Sandwich Estimator Research Articles

The lack of opportunity to eat together is associated with an increased risk of weight loss among independent older adults: a prospective cohort study based on the JAGES.

the present study aimed to investigate the relationship between the frequency of eating together and the risk of weight loss in older adults. this was a three-year follow-up prospective cohort study based on a self-reported questionnaire. We used data from the Japan Gerontological Evaluation Study (JAGES) conducted in 2016 and 2019. The participants were independent older adults aged ≥65years in Japan. We used >5% weight loss during follow-up as the outcome variable and frequency of eating together as the explanatory variable. The relative risks (RRs) and 95% confidence intervals (95% CIs) were estimated based on the Poisson regression model with a Huber-White sandwich estimator for standard errors, including possible confounders. among 56,919 participants, the mean age was 73.0years (1SD = 5.5) at baseline, and 47.9% were male. About 15.1% (n = 8,596) of the participants experienced >5% weight loss during follow-up. The proportion of each category of the eating together frequency was 36.6% for 'every day', 10.3% for 'several times a week', 26.8% for 'several times a month', 20.5% for 'several times a year' and 5.8% for 'seldom'. Compared to 'every day', only 'several times a year (RR = 1.07, 95% CI = 1.01-1.13)' and 'seldom (RR = 1.17, 95% CI = 1.08-1.27)' were significantly associated with the increased risk of >5% weight loss. there is a temporal association between less frequent opportunities to eat together and the increased risk of weight loss among independent older adults.

Adding non-contrast and delayed phases increases the diagnostic performance of arterial CTA for suspected active lower gastrointestinal bleeding.

When assessing for lower gastrointestinal bleed (LGIB) using CTA, many advocate for acquiring non-contrast and delayed phases in addition to an arterial phase to improve diagnostic performance though the potential benefit of this approach has not been fully characterized. We evaluate diagnostic accuracy among radiologists when using single-phase, biphasic, and triphasic CTA in active LGIB detection. A random experimental block design was used where 3 blinded radiologists specialty trained in interventional radiology retrospectively interpreted 96 CTA examinations completed between Oct 2012 and Oct 2017 using (1) arterial only, (2) arterial/non-contrast, and (3) arterial/non-contrast/delayed phase configurations. Confirmed positive and negative LGIB studies were matched, balanced, and randomly ordered. Sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, positive and negative predictive values, and time to identify the presence/absence of active bleeding were examined using generalized estimating equations (GEE) with sandwich estimation assuming a binary distribution to estimate relative benefit of diagnostic performance between phase configurations. Specificity increased with additional contrast phases (arterial 72.2; arterial/non-contrast 86.1; arterial/non-contrast/delayed 95.1; p < 0.001) without changes in sensitivity (arterial 77.1; arterial/non-contrast 70.2; arterial/non-contrast/delayed 73.1; p = 0.11) or mean time required to identify bleeding per study (s, arterial 34.8; arterial/non-contrast 33.1; arterial/non-contrast/delayed 36.0; p = 0.99). Overall agreement among readers (Kappa) similarly increased (arterial 0.47; arterial/non-contrast 0.65; arterial/non-contrast/delayed 0.79). The addition of non-contrast and delayed phases to arterial phase CTA increased specificity and inter-reader agreement for the detection of lower gastrointestinal bleeding without increasing reading times. • A triphasic CTA including non-contrast, arterial, and delayed phase has higher specificity for the detection of lower gastrointestinal bleeding than arterial-phase-only protocols. • Inter-reader agreement increases with additional contrast phases relative to single-phase CTA. • Increasing the number of contrast phases did not increase reading times.

On Variance of the Treatment Effect in the Treated When Estimated by Inverse Probability Weighting.

In the analysis of observational studies, inverse probability weighting (IPW) is commonly used to consistently estimate the average treatment effect (ATE) or the average treatment effect in the treated (ATT). The variance of the IPW ATE estimator is often estimated by assuming that the weights are known and then using the so-called "robust" (Huber-White) sandwich estimator, which results in conservative standard errors (SEs). Here we show that using such an approach when estimating the variance of the IPW ATT estimator does not necessarily result in conservative SE estimates. That is, assuming the weights are known, the robust sandwich estimator may be either conservative or anticonservative. Thus, confidence intervals for the ATT using the robust SE estimate will not be valid, in general. Instead, stacked estimating equations which account for the weight estimation can be used to compute a consistent, closed-form variance estimator for the IPW ATT estimator. The 2 variance estimators are compared via simulation studies and in a data analysis of the association between smoking and gene expression.

