Case-crossover design and its implementation in R

Trigger factors in patients with a patent foramen ovale-associated stroke: A case-crossover study

BACKGROUND

Patent foramen ovale (PFO) is a congenital anatomical variant which is associated with strokes in young adults. Contrary to vascular risk factors and atherosclerosis, a PFO is present from birth. However, it is completely unknown how an anatomical structure that is already present at birth in a large proportion of the population can convert into a PFO that causes stroke in a few. Recent studies reported a significant association between certain trigger factors and ischemic stroke in young adults. This study aims to investigate these triggers in PFO-associated stroke.

METHODS

The ODYSSEY study, a multicenter prospective cohort study between 2013 and 2021, included patients aged 18-49 years experiencing their first-ever ischemic event. Participants completed a questionnaire about exposure to potential trigger factors. A case-crossover design was used to assess the relative risks (RR) with 95% confidence intervals (95% CI). The primary outcome was the RR of potential trigger factors for PFO-associated stroke.

RESULTS

Overall, 1043 patients completed the questionnaire and had an ischemic stroke, of which 124 patients had a PFO-associated stroke (median age 42.1 years, 45.2% men). For patients with PFO-associated stroke, the RR was 26.0 (95% CI 8.0-128.2) for fever, 24.2 (95% CI 8.5-68.7) for flu-like disease, and 3.31 (95% CI 2.2-5.1) for vigorous exercise.

CONCLUSION

In conclusion, flu-like disease, fever, and vigorous exercise may convert an asymptomatic PFO into a stroke-causing PFO in young adults.

DATA ACCESS STATEMENT

The raw and anonymized data used in this study can be made available to other researchers on request. Written proposals can be addressed to the corresponding author and will be assessed by the ODYSSEY investigators for appropriateness of use, and a data sharing agreement in accordance with Dutch regulations will be put in place before data are shared.

Power outages and pediatric unintentional injury hospitalizations in New York State

Background

In the past decade, electrical power disruptions (outages) have increased in the United States, especially those attributable to weather events. These outages have a range of health impacts but are largely unstudied in children. Here, we investigated the association between outages and unintentional injury hospitalizations, a leading cause of childhood morbidity.

Methods

The study setting was New York State (NYS) from 2017 to 2020. Outage exposure was defined as ≥10%, ≥20%, and ≥50% of customers from a power operating locality without power, ascertained from NYS Department of Public Service records and stratified by rural, urban non-New York City (NYC), and NYC regions. Outcome daily block group-level pediatric injury hospitalization data was from the Statewide Planning and Research Cooperative System (SPARCS). We leveraged a case-crossover study design with logistic conditional regression.

Results

We identified 23,093 unintentional injury hospitalizations in children <18 years with complete block group and exposure data. Most hospitalizations occurred in urban regions (90%), whereas outages were more likely in rural than urban areas. In urban non-NYC regions, outages ≥4 hours were associated with 30% increased odds of all-cause unintentional injury hospitalizations when ≥50% of customers were without power. Analyses by injury subtype revealed increasing point estimates as the proportion of customers exposed increased. These results, however, had wide confidence intervals.

Conclusions

Outage exposure differed significantly across rural, urban non-NYC, and NYC regions across New York. Especially at the highest outage threshold, we observed an increased risk of pediatric unintentional injury hospitalizations.

Associations between outdoor air pollutants and non-viral asthma exacerbations and airway inflammatory responses in children and adolescents living in urban areas in the USA: a retrospective secondary analysis

BACKGROUND

Asthma prevalence and severity have markedly increased with urbanisation, and children in low-income urban centres have among the greatest asthma morbidity. Outdoor air pollution has been associated with adverse respiratory effects in children with asthma. However, the mechanisms by which air pollution exposure exacerbates asthma, and how these mechanisms compare with exacerbations induced by respiratory viruses, are poorly understood. We aimed to investigate the associations between regional air pollutant concentrations, respiratory illnesses, lung function, and upper airway transcriptional signatures in children with asthma, with particular focus on asthma exacerbations occurring in the absence of respiratory virus.

