Abstract 40: Empirical Comparison of Approaches to Mediation Analyses in Nutrition Research With Applications to the Quality Cohort

Circulation ◽  
2020 ◽  
Vol 141 (Suppl_1) ◽  
Author(s):  
Karine Suissa ◽  
Andrea Benedetti ◽  
Melanie Henderson ◽  
Katherine Gray-Donald ◽  
Gilles Paradis
Keyword(s):  
Lipid Profile ◽  
Direct Effect ◽  
Fat Mass ◽  
Mediation Analysis ◽  
Blood Lipids ◽  
Glycemic Load ◽  
Small Sample ◽  
Conventional Approach ◽  
Nutrition Research ◽  
Nutritional Studies

Background: Adiposity may mediate the effect of dietary glycemic load (GL) on lipid profile in children as studies have shown an association between dietary GL and adiposity and between adiposity and lipid profiles. While several approaches for conducting mediation analysis have been proposed, given the strict assumptions and restrictions for each method, it is not clear which method is ideal in nutrition research. Our objective was to compare the conventional approach and the causal approach using marginal structural models (MSM) with inverse probability weights (IPW) to examine adiposity as a mediator in the association between baseline dietary GL and lipid profile after 2 years. Methods: The QUALITY cohort included 630 children, 8-10 years at recruitment with at least one obese parent followed-up 2 years later. Three separate 24-hour dietary recalls were administered by a dietitian at baseline. Child and parent characteristics were obtained through direct measurement (blood lipids, anthropometrics) or questionnaires (socio-economic characteristics). Indicators of adiposity, including BMI z-score and percent fat mass, were the mediators of interest. A conventional approach of statistical adjustment for the mediator was used. As well, we used MSM to estimate the controlled direct effect (CDE) between GL and blood lipids 2 years later not mediated by adiposity. The MSM models were constructed using IPW truncated at 5-95 percentile. Results: Mean age of children at baseline was 9.6 years and 33% were overweight or obese. Both mediation analysis methods revealed that most of the effect of GL on blood lipids was mediated by adiposity. Results obtained from both methods differed; while the MSM with IPW did not show evidence of a direct effect of GL on TG not via percent fat mass (TG: β=0.06, 95%CI=-0.01, 0.12; HDL: β=-0.01, 95%CI=-0.04, 0.03), the conventional method did (TG: β=0.03, 95%CI=0.003, 0.06; HDL: β=-0.02, 95%CI=-0.03, -0.001). Conclusion: In conclusion, our results suggest that adiposity contributes substantially to the association between GL and blood lipids after 2 years. MSM with IPW may not be appropriate in nutritional studies with small sample sizes.

AP2-microRNA-26a overexpression reduces visceral fat mass and blood lipids

2021 ◽  
pp. 111217
Author(s):  
Hailuan Zeng ◽  
Weihong Sun ◽  
Xinping Ren ◽  
Nan Xia ◽  
Sheng Zheng ◽  
...  
Keyword(s):  
Fat Mass ◽  
Visceral Fat ◽  
Blood Lipids

scholarly journals AB0875-HPR NON-PHARMACOLOGICAL AND PHARMACOLOGICAL APPROACHES TO THE MANAGEMENT OF PATIENTS WITH OSTEOARTHRITIS AND LIPID METABOLISM DISORDERS

2021 ◽  
Vol 80 (Suppl 1) ◽  
pp. 1461.1-1461
Author(s):  
T. Rogatkina ◽  
O. Korolik ◽  
V. Polyakov ◽  
G. Kravtsov ◽  
Y. Polyakova
Keyword(s):  
Physical Activity ◽  
Cardiovascular Disease ◽  
Lipid Metabolism ◽  
Lipid Profile ◽  
Fat Mass ◽  
Prognostic Significance ◽  
Lipid Lowering ◽  
Lipid Lowering Therapy ◽  
Lipid Metabolism Disorders ◽  
Group I