Quantifying the effect of probability model misspecification in seismic collapse risk assessment

One of the main steps in probabilistic seismic collapse risk assessment is estimating the fragility function parameters. The maximum likelihood estimation (MLE) approach, which is widely used for this purpose, contains the underlying assumption that the likelihood function is known to follow a specified parametric probability distribution. However, this assumed distribution may not always be consistent with the “true” probability distribution of the collapse data. This paper implements the Information matrix equivalence theorem to identify the presence of model misspecification i.e., if the assumed collapse probability distribution is, in fact, the “true” one. In the presence of model misspecification, the fragility parameter estimates continue to be asymptotically normally distributed but the variance–covariance matrix is no longer equal to the inverse of the Fisher’s Information matrix. To increase the robustness of the variance–covariance matrix, the Huber–White sandwich estimator is implemented. Using collapse data from eight woodframe buildings, the effect of model misspecification on fragility parameter estimates and collapse rate is quantified. For the considered building cases, the parameter estimation uncertainty in the collapse risk did not increase when the “sandwich” estimator was used compared to when probability model misspecification was not considered (i.e., using MLE). The proposed framework should be used to further investigate the issue of probability model misspecification as it relates to fragility parameter estimation since only a single construction type (woodframe buildings) and limit state (collapse) was considered in the current study.

Digital Breast Tomosynthesis and Digital Mammography Recall and False-Positive Rates by Time of Day and Reader Experience.

Background Digital breast tomosynthesis (DBT) image interpretation might be more cognitively demanding than interpretation of digital mammography (DM) images. The time of day of interpretation might affect recall and false-positive (FP) rates, especially for DBT. Purpose To determine whether recall and FP rates vary by time of day of interpretation for screening mammography for breast cancer performed with DM and DBT. Materials and Methods This is a retrospective study examining 97 671 screening mammograms interpreted by 18 radiologists between January 2018 and December 2019 at one of 12 community radiology sites. The association between the time of day of interpretation, the type of image interpreted (DM vs DBT), and radiologist experience (≤5 posttraining years vs >5 posttraining years) and the likelihood of a patient being recalled from screening mammography and the likelihood of whether the interpretation was FP or true positive were analyzed. Analyses were conducted using generalized linear mixed modeling with a binary distribution and sandwich estimation where observations were nested by radiologist. Results Screening mammograms interpreted by 18 radiologists were reviewed (40 220 DBTs, 57 451 DMs). Nine radiologists had 5 or fewer posttraining years of experience, and nine had more than 5 posttraining years of experience. The overall recall rates for DM (10.2%) and DBT (9.0%) were different (P = .006); FP rate also differed (9.8% DM, 8.6% DBT; P = .004). For radiologists with 5 or fewer posttraining years of experience, odds of recall increased 11.5% (odds ratio [OR] = 1.12, P = .01) with every hour when using DBT, but this was not found for DM (OR = 1.09, P = .06); DBT and DM were different (OR = 1.12 vs 1.09, P = .02). For radiologists with more than 5 posttraining years of experience, no evidence of increase in recall was observed for DBT (OR = 1.02, P = .27) or DM (OR = 1.0, P = .80), and there was no evidence that these were different (OR = 1.02 vs 1.0, P = .13). Conclusion Patients were more likely to be recalled when their screening digital breast tomosynthesis images were interpreted later in the day by less-experienced radiologists. © RSNA, 2022 Online supplemental material is available for this article.

Escitalopram administration, relearning, and neuroplastic effects: A diffusion tensor imaging study in healthy individuals

BackgroundNeuroplastic processes are influenced by serotonergic agents, which reportedly alter white matter microstructure in humans in conjunction with learning. The goal of this double-blind, placebo-controlled imaging study was to investigate the neuroplastic properties of escitalopram and cognitive training on white matter plasticity during (re)learning as a model for antidepressant treatment and environmental factors. MethodsSeventy-one healthy individuals (age=25.6 ± 5.0, 43 females) underwent three diffusion magnetic resonance imaging scans: at baseline, after 3 weeks of associative learning (emotional/non-emotional content), and after relearning shuffled associations for an additional 3 weeks. During the relearning phase, participants received a daily dose of 10 mg escitalopram or placebo orally. Fractional anisotropy (FA), and mean (MD), axial (AD), and radial diffusivity (RD) were calculated within the FMRIB software library and analyzed using tract-based spatial statistics. ResultsIn a three-way repeated-measures marginal model with sandwich estimator standard errors, we found no significant effects of escitalopram and content on AD, FA, MD, and RD during both learning and relearning periods (pFDR>0.05). When testing for escitalopram or content effects separately, we also demonstrated no significant findings (pFDR>0.05) for any of the diffusion tensor imaging metrics. LimitationsThe intensity of the study interventions might have been too brief to induce detectable white matter changes. DiscussionPrevious studies examining the effects of SSRIs on white matter tracts in humans have yielded inconclusive outcomes. Our results indicate that relearning under escitalopram does not affect the white matter microstructures in healthy individuals when administered for 3 weeks.