METHODS

We performed a retrospective analysis of data from the MUPPITS1 cohort and validated our findings in the ICATA cohort. The MUPPITS1 cohort recruited 208 children aged 6-17 years living in urban areas across nine US cities with exacerbation-prone asthma between Oct 7, 2015, and Oct 18, 2016, and monitored them during reported respiratory illnesses. The last MUPPITS1 study visit occurred on Jan 6, 2017. The ICATA cohort recruited 419 participants aged 6-20 years with persistent allergic asthma living in urban sites across eight US cities between Oct 23, 2006, and March 25, 2008, and the last study visit occurred on Dec 30, 2009. We included participants from the MUPPITS1 cohort who reported a respiratory illness at some point during the follow-up and participants from the ICATA cohort who had nasal samples collected during respiratory illness or at a scheduled visit. We used air quality index values and air pollutant concentrations for PM2·5, PM10, O3, NO2, SO2, CO, and Pb from the US Environmental Protection Agency spanning the years of both cohorts, and matched values and concentrations to each illness for each participant. We investigated the associations between regional air pollutant concentrations and respiratory illnesses and asthma exacerbations, pulmonary function, and upper airway transcriptional signatures by use of a combination of generalised additive models, case crossover analyses, and generalised linear mixed-effects models.

FINDINGS

Of the 208 participants from the MUPPITS1 cohort and 419 participants from the ICATA cohort, 168 participants in the MUPPITS1 cohort (98 male participants and 70 female participants) and 189 participants in the ICATA cohort (115 male participants and 74 female participants) were included in our analysis. We identified that increased air quality index values, driven predominantly by increased PM2·5 and O3 concentrations, were significantly associated with asthma exacerbations and decreases in pulmonary function that occurred in the absence of a provoking viral infection. Moreover, individual pollutants were significantly associated with altered gene expression in coordinated inflammatory pathways, including PM2·5 with increased epithelial induction of tissue kallikreins, mucus hypersecretion, and barrier functions and O3 with increased type-2 inflammation.

INTERPRETATION

Our findings suggest that air pollution is an important independent risk factor for asthma exacerbations in children living in urban areas and is potentially linked to exacerbations through specific inflammatory pathways in the airway. Further investigation of these potential mechanistic pathways could inform asthma prevention and management approaches.

FUNDING

National Institutes of Health, National Institute of Allergy and Infectious Diseases.

The effect of temperature, humidity, precipitation and cloud coverage on the risk of COVID-19 infection in temperate regions of the USA—A case-crossover study

Background

Observations based on the spread of SARS-CoV-2 early into the COVID-19 pandemic have suggested a reduced burden in tropical regions leading to the assumption of a dichotomy between cold and dry and wet and warm climates.

Objectives

Analyzing more than a whole year of COVID-19 infection data, this study intents to refine the understanding of meteorological variables (temperature, humidity, precipitation and cloud coverage) on COVID-19 transmission in settings that experience distinct seasonal changes.

Methods and findings

A time stratified case-crossover design was adopted with a conditional Poisson model in combination with a distributed lag nonlinear model to assess the short-term impact of mentioned meteorological factors on COVID-19 infections in five US study sites (New York City (NYC); Marion County, Indiana (MCI); Baltimore and Baltimore County, Maryland (BCM); Franklin County, Ohio (FCO); King County, Washington (KCW)). Higher-than-average temperatures were consistently associated with a decreased relative risk (RR) of COVID-19 infection in four study sites. At 20 degrees Celsius COVID-19 infection was associated with a relative risk of 0.35 (95%CI: 0.20–0.60) in NYC, 1.03 (95%CI:0.57–1.84) in MCI, 0.34 (95%CI: 0.20–0.57) in BCM, 0.52 (95%CI: 0.31–0.87) in FCO and 0.21 (95%CI: 0.10–0.44) in KCW. Higher-than-average humidity levels were associated with an increased relative risk of COVID-19 infection in four study sites. Relative to their respective means, at a humidity level of 15 g/kg (specific humidity) the RR was 5.83 (95%CI: 2.05–16.58) in BCM, at a humidity level of 10 g/kg the RR was 3.44 (95%CI: 1.95–6.01) in KCW.

Conclusions

The results of this study suggest opposed effects for higher-than-average temperature and humidity concerning the risk of COVID-19 infection. While a distinct seasonal pattern of COVID-19 has not yet emerged, warm and humid weather should not be generally regarded as a time of reduced risk of COVID-19 infections.

Using Consumer-Wearable Activity Trackers for Risk Prediction of Life-Threatening Heart Arrhythmia in Patients with an Implantable Cardioverter-Defibrillator: An Exploratory Observational Study

Ventricular arrhythmia (VA) is a leading cause of sudden death and health deterioration. Recent advances in predictive analytics and wearable technology for behavior assessment show promise but require further investigation. Yet, previous studies have only assessed other health outcomes and monitored patients for short durations (7−14 days). This study explores how behaviors reported by a consumer wearable can assist VA risk prediction. An exploratory observational study was conducted with participants who had an implantable cardioverter-defibrillator (ICD) and wore a Fitbit Alta HR consumer wearable. Fitbit reported behavioral markers for physical activity (light, fair, vigorous), sleep, and heart rate. A case-crossover analysis using conditional logistic regression assessed the effects of time-adjusted behaviors over 1−8 weeks on VA incidence. Twenty-seven patients (25 males, median age 59 years) were included. Among the participants, ICDs recorded 262 VA events during 8093 days monitored by Fitbit (median follow-up period 960 days). Longer light to fair activity durations and a higher heart rate increased the odds of a VA event (p < 0.001). In contrast, lengthier fair to vigorous activity and sleep durations decreased the odds of a VA event (p < 0.001). Future studies using consumer wearables in a larger population should prioritize these outcomes to further assess VA risk.

