scholarly journals Comparisons Between Hypothesis- and Data-Driven Approaches for Multimorbidity Frailty Index: A Machine Learning Approach (Preprint)

2019 ◽  
Author(s):  
Li-Ning Peng ◽  
Fei-Yuan Hsiao ◽  
Wei-Ju Lee ◽  
Shih-Tsung Huang ◽  
Liang-Kung Chen
Keyword(s):  
Machine Learning ◽  
Dose Response ◽  
Frailty Index ◽  
Risk Groups ◽  
Adverse Outcomes ◽  
Response Relationship ◽  
Dose Response Relationship ◽  
Machine Learning Approach ◽  
Unplanned Hospitalizations

BACKGROUND Using big data and the theory of cumulative deficits to develop the multimorbidity frailty index (mFI) has become a widely accepted approach in public health and health care services. However, constructing the mFI using the most critical determinants and stratifying different risk groups with dose-response relationships remain major challenges in clinical practice. OBJECTIVE This study aimed to develop the mFI by using machine learning methods that select variables based on the optimal fitness of the model. In addition, we aimed to further establish 4 entities of risk using a machine learning approach that would achieve the best distinction between groups and demonstrate the dose-response relationship. METHODS In this study, we used Taiwan’s National Health Insurance Research Database to develop a machine learning multimorbidity frailty index (ML-mFI) using the theory of cumulative diseases/deficits of an individual older person. Compared to the conventional mFI, in which the selection of diseases/deficits is based on expert opinion, we adopted the random forest method to select the most influential diseases/deficits that predict adverse outcomes for older people. To ensure that the survival curves showed a dose-response relationship with overlap during the follow-up, we developed the distance index and coverage index, which can be used at any time point to classify the ML-mFI of all subjects into the categories of fit, mild frailty, moderate frailty, and severe frailty. Survival analysis was conducted to evaluate the ability of the ML-mFI to predict adverse outcomes, such as unplanned hospitalizations, intensive care unit (ICU) admissions, and mortality. RESULTS The final ML-mFI model contained 38 diseases/deficits. Compared with conventional mFI, both indices had similar distribution patterns by age and sex; however, among people aged 65 to 69 years, the mean mFI and ML-mFI were 0.037 (SD 0.048) and 0.0070 (SD 0.0254), respectively. The difference may result from discrepancies in the diseases/deficits selected in the mFI and the ML-mFI. A total of 86,133 subjects aged 65 to 100 years were included in this study and were categorized into 4 groups according to the ML-mFI. Both the Kaplan-Meier survival curves and Cox models showed that the ML-mFI significantly predicted all outcomes of interest, including all-cause mortality, unplanned hospitalizations, and all-cause ICU admissions at 1, 5, and 8 years of follow-up (<i>P</i>&lt;.01). In particular, a dose-response relationship was revealed between the 4 ML-mFI groups and adverse outcomes. CONCLUSIONS The ML-mFI consists of 38 diseases/deficits that can successfully stratify risk groups associated with all-cause mortality, unplanned hospitalizations, and all-cause ICU admissions in older people, which indicates that precise, patient-centered medical care can be a reality in an aging society.

scholarly journals Comparisons Between Hypothesis- and Data-Driven Approaches for Multimorbidity Frailty Index: A Machine Learning Approach

10.2196/16213 ◽  
2020 ◽  
Vol 22 (6) ◽  
pp. e16213
Author(s):  
Li-Ning Peng ◽  
Fei-Yuan Hsiao ◽  
Wei-Ju Lee ◽  
Shih-Tsung Huang ◽  
Liang-Kung Chen
Keyword(s):  
Machine Learning ◽  
Dose Response ◽  
Frailty Index ◽  
Risk Groups ◽  
Adverse Outcomes ◽  
Response Relationship ◽  
Dose Response Relationship ◽  
Machine Learning Approach ◽  
Unplanned Hospitalizations