Inattention and hyperactivity symptoms in childhood predict physical activity in adolescence

BackgroundPhysical activity has been documented to influence several aspects of physical and mental health. Growing evidence shows that physical activity can improve attention. Less is known about how symptoms of inattention and hyperactivity / impulsivity in childhood are associated with physical activity in adolescence. We aimed to explore this relationship further.MethodsWe used a cohort of 3949 Swedish children (1884 boys and 2065 girls) with data collected at ages 9 (or 12) and 15. We investigated the influence of symptoms of inattention and hyperactivity / impulsivity in childhood – age 9/12 (inattention and hyperactivity/impulsivity separately) on self-rated physical activity at age 15, using multiple logistic regression models. We considered potential confounders such as sex, parental education level, physical activity in childhood and neurodevelopmental comorbidity. A cluster robust sandwich estimator was applied to adjust the standard errors for the nested twin data when computing the regression models.ResultsSymptoms of inattention in childhood (9/12) predicted less physical activity in adolescence (age 15) (OR = 0.83 CI = 0.78–0.89), whereas the opposite was true for hyperactivity/impulsivity (OR = 1.08 CI = 1.02–1.10). These associations still remained when taking possible confounders into account including neurodevelopmental and neurodevelopmental related comorbidity.ConclusionsThese findings support the importance of helping children and adolescents with inattention symptoms to engage in physical activity in suitable settings.

Trajectories of energy intake distribution and subsequent risk of hyperglycemia among Chinese adults: findings from the China Health and Nutrition Survey (1997\u20132018)

AimsFew studies have examined the secular trend of the energy intake distribution, and its effect on future risk of hyperglycemia. This study aims to describe trajectories of energy intake distribution over 12 years and relate them to subsequent risk of hyperglycemia over 9 years of follow-up.MethodsOur study used ten waves of data from the CHNS survey, a population-based longitudinal survey in China, ongoing since 1989. We examined a cohort of adult participants who were free from diabetes but had at least three waves of dietary data from 1997 to 2009. We assessed energy intake using three consecutive 24 h recalls. We used these data to identify trajectory groups of energy intake distribution by multi-trajectory model based on energy intake proportions of breakfast, lunch, and dinner. We followed up participants for hyperglycemia, diabetes, and impaired fasting glucose for 9 years from 2009 to 2018. Outcomes were ascertained with fasting glucose, serum HbA1c, and self-report of diabetes and/or glucose-lowering medication. We estimated relative risk (RR) for hyperglycemia, diabetes, and impaired fasting glucose by identified trajectory groups using multilevel mixed-effects modified Poisson regression with robust (sandwich) estimation of variance. Gender difference was additionally examined.ResultsA total of 4417 participants were included. Four trajectory groups were identified, characterized and labeled by “Energy evenly distributed with steady trend group” (Group 1), “Dinner and lunch energy dominant with relatively steady trend group” (Group 2), “Dinner energy dominant with increasing trend and breakfast energy with declining trend group” (Group 3), and “breakfast and dinner energy dominant with increasing trend group” (Group 4). During 48,091 person-years, 1053 cases of incident hyperglycemia occurred, 537 cases of incident diabetes occurred, and 516 cases of impaired fasting glucose occurred. Compared with Group 1, Group 3 was associated with higher subsequent risk of incident hyperglycemia in 9 years of follow-up (RR = 1.28, 95% CI = 1.02, 1.61). No association was found for incident diabetes and impaired fasting glucose. Among males, Group 3 was associated with higher risk of incident hyperglycemia in 9 years of follow-up (RR = 1.44, 95% CI = 1.07, 1.94). No relationship was found in females.ConclusionsEnergy intake distribution characterized by over 40% of energy intake from dinner with a rising trend over years was associated with higher long-term risk of hyperglycemia in Chinese adults.

Abstract 10137: Functional Brain Imaging Reveals Decreased Cerebellar Volume and Altered Brain Connectivity in Pulmonary Arterial Hypertension

Introduction: Psychiatric and cognitive deficiencies are increasingly appreciated in pulmonary arterial hypertension (PAH). Brain abnormalities potentially underlying these disease manifestations have been ill-defined. Hypothesis: Altered brain structure and functional connectivity correlate with neurocognitive changes in PAH patients. Methods: PAH patients at UPMC were enrolled from 2017 to 2018. Participants underwent T1/T2 structural and resting functional brain MRI (fMRI) in addition to cognitive testing. Case data were compared in a 1:4 ratio to that of age-, gender-, and education-matched non-PAH controls. Linear regression analysis with clustered sandwich estimator and adjustment for multiple hypothesis testing was used for comparisons. Most recent hemodynamic data for PAH patients were reviewed including pulmonary vascular resistance (PVR). Pearson correlations were computed between PVR and brain volumes. Results: 12 PAH patients (median age, 45 years [range 26-61]; 9 women) were enrolled. T1/T2 MRI analysis identified 18 regions with volumetric differences (q&lt;0.05), with notable involvement of limbic (6/18, 33%) and basal ganglia (4/18, 22%) structures. PAH patients showed decreased left cerebellar cortical (45178 ± 1417 vs 53295 ± 583 mm 3 ; q=0.01) and right cerebellar white matter volumes (16458 ± 727 vs 39675 ± 5846 mm 3 ; q=0.018). Structural changes concorded with results from fMRI analysis: PAH patients had decreased functional connectivity in the left cerebellum (q=0.003). Corresponding with modulation of executive functions by the cerebellum, a trend was seen toward decreased inhibitory control and attention (83 ± 3 vs 95 ± 1; q=0.07) and pattern comparison processing speed (93 ± 5 vs 106 ± 4; q=0.07) in PAH patients. A trend toward negative correlation of PVR with left cerebellar white matter volume (r = -0.82, q = 0.06) was also observed. Conclusion: Limbic, basal ganglia, and cerebellar regions were smaller and less functionally connected in PAH cases than in matched controls. Decreased cerebellar volume may correlate with more severe hemodynamic disease. Any links among brain changes, cognitive dysfunction, and PAH pathophysiology deserve further study and may lead to neurologic modalities to diagnose and treat disease.