Background:Attention is drawn to the frequent combination of osteoarthritis (OA) with cardiovascular disease. Non-specific inflammation plays a significant role in the pathogenesis of OA and atherosclerosis. Limiting the physical activity of patients with OA is an additional important factor aggravating the course of cardiovascular disease (CVD). Chronic pain syndrome, causing a neuroendocrine response, is often the cause of the development of complications of atherosclerotic disease. Dyslipidemia is the main cause of atherosclerosis and vascular thrombosis.Objectives:To study variants of lipid metabolism disorders in female and male patients of different age groups with osteoarthritis.Methods:Case histories of 90 patients with OA were analyzed. The average age of patients was 63.27 ± 11.31 years. The average body mass index (BMI) is 39.8 ± 3.2. All patients underwent questionnaires, general clinical and biochemical blood tests with lipid profile determination, anthropometry, bioimpedansometry, and the main metabolic rate assessment using indirect calorimetry in dynamics (at the beginning of the study and after 3 months).Results:Burdened heredity for obesity, arterial hypertension (AH), diabetes mellitus (DM) was revealed. AH was diagnosed in 76 patients (84.4%), type II diabetes in 17 (18.9%), dyslipidemia and hypercholesterolemia in 56 (62.2%). Statins were taken by 43 patients (47.8%) - group I patients, which is associated with low adherence to therapy, group II included patients who did not initially take statins or stopped taking them at least 6 months before inclusion in the study.Against the background of diet therapy and physiotherapy exercises, BMI (R0.99; p <0.05), fat mass (R0.95; p <0.05) significantly decreased, lipid profile normalization was noted: total cholesterol (R0.66; p <0 .05), LDL (R0.69; p <0.05), HDL (R0.95; p <0.05), TG (R0.57; p <0.05), AST decreased (R0.64; p <0.05) and ALT (R0.76; p <0.05) in both groups of patients, regardless of lipid-lowering therapy. A decrease in fat mass correlated with TG levels (R0.51; p <0.05), an increase in skeletal muscle mass (R0.60; p <0.05), lean mass (R0.72; p <0.05), and active cell mass (R0.59; p <0.05). The lipid profile in the I group of patients was significantly better before and at the end of the study. Long-term effects have not been investigated due to the short duration of the study.Conclusion:In patients with OA, a high frequency of concomitant diseases of the cardiovascular system, lipid metabolism disorders was found. Non-drug therapy has a positive effect on the lipid profile and the level of transaminases. The decrease in body weight due to loss of fat mass reliably correlates with the level of TG. Timely use of statins contributes to the normalization of the lipid profile, reduces the risk of cardiovascular disease in patients with OA. It is necessary to study lipid profile disorders in patients with OA with recommendations for lifestyle modification (diet, physical activity), and if necessary, prescribe lipid-correcting therapy.References:[1]E. Simakova, B. Zavodovsky, L. Sivordova [et al]. Prognostic significance of lipid disorders markers determination in pathogenesis of osteoarthritis. Vestnik Rossijskoj voenno-medicinskoj akademii. 2013. No. 2 (42). P.29-32.[2]Zavodovsky B.V., Sivordova L.E. Prognostic significance value of definition of leptin level determination in osteoarthritis. Siberian Medical Journal (Irkutsk). 2012; 115(8):069-072.Disclosure of Interests:None declared

scholarly journals Exact parametric causal mediation analysis for a binary outcome with a binary mediator

2021 ◽  
Author(s):  
Marco Doretti ◽  
Martina Raggi ◽  
Elena Stanghellini
Keyword(s):  
Logistic Regression ◽  
Direct Effect ◽  
Mediation Analysis ◽  
Delta Method ◽  
Binary Outcome ◽  
Effect Decomposition ◽  
Causal Mediation ◽  
Regression Parameters ◽  
Causal Mediation Analysis ◽  
Parametric Expression

AbstractWith reference to causal mediation analysis, a parametric expression for natural direct and indirect effects is derived for the setting of a binary outcome with a binary mediator, both modelled via a logistic regression. The proposed effect decomposition operates on the odds ratio scale and does not require the outcome to be rare. It generalizes the existing ones, allowing for interactions between both the exposure and the mediator and the confounding covariates. The derived parametric formulae are flexible, in that they readily adapt to the two different natural effect decompositions defined in the mediation literature. In parallel with results derived under the rare outcome assumption, they also outline the relationship between the causal effects and the correspondent pathway-specific logistic regression parameters, isolating the controlled direct effect in the natural direct effect expressions. Formulae for standard errors, obtained via the delta method, are also given. An empirical application to data coming from a microfinance experiment performed in Bosnia and Herzegovina is illustrated.

scholarly journals Smartphone Addiction, Narcissism and Aggression: A Mediation Analysis Method

2019 ◽  
Author(s):  
Anjelica Simsek ◽  
Cahit Nuri ◽  
Cemaliye Direktor ◽  
Ahmet Arnavut
Keyword(s):  
Direct Effect ◽  
Indirect Effect ◽  
Mediation Analysis ◽  
Total Effect ◽  
Mediation Effect ◽  
Smartphone Addiction ◽  
Analysis Method ◽  
Variable Effect ◽  
Independent Variable ◽  
Mediator Variable