Background Using big data and the theory of cumulative deficits to develop the multimorbidity frailty index (mFI) has become a widely accepted approach in public health and health care services. However, constructing the mFI using the most critical determinants and stratifying different risk groups with dose-response relationships remain major challenges in clinical practice. Objective This study aimed to develop the mFI by using machine learning methods that select variables based on the optimal fitness of the model. In addition, we aimed to further establish 4 entities of risk using a machine learning approach that would achieve the best distinction between groups and demonstrate the dose-response relationship. Methods In this study, we used Taiwan’s National Health Insurance Research Database to develop a machine learning multimorbidity frailty index (ML-mFI) using the theory of cumulative diseases/deficits of an individual older person. Compared to the conventional mFI, in which the selection of diseases/deficits is based on expert opinion, we adopted the random forest method to select the most influential diseases/deficits that predict adverse outcomes for older people. To ensure that the survival curves showed a dose-response relationship with overlap during the follow-up, we developed the distance index and coverage index, which can be used at any time point to classify the ML-mFI of all subjects into the categories of fit, mild frailty, moderate frailty, and severe frailty. Survival analysis was conducted to evaluate the ability of the ML-mFI to predict adverse outcomes, such as unplanned hospitalizations, intensive care unit (ICU) admissions, and mortality. Results The final ML-mFI model contained 38 diseases/deficits. Compared with conventional mFI, both indices had similar distribution patterns by age and sex; however, among people aged 65 to 69 years, the mean mFI and ML-mFI were 0.037 (SD 0.048) and 0.0070 (SD 0.0254), respectively. The difference may result from discrepancies in the diseases/deficits selected in the mFI and the ML-mFI. A total of 86,133 subjects aged 65 to 100 years were included in this study and were categorized into 4 groups according to the ML-mFI. Both the Kaplan-Meier survival curves and Cox models showed that the ML-mFI significantly predicted all outcomes of interest, including all-cause mortality, unplanned hospitalizations, and all-cause ICU admissions at 1, 5, and 8 years of follow-up (P<.01). In particular, a dose-response relationship was revealed between the 4 ML-mFI groups and adverse outcomes. Conclusions The ML-mFI consists of 38 diseases/deficits that can successfully stratify risk groups associated with all-cause mortality, unplanned hospitalizations, and all-cause ICU admissions in older people, which indicates that precise, patient-centered medical care can be a reality in an aging society.

The unborn smoker: association between smoking during pregnancy and adverse perinatal outcomes

2015 ◽  
Vol 43 (5) ◽  
Author(s):  
Elad Mei-Dan ◽  
Asnat Walfisch ◽  
Boaz Weisz ◽  
Mordechai Hallak ◽  
Richard Brown ◽  
...  
Keyword(s):  
Preterm Delivery ◽  
Dose Response ◽  
Reference Group ◽  
Adverse Outcomes ◽  
Delivery Rate ◽  
Response Relationship ◽  
Dose Response Relationship ◽  
Fetal Demise ◽  
Smoking During Pregnancy ◽  
Intrauterine Fetal Demise

AbstractTo evaluate a possible dose-response relationship between active maternal smoking during pregnancy and adverse perinatal outcome.Retrospective cohort study.Population-based in Montreal, Quebec, Canada.Women who gave birth to a liveborn or stillborn infant during the period of January 2001 to December 2007.Active smokers of different daily cigarette consumption (n=1646) were identified through maternal self-reporting. The reference group comprised 19,292 non-smoking women who delivered during the same period.Birth weight, preterm delivery rate, fetal and neonatal mortality and morbidity, and congenital malformations.Preterm delivery rate was significantly higher in the smoking group compared with controls (22.2% vs. 12.4%, P<0.05), as was intrauterine fetal demise (1.4% vs. 0.3%, P<0.05). Newborns of active smokers were more likely to weigh less (3150±759 g vs. 3377±604 g, P<0.05), suffer from respiratory distress syndrome (2.5% vs. 1.3%, P<0.05), suffer from a cardiac malformation (1.5% vs. 0.8%, P<0.05), and die (neonatal death 1.2% vs. 0.6%, P<0.05). A dose-response relationship was demonstrated between levels of daily cigarette smoking and several adverse outcomes. Using multiple regression models, smoking was found to be an independent predictor of preterm delivery (odds ratios (OR) 1.9, 95% confidence intervals (95%CI) 1.6–2), and intrauterine fetal demise (OR 2.4, 95%CI 1.4–4.2).Any amount of daily smoking appears to harm the fetus and newborn. As pregnancy may be a “window of opportunity” for behavioural changes, efforts to promote smoking cessation should be encouraged.

