS, presenteeism score was elevated by oral ulcer activity in the direct path (P = 0.0073) and long disease duration as a mediator in the indirect path (P = 0.0191). Patients with active joint involvement had poor scores in absenteeism, presenteeism, overall impairment and activity impairment compared with those of inactive patients (P < 0.05). Using mediation analysis, the activity impairment score was directly mediated by joint activity (P = 0.0001) and indirectly mediated through oral ulcer-related pain in BS (P = 0.0309). In BS, presenteeism was associated with disease activity, presence of comorbidities and being a current smoker, whereas in RAS, presenteeism was associated with oral ulcer activity and increased length of the disease. Moreover, activity impairment was adversely affected by joint activity and oral ulcer related pain in BS. Patients need to be empowered by using appropriate treatment strategies in their working environment and daily life.

Effect of specific non-pharmaceutical intervention policies on SARS-CoV-2 transmission in the counties of the United States

Non-pharmaceutical interventions (NPIs) remain the only widely available tool for controlling the ongoing SARS-CoV-2 pandemic. We estimated weekly values of the effective basic reproductive number (Reff) using a mechanistic metapopulation model and associated these with county-level characteristics and NPIs in the United States (US). Interventions that included school and leisure activities closure and nursing home visiting bans were all associated with a median Reff below 1 when combined with either stay at home orders (median Reff 0.97, 95% confidence interval (CI) 0.58–1.39) or face masks (median Reff 0.97, 95% CI 0.58–1.39). While direct causal effects of interventions remain unclear, our results suggest that relaxation of some NPIs will need to be counterbalanced by continuation and/or implementation of others.

The intestinal microbiota contributes to the growth and physiological state of muscle tissue in piglets

Although the importance of the intestinal microbiota in host growth and health is well known, the relationship between microbiota colonization and muscle development is unclear. In this study, the direct causal effects of the colonization of gut microorganisms on the muscle tissue of piglets were investigated. The body weight and lean mass of germ-free (GF) piglets were approximately 40% lower than those of normal piglets. The deletion of the intestinal microbiota led to weakened muscle function and a reduction in myogenic regulatory proteins, such as MyoG and MyoD, in GF piglets. In addition, the blinded IGF1/AKT/mTOR pathway in GF piglets caused muscle atrophy and autophagy, which were characterized by the high expression of Murf-1 and KLF15. Gut microbiota introduced to GF piglets via fecal microbiota transplantation not only colonized the gut but also partially restored muscle growth and development. Furthermore, the proportion of slow-twitch muscle fibers was lower in the muscle of GF piglets, which was caused by the reduced short-chain fatty acid content in the circulation and impaired mitochondrial function in muscle. Collectively, these findings suggest that the growth, development and function of skeletal muscle in animals are mediated by the intestinal microbiota.

Genomic Approaches Reveal Pleiotropic Effects in Crossbred Beef Cattle.

Carcass and meat quality are two important attributes for the beef industry because they drive profitability and consumer demand. These traits are of even greater importance in crossbred cattle used in subtropical and tropical regions for their superior adaptability because they tend to underperform compared to their purebred counterparts. Many of these traits are challenging and expensive to measure and unavailable until late in life or after the animal is harvested, hence unrealistic to improve through traditional phenotypic selection, but perfect candidates for genomic selection. Before genomic selection can be implemented in crossbred populations, it is important to explore if pleiotropic effects exist between carcass and meat quality traits. Therefore, the objective of this study was to identify genomic regions with pleiotropic effects on carcass and meat quality traits in a multibreed Angus–Brahman population that included purebred and crossbred animals. Data included phenotypes for 10 carcass and meat quality traits from 2,384 steers, of which 1,038 were genotyped with the GGP Bovine F-250. Single-trait genome-wide association studies were first used to investigate the relevance of direct additive genetic effects on each carcass, sensory and visual meat quality traits. A second analysis for each trait included all other phenotypes as covariates to correct for direct causal effects from identified genomic regions with pure direct effects on the trait under analysis. Five genomic windows on chromosomes BTA5, BTA7, BTA18, and BTA29 explained more than 1% of additive genetic variance of two or more traits. Moreover, three suggestive pleiotropic regions were identified on BTA10 and BTA19. The 317 genes uncovered in pleiotropic regions included anchoring and cytoskeletal proteins, key players in cell growth, muscle development, lipid metabolism and fat deposition, and important factors in muscle proteolysis. A functional analysis of these genes revealed GO terms directly related to carcass quality, meat quality, and tenderness in beef cattle, including calcium-related processes, cell signaling, and modulation of cell–cell adhesion. These results contribute with novel information about the complex genetic architecture and pleiotropic effects of carcass and meat quality traits in crossbred beef cattle.

Contributions of Age-Related and Audibility-Related Deficits to Aided Consonant Identification in Presbycusis: A Causal-Inference Analysis.