<p>At this study, meditation effect of aggression was analyzed using Baron and Kenny’s mediation analysis method. Baron and Kenny (1986) indicates that to analyze the effect of mediator variable 3 criteria have to be actualized:</p> <p>1. Independent variable have a significant effect on a mediator variable (way a)</p> <p>2. Mediator variable have a significant effect on a dependent variable (way b)</p> <p>3. Independent variable have a significant effect on a dependent variable (way c)</p> <p>PROCESS program were used the meditational effect, it is an extra macro that is downloading to the Daniel and Hayes’s (2016) SPSS program. In this program mediation effect could be evaluated as; total effect, direct effect and indirect effect scores of mediation variable effect on dependent variable (Preacher & Hayes, 2008).</p>

scholarly journals Obesity, smoking status and their relationships in selected population groups

2021 ◽  
pp. 291-299
Keyword(s):  
Body Composition ◽  
Lipid Profile ◽  
Fat Mass ◽  
Severe Obesity ◽  
Overweight And Obesity ◽  
Fat Free Mass ◽  
Chronic Obstructive ◽  
Average Value ◽  
Copd Patients ◽  
The Mean

Background. Overweight and obesity are defined as abnormal or excessive fat accumulation that may impair health. Obesity is associated with many chronic diseases, including cardiovascular disease and diabetes, and recently the role of overweight and obesity in lung disease has received new interest. Chronic obstructive lung disease is the third-leading cause of death globally, and both obesity and diet appear to play roles in its pathophysiology. Cross-sectional studies have demonstrated an inverse association between obesity and the prevalence of chronic obstructive pulmonary disease (COPD). Objective. This study aims to evaluate the relationship between smoking, lipid profile and obesity (body composition changes) in a selected groups of population (30 non-smokers, 30 smokers and 60 COPD patients). Material and Methods. We evaluated fat mass, fat free mass, body mass index and lipid profile in a group of 120 randomly selected probands (60 COPD patients; 30 smokers without COPD; 30 non-smokers without COPD) to identify possible negative relationships of smoking to body composition. To the measurement of fat mass (FM) and fat free mass (FFM) was used a device Bodystat Quadscan 4000 (Bodystat Ltd, British Isles). The device works by using four-frequency bioelectrical impedance analysis. Laboratory parameters as total cholesterol (T-C), high-density cholesterol (HDL-C), low-density cholesterol (LDL-C) and triacylglycerols (TG) were investigated by automated clinical chemistry analyzer LISA 200th. The measured values were statistically processed and evaluated in a statistical program STATISTICA Cz. Version 7.1. (Kruskall-Wallis test). Results. A comparison of the mean fat mass we found statistically highly significant differences between the group of COPD patients and non-smokers (P <0.001) and insignificant differences (P ≥ 0.05) between the other groups of our experiment. A comparison of the mean fat mass values of all three groups of the experiment shows a steady increase in fat from non-smokers (17.66 ± 10.04 kg) to COPD patients with the highest mean value (25.08 ± 10.14 kg). In the group of COPD patients we recorded the lowest average value of FFM (51.76 ± 13.84 kg), in group of smokers the middle (56.06 ± 10.76 kg) and in non-smokers the highest average value of FFM (59.91 ± 9.90 kg) at relatively the same body weight in the groups. Based on calculated body mass index (BMI), we found in group of COPD patients overweight in 15 cases (25%), obesity in 7 patients (11.67%), severe obesity in 14 patients (23.3%) and morbid obesity in 2 patients (3.33%); in the group of smokers overweight in 16 cases (53.33%), obesity in 5 cases (16.6%) and severe obesity in 1 case (3.33%); in non-smokers we recording overweight in 14 cases (46.67%), obesity in 5 cases (16.67%) and severe obesity in 2 cases (6.67%). In the lipid profile of the monitored groups of probands, we observed statistically significant differences only for LDL cholesterol (LDL-C). There was a statistically significant difference (P <0.001) between the group with COPD and smokers, as well as between the group of smokers and non-smokers (P < 0.05). Conclusions. In the vast majority of patients with COPD, the lung damage that leads to COPD is caused by long-term cigarette smoking. The presence and absence of risk factors such as smoking, inappropriate lipid profile and obesity (amount of fat mass) in selected population groups were observed. Additional studies to explore both the quantitative and qualitative changes in body composition with disease process of COPD are required.

scholarly journals Dissecting the Role of the Gut Microbiota and Diet on Visceral Fat Mass Accumulation (OR31-07-19)