scholarly journals Serum Uric Acid Might Be Positively Associated With Hypertension in Chinese Adults: An Analysis of the China Health and Nutrition Survey

2022 ◽  
Vol 8 ◽  
Author(s):  
Yingdong Han ◽  
Kaidi Han ◽  
Xinxin Han ◽  
Yue Yin ◽  
Hong Di ◽  
...  
Keyword(s):  
Uric Acid ◽  
Dose Response ◽  
Response Relationship ◽  
Chinese Adults ◽  
Dose Response Relationship ◽  
Nutrition Survey ◽  
Health And Nutrition ◽  
Increased Risk ◽  
Stratified Analysis

Background: Previous studies have clarified the relationship between serum uric acid (SUA) and hypertension; most of previous studies suggest that elevated uric acid levels are associated with an increased risk of hypertension, while in China, there are relatively few studies to explore above association. The objective of this longitudinal study is to investigate the correlation of SUA and hypertension in Chinese adults with a nationwide large-scale sample.Methods: Data from the China Health and Nutrition Survey 2009, 2011, and 2016 were used; a total of 8,469 participants (3,973 men and 4,496 women) were involved. This study was conducted separately by gender. Clinical characteristics of the participants among different uric acid groups are compared. The binary logistic regression analysis was conducted to examine the association between SUA and hypertension. Restricted cubic spline analysis with three knots of the SUA concentration were used to characterize the dose-response relationship. Additionally, we compared the incidence of hypertension in the different baseline uric acid groups during follow-up in 2011 and 2015.Results: After the covariates were fully adjusted, we found that elevated uric acid levels were correlated with increased risk of hypertension in both males (p &lt; 0.01) and females (p &lt; 0.01). With 2-year or 6-year of follow-up, we found participants with higher baseline uric acid levels had a higher incidence of hypertension (p &lt; 0.01). In stratified analysis by obesity, above relationship remained significant in nonobesity population (males: p &lt; 0.05, females: p &lt; 0.01) and became nonsignificant in obesity people. In stratified analysis by age, above positively correlation remained significant in middle-aged men (p &lt; 0.05) and elderly women (p &lt; 0.01). Restricted cubic spline revealed the dose-response relationship between SUA and hypertension; we also found that above relationship was much stronger in females.Conclusion: This study suggests that elevated SUA levels might be positively associated with an increased risk of hypertension in general Chinese adults.

Treated Versus Untreated Mental Health Problems in Adolescents: A Six-year Comparison of Emotional and Behavioral Problem Trajectories

2017 ◽  
Vol 41 (S1) ◽  
pp. s130-s131
Author(s):  
F. Jörg ◽  
D. Raven ◽  
E. Visser ◽  
R. Schoevers ◽  
T. Oldehinkel
Keyword(s):  
Mental Health ◽  
Dose Response ◽  
Real World ◽  
Response Relationship ◽  
Dose Response Relationship ◽  
Problem Behaviour ◽  
The Real ◽  
Clinical Level ◽  
Pre Treatment

IntroductionMultidisciplinary guidelines in adolescent mental health care are based on RCTs, while treatment efficacy can be different from effectiveness seen in ‘the real world’. Studies in the real world conducted so far suggest that treatment has a negligible effect on follow-up symptomatology. However, these studies did not incorporate the pre-treatment trajectory of symptoms nor investigated a dose-response relationship.ObjectivesTo test whether future treatment users and non-users differed in emotional and behavioural problem scores, whether specialist mental health treatment (SMHT) was effective in reducing problem levels while controlling for pre-treatment trajectory, and to seek evidence of a dose-response relationship.MethodsSix-year follow up data were used from the Tracking Adolescents’ Individual Lives Survey (TRAILS). We identified adolescents with a clinical level of problem behaviour on the Child Behaviour Checklist or Youth Self Report and first SMHT between the ages 13 and 16. Adolescents with a clinical level of problem behaviour but without SMHT use served as control group. A psychiatric case register provided data on number of treatment contacts. Using regression analysis, we predicted the effect of treatment on post-treatment problem scores.ResultsTreated adolescents more often had a (severe) diagnosis than untreated adolescents. Pre-treatment trajectories barely differed between treated and untreated adolescents. Treatment predicted an increase in follow-up problem scores, regardless of the number of sessions.ConclusionThe quasi-experimental design calls for modest conclusions. We might however need to take a closer look at real-world service delivery, and invest in developing treatments that can achieve sustainable benefits.Disclosure of interestThe authors have not supplied their declaration of competing interest.