The decline of speech intelligibility in presbycusis can be regarded as resulting from the combined contribution of two main groups of factors: (1) audibility-related factors and (2) age-related factors. In particular, there is now an abundant scientific literature on the crucial role of suprathreshold auditory abilities and cognitive functions, which have been found to decline with age even in the absence of audiometric hearing loss. However, researchers investigating the direct effect of aging in presbycusis have to deal with the methodological issue that age and peripheral hearing loss covary to a large extent. In the present study, we analyzed a dataset of consonant-identification scores measured in quiet and in noise for a large cohort (n = 459, age = 42–92) of hearing-impaired (HI) and normal-hearing (NH) listeners. HI listeners were provided with a frequency-dependent amplification adjusted to their audiometric profile. Their scores in the two conditions were predicted from their pure-tone average (PTA) and age, as well as from their Extended Speech Intelligibility Index (ESII), a measure of the impact of audibility loss on speech intelligibility. We relied on a causal-inference approach combined with Bayesian modeling to disentangle the direct causal effects of age and audibility on intelligibility from the indirect effect of age on hearing loss. The analysis revealed that the direct effect of PTA on HI intelligibility scores was 5 times higher than the effect of age. This overwhelming effect of PTA was not due to a residual audibility loss despite amplification, as confirmed by a ESII-based model. More plausibly, the marginal role of age could be a consequence of the relatively little cognitively-demanding task used in this study. Furthermore, the amount of variance in intelligibility scores was smaller for NH than HI listeners, even after accounting for age and audibility, reflecting the presence of additional suprathreshold deficits in the latter group. Although the non-sense-syllable materials and the particular amplification settings used in this study potentially restrict the generalization of the findings, we think that these promising results call for a wider use of causal-inference analysis in audiology, e.g., as a way to disentangle the influence of the various cognitive factors and suprathreshold deficits associated to presbycusis.

Simpson's Paradox in COVID-19 Case Fatality Rates: A Mediation Analysis of Age-Related Causal Effects.

We point out an instantiation of Simpson's paradox in COVID-19 case fatality rates (cfrs): comparing a large-scale study from China (February 17) with early reports from Italy (March 9), we find that cfrs are lower in Italy for every age group, but higher overall. This phenomenon is explained by a stark difference in case demographic between the two countries. Using this as a motivating example, we introduce basic concepts from mediation analysis and show how these can be used to quantify different direct and indirect effects when assuming a coarse-grained causal graph involving country, age, and case fatality. We curate an age-stratified cfr dataset with n}{}>750 k cases and conduct a case study, investigating total, direct, and indirect (age-mediated) causal effects between different countries and at different points in time. This allows us to separate age-related effects from others unrelated to age and facilitates a more transparent comparison of cfrs across countries at different stages of the COVID-19 pandemic. Using longitudinal data from Italy, we discover a sign reversal of the direct causal effect in mid-March, which temporally aligns with the reported collapse of the healthcare system in parts of the country. Moreover, we find that direct and indirect effects across 132 pairs of countries are only weakly correlated, suggesting that a country's policy and case demographic may be largely unrelated. We point out limitations and extensions for future work, and finally, discuss the role of causal reasoning in the broader context of using AI to combat the COVID-19 pandemic.Impact Statement—During a global pandemic, understanding the causal effects of risk factors such as age on COVID-19 fatality is an important scientific question. Since randomised controlled trials are typically infeasible or unethical in this context, causal investigations based on observational data—such as the one carried out in this article—will, therefore, be crucial in guiding our understanding of the available data. Causal inference, in particular mediation analysis, can be used to resolve apparent statistical paradoxes; help educate the public and decision-makers alike; avoid unsound comparisons; and answer a range of causal questions pertaining to the pandemic, subject to transparently stated assumptions. Our exposition helps clarify how mediation analysis can be used to investigate direct and indirect effects along different causal paths and thus serves as a stepping stone for future studies of other important risk factors for COVID-19 besides age.

The impact of antenatal care on under\u2010five mortality in Ethiopia: a difference\u2010in\u2010differences analysis

BackgroundSub-Saharan Africa, as opposed to other regions, has the highest under-five mortality rates yet makes the least improvement in reducing under-five mortality. Despite the decline, Ethiopia is among the top ten countries contributing the most to global under-five mortalities. This article examines the impact of the number of antenatal care and the timing of first antenatal care on child health outcomes. We specifically investigated if the utilization of antenatal care services positively affects the reduction of under-five mortality.MethodsWe employ a difference-in-differences design with propensity score matching to identify direct causal effects of antenatal care on under-five mortality based on the Ethiopian Demographic Health Survey data of 2011 and 2016. Our sample includes 22 295 women between the ages of 14–49 who had antenatal care visits at different times before delivery.ResultsThe study revealed 1 481 cases of reported under-five mortality. 99.0% of that under-five mortality cases are women who had less than eight antenatal care visits, while only 1% of that is by women who had eight or more antenatal care visits. Antenatal care visit decreases the likelihood of under-five mortality in Ethiopia by 45.2% (CI = 19.2–71.3%, P-value < 0.001) while the timing of first antenatal care within the first trimester decreases the likelihood of under-five mortality by 10% (CI = 5.7–15.6%, P-value < 0.001).ConclusionsTo achieve a significant reduction in the under-five mortality rate, Intervention programs that encourages more antenatal care visits should be considered. This will improve child survival and help in attaining Sustainable Development Goal targets.