2019 ◽  
Vol 3 (Supplement_1) ◽  
Author(s):  
Caroline Le Roy ◽  
Ruth Bowyer ◽  
Claire Steves ◽  
Tim Spector ◽  
Bell Jordana
Keyword(s):  
Vitamin E ◽  
Random Forest ◽  
Gut Microbiota ◽  
Fat Mass ◽  
Mediation Analysis ◽  
Visceral Fat ◽  
Nutrient Intake ◽  
Microbiota Composition ◽  
Gut Microbiota Composition

Abstract Objectives Accumulation of visceral fat mass (VFM) is a major risk factor for cardiovascular and metabolic disease. Both gut microbiota and diet have been shown to impact host adiposity in an interdependent manner, but the exact nature of their joint contributions has not been characterised. Here, we aimed to estimate and separate the effect of gut microbiota composition from that of nutrient intake on host VFM in of 1760 older female twins. Methods The gut microbiome profile was assessed by 16S sequencing. VFM was measured by DEXA whole body scan and nutrient intake was assessed through food frequency questionnaires. We used a combination of pair-wise associations, random forest modelling and mediation analysis to separate the effect of the gut microbiota and nutrients on VFM. Results Pairwise analyses revealed that 93 OTUs and 10 nutrients were significantly linked to VFM. Five of the 10 nutrients (fibre, trans fatty acids, magnesium, vitamin E and biotin) were also associated with 23% of the 93 VFM-associated OTUs. To separate the effects of the gut microbiota from nutrients on VFM we carried out conditional analyses. We observed that the majority (87%) of the 93 OTUs remained significantly associated with VFM irrespective of nutrient intake correction. In contrast, we observed that fibre, magnesium, biotin and vitamin E were no longer significantly associated with VFM when adjusting models for OTUs (P > 0.05), implying a role of the gut microbiota in mediating these nutrient effects on VFM. Formal mediation analysis revealed that the individual effect of fibre, biotin, magnesium and vitamin E on VFM were mediated at 69, 43, 41 and 31% respectively by OTUs. Moreover, we estimated that accumulated effect of OTUs on VFM (R2 = 0.19) was twice the one of nutrients (R2 = 0.11) and so were their prediction potential determined using random forest classification. Conclusions Our results suggest that while the role of certain nutrients on VFM appears to depend on gut microbiota composition, specific gut microbes may affect host adiposity regardless of dietary intake. The findings imply that the gut microbiota may have a greater contribution towards shaping host adiposity and VFM, compared to diet alone. Funding Sources We gratefully acknowledge support provided by the JPI HDHL funded DINAMIC consortium (administered by the MRC UK, MR/N030125/1). Supporting Tables, Images and/or Graphs

scholarly journals Aerobic exercise-induced changes in body composition and blood lipids in young women

Medicina ◽  
2010 ◽  
Vol 46 (2) ◽  
pp. 129 ◽  
Author(s):  
Arvydas Stasiulis ◽  
Asta Mockienė ◽  
Daiva Vizbaraitė ◽  
Pranas Mockus
Keyword(s):  
Body Mass Index ◽  
Body Composition ◽  
Body Weight ◽  
Body Mass ◽  
Body Fat ◽  
Fat Mass ◽  
Blood Lipids ◽  
Density Lipoprotein ◽  
Body Fat Mass ◽  
Cycling Training

The objective of the study was to assess changes in body composition, blood lipid and lipoprotein concentrations in 18–24-year-old women during the period of two-month aerobic cycling training. Material and methods. Young, healthy, nonsmoking women (n=19) volunteered to participate in this study. They were divided in two groups: experimental (E, n=10) and control (C, n=9). The subjects of group E exercised 3 times a week with intensity of the first ventilatory threshold and duration of 60 min. The group C did not exercise regularly over a two-month period of the experiment. The subjects of group E were tested before and after 2, 4, 6 and 8 weeks of the experiment. The participants of group C were tested twice with an eight-week interval. Results. Body weight, body mass index, body fat mass, and triacylglycerol (TAG) concentration decreased and high-density lipoprotein cholesterol (HDL-ch) concentration increased after the 8-week training program in the experimental group (P<0.05). Blood total cholesterol (Tch) and low-density lipoprotein cholesterol (LDL-ch) concentrations did not change significantly. Body weight and body mass index started to decrease after 2 weeks of the experiment, but significant changes were observed only after 6 and 8 weeks. Body fat mass was significantly decreased after 2 and 8 weeks of aerobic training. A significant increase in HDL-ch concentration was observed after 4, 6, and 8 weeks. A significant decrease in TAG concentration was observed after 2-week training. No significant changes in all the parameters except TAG (it was slightly increased) were seen in the control group. Conclusions. The two-month aerobic cycling training (within VT1, 60-min duration, three times a week) may induce significant changes in the parameters of body composition – body weight, body mass index, body fat mass, and blood lipids – in young women. The following significant changes were observed: TAG level decreased after 2 weeks, body mass and body mass index decreased after 6 weeks, body fat mass decreased and HDL-ch level increased after 8 weeks. Peak oxygen uptake increased after 4 weeks.