Thromboembolism, Bleeding, and Vascular Death Among Older and Younger Nonvalvular Atrial Fibrillation Patients with Type 2 Diabetes Receiving Rivaroxaban or Warfarin

Background: Advancing age and diabetes are common comorbidities among patients diagnosed with nonvalvular atrial fibrillation (NVAF). In separate subanalyses of the Rivaroxaban Once Daily, Oral, Direct Factor Xa Inhibition Compared With Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET AF), the relative efficacy and safety of rivaroxaban versus warfarin did not vary with by older (≥75 years old) or younger age (<75 years old) or with the presence or absence of diabetes. We sought to evaluate the impact of age on the comparative effectiveness and safety of rivaroxaban compared to warfarin in NVAF patients with comorbid diabetes treated in routine practiceMethods: We performed an analysis using Optum® De-Identified electronic health record (EHR) data from November 2010 to December 2019. This study included adults with NVAF and type 2 diabetes, newly started on either rivaroxaban or warfarin and who had ≥12-months of EHR activity. Patients who were pregnant, had alternative indications for oral anticoagulation or valvular heart disease were excluded. To allow for the assessment of statistical interaction in outcome rates across age groups, we propensity score-overlap weighted patients after stratification of eligible patients into older (≥80 years old) and younger (<80 years old) cohorts. Weighted incidence rates (%/year) of developing stroke or systemic embolism, or vascular death; hospitalization for major or clinically relevant nonmajor (CRNM) bleeding; and major adverse limb events (need for revascularization or major amputation of the lower limbs) for rivaroxaban and warfarin users with type 2 diabetes patients either ≥80 or <80 years old were reported. Propensity score-overlap weighted hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated using Cox regression employing a robust sandwich estimator. Patients were followed in the Cox regression models until time of outcome occurrence, end-of-EHR activity or end-of-data availability. P-values for interaction across age subgroups were adjusted to control for false discovery rates due to multiple hypothesis testing. A p-value <0.05 was considered statistically significant in all cases.Results: A total of 32,078 rivaroxaban (31% initiated on 15 mg dose) and 83,971 warfarin users (time-in-therapeutic range = 47±28%) with NVAF and comorbid type 2 diabetes were included in the present study. Mean follow-up time was 2.9±1.9 years for rivaroxaban and 2.9±2.0 years for warfarin patients. There were 31,941 patients (28%) aged ≥80 years who were initiated on either rivaroxaban (n=6,606) or warfarin (n=25,335). Older patients had a higher mean CHA 2DS 2VASc (4.4±1.2 versus 3.8±1.3) and modified HASBLED (1.7±0.7 versus 1.4±0.8) score compared to younger patients. Propensity score-overlap weighted analyses found no statistically significant interaction for the relative effectiveness or safety of rivaroxaban versus warfarin across the older or younger age groups for the outcomes of stroke, systemic embolism, or vascular death (HR=0.93; 95%CI, 0.87-1.00 versus HR=0.91; 95%CI, 0.86-0.96), hospitalization for major or CRNM bleeding (HR=1.06; 95%CI, 0.96-1.18 versus HR=0.90; 95%CI, 0.84-0.96) or major adverse limb events (HR=0.79; 95%CI, 0.68-0.94 versus HR=0.76; 95%CI, 0.70-0.83) (Table). Major and intracranial bleeding were also observed significantly less frequently with rivaroxaban compared to warfarin regardless of patient age.Conclusions: The effectiveness and safety of rivaroxaban relative to warfarin remained consistent across older and younger patient subgroups, supporting rivaroxaban as an alternative for elderly NVAF patients with concomitant type 2 diabetes (NCT04509193). [Display omitted] DisclosuresColeman: Alexion Pharmaceuticals: Research Funding; Janssen Scientific Affairs LLC: Consultancy, Honoraria, Research Funding; Bayer AG: Consultancy, Honoraria, Research Funding. Vardar: Bayer AG: Current Employment. Abdelgawwad: Bayer AG: Current Employment.