scholarly journals Impact of cardiovascular health status on the dose-response relationship between physical activity and incident morbidity and mortality

2020 ◽  
Vol 41 (Supplement_2) ◽  
Author(s):  
E.A Bakker ◽  
E.J Oymans ◽  
M.T.E Hopman ◽  
D.H.J Thijssen ◽  
T.M.H Eijsvogels
Keyword(s):  
Health Status ◽  
The Netherlands ◽  
Dose Response ◽  
Cardiovascular Health ◽  
Adverse Outcomes ◽  
Response Relationship ◽  
Healthy Individuals ◽  
Dose Response Relationship ◽  
Active Lifestyle ◽  
Physically Active

Abstract Background Although the health benefits of a physically active lifestyle are well-known, there is ongoing discussion about the dose-response relationship between physical activity (PA) and incident morbidity/mortality, and whether this association may be affected by cardiovascular health status. Purpose We compared the dose-response relationship of PA, incident major cardiovascular events (MACE) and all-cause mortality between healthy individuals and individuals with cardiovascular disease risk factors (CVRF) or cardiovascular diseases (CVD). Methods This study used data from Lifelines, which is a multi-disciplinary population-based cohort including 167,729 participants from the northern population of the Netherlands. Adults (&gt;18 yrs old) without severe illnesses or limited life expectancy (&lt;5 yrs) were included (N=143,059). PA volumes were presented as Metabolic Equivalent of Task (MET) minutes/week, and divided into quartiles (Q1-Q4). The primary outcome was a composite endpoint of incident MACE (i.e. myocardial infarction, stroke, heart failure, CABG or PCI) and all-cause mortality. Results Age (42±12 yrs) and proportion of male (40%) was lower in healthy individuals compared to individuals with CVRF (54±11 yrs, 45% male) or with CVD (57±13 yrs, 62% male). During a median follow-up of 7 years (IQR 6–9), 2,485 events occurred in healthy individuals (2% of n=112,018), 2,214 in individuals with CVRF (8% of n=27,982) and 1,100 (36% of n=3,059) in those with CVD. Higher PA volumes were associated with a lower risk of adverse outcomes in healthy individuals and in individuals with CVRF (Table 1). In contrast, only the highest PA quartile was associated with a risk reduction for adverse outcomes in individuals with CVD (Table 1). Also, effect-modification was present in the dose-response relationship between PA volumes and health outcomes for CVD (P-interaction&lt;0.05), but not for healthy or CVRF. Conclusions Cardiovascular health status impacts the dose-response relationship between PA volumes and adverse outcomes. These findings indicate that PA recommendations should be adjusted to an individual's health status for achieving maximal health benefits from a physically active lifestyle. Funding Acknowledgement Type of funding source: Foundation. Main funding source(s): The work of T.M.H.E is supported by the Netherlands Heart Foundation. The Lifelines Biobank initiative received a subsidy from the Dutch Ministry of Health, Welfare and Sport, the Dutch Ministry of Economic Affairs, the University Medical Center Groningen [UMCG], University Groningen and the Northern Provinces of the Netherlands.

A SUBCUTANEOUS INJECTION ASSAY FOR ANTI-ANDROGENS IN THE CHICK

1962 ◽  
Vol 41 (2) ◽  
pp. 268-273 ◽  
Author(s):  
Ralph I. Dorfman
Keyword(s):  
Dose Response ◽  
Subcutaneous Injection ◽  
Response Relationship ◽  
Dose Response Relationship

ABSTRACT The stimulating action of testosterone on the chick's comb can be inhibited by the subcutaneous injection of 0.1 mg of norethisterone or Ro 2-7239 (2-acetyl-7-oxo-1,2,3,4,4a,4b,5,6,7,9,10,10a-dodecahydrophenanthrene), 0.5 mg of cortisol or progesterone, and by 4.5 mg of Mer-25 (1-(p-2-diethylaminoethoxyphenyl)-1-phenyl-2-p-methoxyphenyl ethanol). No dose response relationship could be established. Norethisterone was the most active anti-androgen by this test.

Sign in / Sign up

Export Citation Format

Share Document