Association Between Exposure to Effervescent Paracetamol and Hospitalization for Acute Heart Failure: A Case-Crossover Study

We investigated whether effervescent paracetamol, as an important source of non-dietary sodium and fluid load, is associated with a transient increase in the risk of hospitalization for acute heart failure (AHF). We conducted a unidirectional case-crossover study using data from the 1/97^th representative sample from the French healthcare database. Subjects aged 18 years or more, hospitalized for AHF during the 2014-2016 period were included. Exposure to effervescent paracetamol was compared between a risk period (i.e. 15 days immediately prior to admission for AHF) and three earlier 15-day control periods, to test a possible trigger effect of effervescent paracetamol intake on AHF. Adjusted odds ratios (aOR) were estimated with a conditional logistic regression. We identified 4,301 patients hospitalized for AHF. We found that 5.7% of AHF subjects were exposed to effervescent paracetamol during the risk period, as compared with 4.1% during the control periods (aOR 1.56 [CI_95% : 1.27 - 1.90], p < 0.001). This association was also found in the subgroup of subjects with hypertension (aOR 1.45 [CI_95% : 1.13 - 1.87], p = 0.004, n = 2,648) and in the subgroup of subjects aged 83 years or more (aOR 1.70 [CI_95% : 1.28 - 2.24], p < 0.001, n = 2,238). A similar analysis, considering exposure to non-effervescent paracetamol, did not support the existence of an indication bias likely to explain the association observed for effervescent paracetamol. This study suggests an association between effervescent paracetamol and admission for AHF and should be confirmed with other complementary study designs. This article is protected by copyright. All rights reserved.

Short-term effects of air pollution on respiratory diseases among young children in Wuhan city, China

BACKGROUND

The high risks for childhood respiratory diseases are associated with exposure to ambient air pollution. However, there are few studies that have explored the association between air pollution exposure and respiratory diseases among young children (particularly aged 0-2 years) based on the entire population in a megalopolis.

METHODS

Daily hospital admission records were obtained from 54 municipal hospitals in Wuhan city, China. We included all children (aged 0-2 years) hospitalized with respiratory diseases between January 2017 and December 2018. Individual air pollution exposure assessment was used in Land Use Regression model and inverse distance weighted. Case-crossover design and conditional logistic regression models were adopted to estimate the hospitalization risk associated with air pollutants.

RESULTS

We identified 62,425 hospitalizations due to respiratory diseases, of which 36,295 were pneumonia. Particulate matter with an aerodynamic diameter less than 2.5 μm (PM_2.5) and nitrogen dioxide (NO₂) were significantly associated with respiratory diseases and pneumonia. ORs of pneumonia were 1.0179 (95% CI 1.0097-1.0260) for PM_2.5 and 1.0131 (95% CI 1.0042-1.0220) for NO₂ at lag 0-7 days. Subgroup analysis suggested that NO₂, Ozone (O₃) and sulfur dioxide (SO₂) only showed effects on pneumonia hospitalizations on male patients, but PM_2.5 had effects on patients of both genders. Except O₃, all pollutants were strongly associated with pneumonia in cold season. In addition, children who aged elder months and who were in central urban areas had a higher hospitalization risk.

CONCLUSIONS

Air pollution is associated with higher hospitalization risk for respiratory diseases, especially pneumonia, among young children, and the risk is related to gender, month age, season and residential location.

Trigger Factors for Spontaneous Intracerebral Hemorrhage: A Case-Crossover Study

BACKGROUND AND PURPOSE

Whether certain activities can trigger spontaneous intracerebral hemorrhage (ICH) remains unknown. Insights into factors that trigger vessel rupture resulting in ICH improves knowledge on the pathophysiology of ICH. We assessed potential trigger factors and their risk for ICH onset.

METHODS

We included consecutive patients diagnosed with ICH between July 1, 2013, and December 31, 2019. We interviewed patients on their exposure to 12 potential trigger factors (eg, Valsalva maneuvers) in the (hazard) period soon before onset of ICH and their normal exposure to these trigger factors in the year before the ICH. We used the case-crossover design to calculate relative risks (RR) for potential trigger factors.