Does Transient Opioid Use Increase Risk of Short-Term Respiratory Exacerbation among Older Adults with Chronic Obstructive Pulmonary Disease?

The objective of this study was to examine the association between transient opioid use and acute respiratory exacerbations among older Medicare beneficiaries with COPD. This study was conducted using national Medicare 5% sample administrative claims data between 2012 and 2016 and employed a case-crossover design. The date of eligible COPD exacerbation events was defined as the index date and the presence of opioid prescriptions during a 7-day exposure window prior to index date was compared to a set of 10 control periods, each 7-days long. The association between opioid exposure and COPD exacerbation was estimated using a conditional logistic regression with robust sandwich estimators, after accounting for known time-varying confounders. Among 16,290 eligible COPD exacerbations included in the study sample, the average patient age was 77.08 years, and 64.2% of events occurred in women. Transient exposure to opioids was associated with a 76% increase in the odds of an acute COPD exacerbation (OR: 1.76, 95%CI: 1.67–1.84), and each 25 mg increase in morphine milligram equivalent dose was associated with a 18% increase in the odds of exacerbation (OR: 1.18, 95% CI: 1.15–1.21). Effect estimates were consistent across subgroup analyses conducted among events identified in the emergency department versus hospital, and among individuals with a single exacerbation event versus those with multiple exacerbations. Transient exposure to opioids was associated with an increased short-term risk of respiratory exacerbation among older adults with COPD. Treatment decisions for breathlessness among individuals with COPD need to account for the benefit-risk profile of opioids. Supplemental data for this article is available online at https://doi.org/10.1080/15412555.2021.2013460 .

Environmental Displacement and Mental Well-Being in Banjarnegara, Indonesia.

Background:Residential moves (displacement) owing to climate- and weather-related disasters may significantly impact mental health. Despite the growing risk from climate change, health impacts of environmental-mobility remain understudied.Objectives:We assessed the effects of displacement on the association between landslides and changes in perceived mental well-being in Banjarnegara, Indonesia. We also investigated whether sociodemographics (age, sex, level of education, household-level income, or employment in agriculture) and landslide characteristics (number and severity of landslides) were associated with differing odds of relocation after experiencing landslides.Methods:In this cross-sectional study, we surveyed 420 individuals who experienced landslides between 2014 and 2018 to assess perceived changes in mental well-being, comparing after landslide exposure to before landslide exposure. We used a novel six-item measure that was created in collaboration with the local community to compare perceived changes between those who were displaced by landslides and those who were not displaced, using logistic and multinomial regressions adjusted for sociodemographic characteristics. We then assessed whether the odds of displacement differed based on sociodemographic characteristics and landslide exposure characteristics, using logistic regressions.Results:Those who were displaced were more likely than those who were not displaced to report perceived increases in economic stability [; 95% confidence interval (CI): 1.45, 6.46], optimism (; 95% CI: 1.87, 8.61), safety (; 95% CI: 1.44, 5.10), religiosity (; 95% CI: 1.03, 3.65), and closeness with community (; 95% CI: 1.10, 3.33) after landslides compared with before their first landslide during the study period. More frequent landslide exposures were associated with reduced odds of relocation, but more severe landslides were associated with increased odds of relocation.Discussion:These findings suggest that landslides affect the mental well-being not only of those who are displaced but also of those who are left behind. Further, this work supports the need for community-based participatory research to fully capture the health impacts of environmental mobility. https://doi.org/10.1289/EHP9391

Epistemic uncertainty quantification in deep learning classification by the Delta method

The Delta method is a classical procedure for quantifying epistemic uncertainty in statistical models, but its direct application to deep neural networks is prevented by the large number of parameters P. We propose a low cost approximation of the Delta method applicable to L2-regularized deep neural networks based on the top K eigenpairs of the Fisher information matrix. We address efficient computation of full-rank approximate eigendecompositions in terms of the exact inverse Hessian, the inverse outer-products of gradients approximation and the so-called Sandwich estimator. Moreover, we provide bounds on the approximation error for the uncertainty of the predictive class probabilities. We show that when the smallest computed eigenvalue of the Fisher information matrix is near the L2-regularization rate, the approximation error will be close to zero even when K≪P. A demonstration of the methodology is presented using a TensorFlow implementation, and we show that meaningful rankings of images based on predictive uncertainty can be obtained for two LeNet and ResNet-based neural networks using the MNIST and CIFAR-10 datasets. Further, we observe that false positives have on average a higher predictive epistemic uncertainty than true positives. This suggests that there is supplementing information in the uncertainty measure not captured by the classification alone.

