scholarly journals Exploring the Risk Factors of Diabetes in Dhaka City: Negative Binomial Regression & Logistic Regression Approach

2020 ◽  
Vol 6 (12) ◽  
pp. 753-758
Author(s):  
  Mohammad Ahsan Uddin
Keyword(s):  
Risk Factors ◽  
Logistic Regression ◽  
Negative Binomial ◽  
Negative Binomial Regression ◽  
Dhaka City ◽  
Binomial Regression ◽  
Regression Approach

Investigation of factors associated with the probability of racehorses being pulled up in steeplechase races at Cheltenham racetrack

2012 ◽  
Vol 8 (2) ◽  
pp. 95-101 ◽  
Author(s):  
F. Mata ◽  
J. Williams ◽  
F. Marks
Keyword(s):  
Risk Factors ◽  
Logistic Regression ◽  
Predictive Models ◽  
Negative Binomial ◽  
Negative Binomial Regression ◽  
Negative Influence ◽  
Factors Associated ◽  
Binomial Regression

Limited research has been conducted to investigate the risk factors associated with horses being pulled up in steeplechase races. The aim of this study was to identify risk factors associated with pulled up horses in steeplechase races at Cheltenham racecourse and utilise these to propose preventative strategies to reduce prospective risks of racehorses being pulled up in steeplechase races. Horse and racetrack factors that could be associated with an increased chance of horses being pulled up, extrapolated from previous research into racehorse falls and clinical injury, were identified and collated via the Racingpost website for all steeplechases (n=1,032) at Cheltenham for a 21 year period (January 1990 - December 2010). A logistic regression was used to model the probability of existence of pulled up horses in a given race. A negative binomial regression was used to model the number of pulled up horses in a given race. Increasing numbers of runners (P<0.001) starting a race and races of longer distances (P<0.001) resulted in more pulled up horses within the race. In contrast, faster race speeds (P<0.01) were associated with the presence of less pulled up horses in a race. Each additional m/s in the speed of the horses running the race in race results in a decreased probability of 38.1% that the race will contain pulled up horses. The influence of other horses within steeplechase races at Cheltenham appears to effect speed within racing and can exert a positive or negative influence on how many horses are pulled up in a race. It is suggested that additional co-variant factors such as going and distance can also impact upon speed, and that it is the interaction of these variables that produce equine fatigue resulting in pulled up horses. The predictive models devised have the potential to be employed to assess risk of horses being pulled up for other racetracks.

scholarly journals Loneliness, physical activity and mental health during Covid-19: a longitudinal analysis of depression and anxiety between 2015 and 2020

2020 ◽  
Author(s):  
Byron Creese ◽  
Zunera Khan ◽  
William Henley ◽  
Siobhan O’Dwyer ◽  
Anne Corbett ◽  
...  
Keyword(s):  
Mental Health ◽  
Physical Activity ◽  
Risk Factors ◽  
Negative Binomial ◽  
Negative Binomial Regression ◽  
Annual Data ◽  
Activity Levels ◽  
Binomial Regression ◽  
Mental Health Questionnaire ◽  
The Impact

BackgroundLoneliness and physical activity are important targets for research into the impact of COVID-19 because they have established links with mental health, could be exacerbated by social distancing policies and are potentially modifiable.MethodWe analysed mental health data collected during COVID-19 from adults aged 50 and over alongside comparable annual data collected between 2015 and 2019 from the same sample. Trajectories of depression (PHQ-9) and anxiety (GAD-7) were analysed with respect to loneliness, physical activity levels and a number of socioeconomic and demographic characteristics using zero-inflated negative binomial regression.Results3,281 people completed the COVID-19 mental health questionnaire, all had at least one data point prior to 2020. In 2020, the adjusted PHQ-9 score for loneliness was 3.2. (95% CI: 3.0-3.4), an increase of one point on previous years and 2 points higher than people not rated lonely, whose score did not change in 2020 (1.2, 95% CI: 1.1-1.3). PHQ-9 was 2.6 (95% CI: 2.4-2.8) in people with decreased physical activity, an increase of 0.5 on previous years. In contrast, PHQ-9 in 2020 for people whose physical activity had not decreased was 1.7 (95% CI: 1.6-1.8), similar to previous years. A similar relationship was observed for GAD-7 though the differences were smaller and the absolute burden of symptoms lower.ConclusionsAfter accounting for pre-COVID-19 trends, we show that experiencing loneliness and decreased physical activity are risk factors for worsening mental health during the pandemic. Our findings highlight the need to examine policies which target these potentially modifiable risk factors.

scholarly journals Injustices in pandemic vulnerability: A spatial-statistical analysis of the CDC Social Vulnerability Index and COVID-19 outcomes in the U.S.

2021 ◽  
Author(s):  
Jamie Song ◽  
Douglas Wiebe ◽  
Sara Solomon ◽  
Eugenia South
Keyword(s):  
Logistic Regression ◽  
Social Vulnerability ◽  
Negative Binomial ◽  
Hot Spot ◽  
Negative Binomial Regression ◽  
Vulnerability Index ◽  
Social Vulnerability Index ◽  
Binomial Regression ◽  
Average Annual Temperature ◽  
The U.S

Background: The COVID-19 pandemic has exacerbated health injustices in the U.S. driven by racism and other forms of structural violence. Research has shown the disproportionate impacts of COVID-19 morbidity and mortality in the most marginalized communities. Objectives: We examined the associations between COVID-19 cumulative incidence (CI) and case-fatality risk (CFR) and the CDC's Social Vulnerability Index (SVI), a composite score assessing historical marginalization and thus vulnerability to disaster events. Methods: Using county-level data from national databases, we used population density, Gini index, percent uninsured, and average annual temperature as covariates, and employed negative binomial regression to evaluate relationships between SVI and COVID-19 outcomes. Optimized hot spot analysis identified hot spots of COVID-19 CI and CFR, which were compared in terms of SVI using logistic regression. Results: As of 2/3/21, 26,452,031 cases of and 448,786 deaths from COVID-19 had been reported in the U.S. Negative binomial regression showed that counties in the top SVI quintile reported 13.7% higher CI (p<0.001) than those in the bottom SVI quintile. Additionally, each unit increase in a county's SVI score was associated with a 0.2% increase in CFR (p<0.001). Logistic regression analysis showed that counties in the lowest SVI quintile had significantly greater odds of being in a CI hot spot than all other counties, yet counties in the highest SVI quintile had 63% greater odds (p=0.008) of being in a CFR hot spot than counties in the lowest SVI quintile. Conclusion: We demonstrated a significant relationship between SVI and CFR, but the relationship between SVI and CI is complex and warrants further investigation. SVI may help elucidate unequal impacts of COVID-19 and guide prioritization of vaccines to communities most impacted by structural injustices.

Sign in / Sign up

Export Citation Format

Share Document