RESULTS

We interviewed 149 patients (mean age 64, 66% male) with ICH. Sixty-seven (45%) had a lobar hemorrhage, 60 (40%) had a deep hemorrhage, 19 (13%) had a cerebellar hemorrhage, and 3 (2%) had an intraventricular hemorrhage. For ICH in general, there was an increased risk within an hour after caffeine consumption (RR=2.5 [95% CI=1.8-3.6]), within an hour after coffee consumption alone (RR=4.8 [95% CI=3.3-6.9]), within an hour after lifting >25 kg (RR=6.6 [95% CI=2.2-19.9]), within an hour after minor head trauma (RR=10.1 [95% CI=1.7-60.2]), within an hour after sexual activity (RR=30.4 [95% CI=16.8-55.0]), within an hour after straining for defecation (RR=37.6 [95% CI=22.4-63.4]), and within an hour after vigorous exercise (RR=21.8 [95% CI=12.6-37.8]). Within 24 hours after flu-like disease or fever, the risk for ICH was also increased (RR=50.7 [95% CI=27.1-95.1]). Within an hour after Valsalva maneuvers, the RR for deep ICH was 3.5 (95% CI=1.7-6.9) and for lobar ICH the RR was 2.0 (95% CI=0.9-4.2).

CONCLUSIONS

We identified one infection and several blood pressure related trigger factors for ICH onset, providing new insights into the pathophysiology of vessel rupture resulting in ICH.

Plasmapheresis reduces cytokine and immune cell levels in COVID-19 patients with acute respiratory distress syndrome (ARDS)

BACKGROUND

In December 2019, pneumonia associated with a novel coronavirus (COVID-19) was reported in Wuhan, China. Acute respiratory distress syndrome (ARDS) is the most frequently observed complication in COVID-19 patients with high mortality rates.

OBJECTIVE OF STUDY

To observe the clinical effect of plasmapheresis on excessive inflammatory reaction and immune features in patients with severe COVID-19 at risk of ARDS.

MATERIALS AND METHODS

In this single-center study, we included 15 confirmed cases of COVID-19 at Masih Daneshvari Hospital, in March 2020 in Tehran, Iran. COVID-19 cases were confirmed by RT-PCR and CT imaging according to WHO guidelines. Plasmapheresis was performed to alleviate cytokine-induced ARDS. The improvement in oxygen delivery (PaO2/FiO2), total number of T cells, liver enzymes, acute reaction proteins, TNF-α and IL-6 levels were evaluated.

RESULTS

Inflammatory cytokine levels (TNF-α, IL-6), and acute phase reaction proteins including ferritin and CRP were high before plasmapheresis. After plasmapheresis, the levels of PaO2/FiO2, acute phase reactants, inflammatory mediators, liver enzymes and bilirubin were significantly reduced within a week (p < 0.05). In contrast, although the number of T helper cells decreased immediately after plasmapheresis, they rose to above baseline levels after 1 week. Nine out of fifteen patients on non-invasive positive-pressure ventilation (NIPPV) survived whilst the six patients undergoing invasive mechanical ventilation (IMV) died.

CONCLUSION

Our data suggests that plasmapheresis improves systemic cytokine and immune responses in patients with severe COVID-19 who do not undergo IMV. Further controlled studies are required to explore the efficacy of plasmapheresis treatment in patients with COVID-19.

Effect Modification by Environmental Quality on the Association between Heatwaves and Mortality in Alabama, United States

Background: Previous studies have shown that heatwaves are associated with increased mortality. However, it remains unclear whether the associations between heatwaves and mortality are modified by the environmental quality. Methods: We used the United States (US) Environmental Protection Agency's Environmental Quality Index (EQI) and its five domain indices (air, water, land, built, and sociodemographic) to represent the cumulative environmental quality. We applied a time-stratified case-crossover design to analyze the disparities in the association between heatwaves and non-accidental deaths (NAD) among counties with different environmental qualities, in metropolitan areas in Alabama (AL), United States. Results: We found significant associations between heatwaves and NAD and a significant effect modification of this relationship by EQI. There were higher odds ratios in counties with the worst cumulative environmental qualities compared to counties with the best cumulative environmental qualities. For example, the percent change in odds ratio (mean and (95% CI)) between heatwave days and non-heatwave days was -10.3% (-26.6, 9.6) in counties with an overall EQI of 1 (the best overall environment) and 13.2% (4.9, 22.2) in counties with an overall EQI of 3 (the worst overall environment). Among the five domains, air quality had the strongest effect modification on the association. Conclusion: Our findings provide evidence that the associations between heatwaves and NAD vary among areas with different environmental qualities. These findings suggest that integration of air quality and heatwave warning systems may provide greater protection to public health.