Physical activity and active recreation before and during COVID-19: The Our Life at Home study

ObjectivesTo examine the effects of COVID-19 related ‘lockdown restrictions’ on Australian's (5–75 years) physical activity recommendation achievement and active recreation participation. DesignCross-sectional online survey with self and proxy-report items (where the participant was a parent). MethodsAdults (n = 1360) and adolescents (n = 1292) reported the frequency they performed 30- or 60-min of moderate-to-vigorous physical activity (MVPA), muscle-strengthening exercises, and participation in 11 active recreation behaviours in February 2020 (pre-COVID-19) and in April/May (during lockdown restrictions). Parents also proxy-reported activity for their child (n = 147, 5–12 years). Mixed effects logistic regressions or a logistic regression (with robust sandwich estimation for variance) assessed recall differences pre- and during lockdown, and interaction by sex. ResultsCompared to February, in April/May children were less likely to meet MVPA recommendations (OR = 0.27, 95%CI = 0.12–0.64); adolescents males, but not females, were less likely to meet MVPA (OR = 0.71, 95%CI 0.43, 1.17) and both recommendations (OR = 0.12, 95%CI = 0.02, 0.79); and adults were more likely to meet MVPA (OR = 1.26, 95%CI = 1.01, 1.57) but less likely to meet muscle-strengthening exercise recommendations (OR = 0.76, 9%CI = 0.65, 0.89). Across age groups more participants reported walking, muscle strengthening exercises at home, and yoga/Pilates/stretching at home, and fewer performed informal sport practice and play, and recreational activities. ConclusionsLockdown restrictions had different effects on physical activity and active recreation among age groups and by sex. Physical activity promotion strategies that target children and adolescents, at home physical activity options, active neighbourhoods, and (re)engagement in informal sport and recreational activities post-COVID-19 are critical for (re)engaging Australians in health-enhancing behaviours.

Abnormal hb-levels associate differently with type 1 and type 2 myocardial infarction in patients visiting the emergency department with chest pain

Abstract Background Previous studies have found an association between myocardial infarction (MI) and abnormal haemoglobin (Hb) but it is unknown whether Hb-levels are associated with type 1 and type 2 MI in unselected patients with chest pain visiting the emergency department (ED). Purpose To investigate the association between abnormal Hb-levels and type 1 and type 2 MI in patientens visiting the ED with chest pain. Methods The study population comprised all consecutive patients visiting four urban ED:s for chest pain between 2013–2016 with available data on Hb. Clinical data from the ED visit were cross-referenced to compulsory national registries retrieving information on previous diagnoses and treatments to identify history of cardiovascular disease (defined as previous MI, stroke or peripheral vascular disease), hypertension, hyperlipidaemia and diabetes mellitus. Patients were categorized depending on the exposure (Hb-level) to anaemia, normal and polycythaemia. The primary outcome (type 1 and type 2 MI) was identified in the Swedish Web-System for Enhancement and Development of Evidence-Based Care in Heart Disease Evaluated According to Recommended Therapies (SWEDEHEART). Relative risk ratio (RRR) was calculated using multinomial logistic regression, with 95% confidence interval (CI) using no infarction as reference. The robust sandwich estimator was used to estimate standard errors. Adjustments were made for risk factors according to HEART-score. Results A total of 64 606 patient with chest pain were included with a mean (SD) age of 56 (19) years and 48% were women. Anaemia was present in 10 204 (15.8%) and polycythaemia in 1598 (2.5%). Overall, type 1 MI occurred in 2 296 patients and type 2 MI in 145. The risk for type 1 MI was higher in both anaemia (RRR 1.8, 95% CI 1.6–2.0) and polycythaemia (RRR 1.6, 95% CI 1.3–2.0) compared to normal Hb. For type 2 MI the risk was markedly higher for both low and high Hb compared to normal (RRR 4.0, 95% CI 2.8–5.6) and (RRR 3.0, 95% CI 1.4–6.9). Taking age, gender and risk factors into account, patients with anaemia had a lower risk (RRR 0.8, 95% CI 0.7–0.9) for type 1 MI compared to patients with normal Hb whereas patients with polycythaemia still had a higher risk (RRR 1.6, CI 1.2–2.0). For type 2 MI, the risk remained higher for both low (RRR 1.9, 95% CI 1.3–2.8) and high Hb (RRR 2.8, 95% CI 1.3–6.2) compared to normal. Conclusion Abnormal Hb-levels in chest pain patients in the ED were significantly associated with an increased risk of type 1 or type 2 MI, however when accounting for risk factors, in a differential pattern. These novel findings indicate that Hb-level may be important when assessing patients for MI symptoms in the ED, however, further investigations are needed to establish the definite predictive value. Funding Acknowledgement Type of funding sources: None. Adjusted results type 1 and type 2 MI

Reduced midbrain functional connectivity and recovery in abstinent heroin users

Dopaminergic pathways from the midbrain to striatum as well as cortex are involved in addiction. However, the alternations of these pathways and whether the recoveries of aberrant circuits would be detected after prolonged abstinence in heroin users are rarely known. The resting-state functional connectivity (RSFC) patterns of midbrain (i.e., the ventral tegmental area (VTA) and substantia nigra (SN)) were compared between 40 abstinent heroin users with opioid use disorder (HUs) and 35 healthy controls (HCs). Then, we tested the functional recovery hypothesis by both cross-sectional and longitudinal design. For cross-sectional design, HUs were separated into short-term abstainers (STs) (3–15 days) and long-term abstainers (LTs) (>15 days). With regard to longitudinal design, 22 subjects among HUs were followed up for 10 months. A sandwich estimator method was used to analyze the differences between baseline HUs and follow-up HUs. HUs showed lower RSFC between midbrain and several cortical areas (medial orbitofrontal cortex (mOFC) and anterior cingulate cortex) compared with HCs. Besides, lower RSFC of VTA-right nucleus accumbens circuit as well as right SN- caudate circuit was also found in HUs. The enhanced RSFC value of VTA-left mOFC circuit was observed in LTs, compared with STs. Additionally, longitudinal design also revealed the increased RSFC values of the midbrain with frontal cortex after 10 months prolonged abstinence. We revealed abnormal functional organizations of midbrain-striato and midbrain-cortical circuits in HUs. More importantly, partially recovery of these dysfunctions can be found after long-term abstinence.

The Influence of Glenoid Retroversion on Posterior Shoulder Instability: A Cadaveric Study (214)

Objectives:Increased glenoid retroversion has been associated with an increased risk of posterior glenohumeral instability. Normal mean glenoid version is between 0-7° of retroversion depending on the population and measurement method. Retroversion can range above 20°, notably in patients with glenoid dysplasia. Increased glenoid retroversion has also been proposed as a risk factor for failure after primary soft tissue repair. Arthroscopic repair is the most common surgical treatment; however, this does not address cases of increased glenoid retroversion. What has not been identified is the degree of glenoid retroversion associated with recurrent instability or failed repair. The goal of our work is to (1) measure how resistance to posterior translation changes as retroversion increases, (2) examine if labral tear results in a greater decrease to resistance at increasing degrees of retroversion, and (3) to determine the degree of retroversion at which labral repair fails to restore the resistance of the intact, neutral version state.Methods:Eight fresh frozen cadaveric shoulder specimens (age 50-64, 4 male) were prepared, maintaining bone and capsulolabral tissue. The scapula and humerus were potted using quick-set polyurethane. CT scans were obtained to establish a scapular 3D coordinate system relative to the potting. Specimens were mounted on a 6 degree of freedom musculoskeletal simulation robotic arm (KUKA KR 6 R700, Augsburg, Germany) and referenced to the coordinate system. The humeral head was centered on the glenoid using a 50N compressive force, and the humerus was translated posterior-inferiorly (30° inferior to the midline) at 1mm/sec in neutral rotation for 10mm. The shoulder was positioned in 30° of abduction and 30° of flexion, based on prior protocol. Custom simVITRO (Cleveland Clinic, Ohio, US) labview-based control software measured peak resistance at 0° of version and then in 5° increments of retroversion until the specimen dislocated, up to 30° of retroversion. Version was adjusted through use of a multiplanar vice. A posterior labral tear was created from the 2 to 6 o’clock position on a left shoulder, and the same testing parameters were performed. Vertical mattress sutures using 4 independent bone tunnels were used to repair the labrum and the same version iterations were tested.Generalized estimating equations were used to compare the peak resistance to translation for each degree of version in the intact, cut and repaired states. The maximum likelihood estimators of the model were adjusted for any model misspecification using classical sandwich estimation. Post hoc pairwise comparisons between conditions were conducted via orthogonal contrasts. The Holm-test was used to calculate adjusted p-values and confidence intervals. Statistical significance was established at the P<0.05 level and all interval estimates were calculated for 95% confidence.Results:The mean peak resistance for the intact labral state decreased significantly for each interval increase in retroversion when the humerus was translated posterior-inferiorly (Figure 1). On average, a 1° increase in retroversion correlated with a 3.5% decrease in resistance to translation. Dislocation with an intact labrum without any posterior force occurred at a mean of 22.7° (range 15-30°) of retroversion. After labral tear, resistance forces to posterior-inferior translation decreased but not significantly from the intact state. However, the percent change of resistance force decreased 41% at 25° of retroversion; this was notably higher than the percent change at 0-15° of retroversion (range 2.7-6.5% decrease) but was not statistically significant (Figure 2). Compared to the intact state at 0° version, there was a 45% and 81% decrease in resistance after labral repair at 20° and 25° of retroversion, respectively (p=0.04 and p=0.004).Conclusions:Glenoid retroversion has a significant effect on resistance to posterior humeral head translation, with each degree increase accounting for 3.5% of resistance to translation. Cutting the labrum at 0-15° of retroversion does not have a significant effect on resistance to posterior inferior humeral translation; however, at 25° of retroversion cutting the labrum results in a 41% decrease in resistance. Similarly, labral repair at 20-25° of retroversion does not recreate peak resistance values of the intact state at 0-5° of retroversion. These findings point to the bony anatomy (retroversion) playing a larger role in preventing posterior instability than the labrum. It also provides evidence that the labrum plays a more significant role in stability at higher degrees of retroversion, and labral repair in patients with >20° of retroversion may be subjected to a relatively greater percentage of force than those at lesser degrees of retroversion.

THE RELATIONSHIP BETWEEN IVF CLINIC VOLUME AND SUCCESS RATES IN SART REPORTING CLINICS

To determine if the volume of IVF cycles performed by a clinic affects the live birth rate Individual clinic reports on the SART.com site were reviewed from the finalized annual report for 2017, and data on IVF cycle volume and success rates were extracted. Clinics reporting more than one location were aggregated into a single report, yielding 341 clinic reports for 248,086 cycles. We explored whether there was a relationship between the clinic size and the outcomes of live births for first embryo transfer, live births for all embryo transfers, singleton births for first embryo transfer, and singleton births for all embryo transfers. For analysis, we fit 4 linear regression models with the birth outcomes for women under 35 in 286 clinics that had greater than 20 cycle starts and had fewer than 2000 total cycles. Due to the non-normality of the data, we performed Box Cox transformations on the percent live birth when necessary. Using clinic volumes as the predictor and a sandwich estimator for the variance to account for decreasing residual variance with increasing clinic volume, as well as an optimal , we found that there were slight but statistically significant positive associations between clinic size and percentage live births for singleton live births for all transfers (=1.3939,) and for singleton births for the first transfer (, but clinic size was not significant for live births for all transfers ( or for live births for the first transfer. Using an illustrative example for the Box Cox transformed outcomes, an increase from 500 to 1500 in clinic volume would correspond a non statistically significant ∼2.25% increase in total live births on average (from 53.72% to 55.97%) and to a statistically significant ∼5.61% increase in singleton live births (from 45.04% to 50.66%). The R2 values are very low for all models (< 0.1), implying low predictive power of clinic size for birth rates. Scatterplots revealed an increased variability of success rates among smaller clinics. Live birth rates from IVF, on average, are not affected by increasing volume of cycles per clinic, but there is greater variability in success rates and a slightly higher multiple pregnancy rate among smaller clinics than larger clinics.

Risk of pain in the neck and shoulders and job change among hairdressers: a combined questionnaire and register-based Danish prospective cohort study

ObjectiveTo investigate whether intensity of work as a hairdresser was associated with treatments for pain, and if musculoskeletal pain was associated with leaving the hairdressing trade.MethodsWe formed two cohorts of hairdressers covered in the PensionDanmark Health Scheme (PDHS). Cohort 1 consisted of 1304 active hairdressers in 2009. Self-reported weekly haircuts were used as work intensity measure and treatments for pain in the neck and shoulders in PDHS were used as outcome. We used a Cox regression model with robust sandwich estimates adjusted for age, sex, employment status and prior treatment < 1 year before July 2009. Cohort 2 consisted of all hairdressers ever covered in the PDHS from 2006 to 2016 (n = 11,162). Exposure were treatments in PDHS within the last year. Outcome was leaving the trade within the following year. Adjustments were made for sex, calendar-year and age in Cox regression models.ResultsThe adjusted hazard ratio of treatments in PDHS compared to the lowest work intensity was 0.95 (95% CI 0.58–1.55) and 0.74 (0.43–1.29) for medium and highest intensity, respectively. The risk of leaving the trade was lower, HR 0.80 (0.72–0.90) among hairdressers with treatments in PDHS within the last year, mainly driven by hairdressers aged < 56 years.ConclusionWe found no association between intensity of work as a hairdresser, measured as self-reported weekly haircuts, and treatments for pain in PDHS. Furthermore, we found a protective effect of treatments in the PDHS within the last year on risk of leaving the trade.

Assessment of a Modified Sandwich Estimator for Generalized Estimating Equations with Application to Opioid Poisoning in MIMIC-IV ICU Patients

Longitudinal data is encountered frequently in many healthcare research areas to include the critical care environment. Repeated measures from the same subject are expected to correlate with each other. Models with binary outcomes are commonly used in this setting. Regression models for correlated binary outcomes are frequently fit using generalized estimating equations (GEE). The Liang and Zeger sandwich estimator is often used in GEE to produce unbiased standard error estimation for regression coefficients in large sample settings, even when the covariance structure is misspecified. The sandwich estimator performs optimally in balanced designs when the number of participants is large with few repeated measurements. The sandwich estimator’s asymptotic properties do not hold in small sample and rare-event settings. Under these conditions, the sandwich estimator underestimates the variances and is biased downwards. Here, the performance of a modified sandwich estimator is compared to the traditional Liang-Zeger estimator and alternative forms proposed by authors Morel, Pan, and Mancl-DeRouen. Each estimator’s performance was assessed with 95% coverage probabilities for the regression coefficients using simulated data under various combinations of sample sizes and outcome prevalence values with independence and autoregressive correlation structures. This research was motivated by investigations involving rare-event outcomes in intensive care unit settings.