Introduction to Simple Mediation Analysis in SPSS

2015 ◽  
Vol 7 (1) ◽  
pp. 62 ◽  
Author(s):  
Nitika Sharma
Keyword(s):  
Causal Relationship ◽  
Direct Effect ◽  
Indirect Effect ◽  
Mediation Analysis ◽  
Indirect Effects ◽  
Regression Analyses ◽  
Mediation Model ◽  
Ols Regression ◽  
Mediator Variable

Data is analyzed using Mediation model which focuses on the estimation of the indirect effect of X on Y through an intermedi - ary mediator variable M causally located between X and Y (i.e., a model of the form X ? M ? Y ) 1 , where X is the input variable, Y is output and M is the Mediating Variable. When researchers want to examine that how X variable exert it effects on Y variable which is commonly intervened by one or two variables denoted by M and this variable has a causal relationship between X &amp; Y as per Figure 1 and termed as Simple Mediation Model. In this casual system there is at least one casual antecedent X variable is projected as influencing an outcome Y through a single inter - vening variable M . Such model establishes two pathways which influences Y by direct effect and indirect effect. In direct effect, pathways lead from X to Y without passing M. In indirect effects, a pathway of X to Y is lead through M. There are two conse - quent variables forming two equations and these equations can be estimated by conducting OLS regression analyses using SPSS or by using PROCESS.sps in SPSS by Andrew F. Hayes. To add PROCESS by Andrew F. Hayes in SPSS.

Association of DXA-derived Fat Mass Indices and Anthropometry with Blood Lipid Profile in Postmenopausal Women

2007 ◽  
Vol 39 (Supplement) ◽  
pp. S233 ◽  
Author(s):  
Hassanali Vatanparast ◽  
Jonathan Little ◽  
Shannon Duff ◽  
Denise Cornish ◽  
Stephen Cornish ◽  
...  
Keyword(s):  
Lipid Profile ◽  
Postmenopausal Women ◽  
Fat Mass ◽  
Blood Lipid ◽  
Blood Lipid Profile

scholarly journals Mango modulates body fat and plasma glucose and lipids in mice fed a high-fat diet

2011 ◽  
Vol 106 (10) ◽  
pp. 1495-1505 ◽  
Author(s):  
Edralin A. Lucas ◽  
Wenjia Li ◽  
Sandra K. Peterson ◽  
Angela Brown ◽  
Solo Kuvibidila ◽  
...  
Keyword(s):  
Insulin Resistance ◽  
Glucose Tolerance ◽  
Lipid Profile ◽  
Body Fat ◽  
Fat Mass ◽  
Model Assessment ◽  
High Fat ◽  
Treatment Groups ◽  
Epididymal Fat ◽  
Freeze Dried

Consumption of fruits and vegetables has been investigated for their role in the prevention of many chronic conditions. Among the fruits, mango provides numerous bioactive compounds such as carotenoids, vitamin C and phenolic compounds, which have been shown to have antioxidant and anti-inflammatory properties. The present study examined the effects of dietary supplementation of freeze-dried mango pulp, in comparison with the hypolipidaemic drug, fenofibrate, and the hypoglycaemic drug, rosiglitazone, in reducing adiposity and alterations in glucose metabolism and lipid profile in mice fed a high-fat (HF) diet. Male C57BL/6J mice were randomly divided into six treatment groups (eight to nine/group): control (10 % energy from fat); HF (60 % energy from fat); HF+1 or 10 % freeze-dried mango (w/w); HF+fenofibrate (500 mg/kg diet); HF+rosiglitazone (50 mg/kg diet). After 8 weeks of treatment, mice receiving the HF diet had a higher percentage body fat (P = 0·0205) and epididymal fat mass (P = 0·0037) compared with the other treatment groups. Both doses of freeze-dried mango, similar to fenofibrate and rosiglitazone, prevented the increase in epididymal fat mass and the percentage of body fat. Freeze-dried mango supplementation at the 1 % dose improved glucose tolerance as shown by approximately 35 % lower blood glucose area under the curve compared with the HF group. Moreover, freeze-dried mango lowered insulin resistance, as indicated by the homeostasis model assessment of insulin resistance, to a similar extent as rosiglitazone and modulated NEFA. The present findings demonstrate that incorporation of freeze-dried mango in the diet of mice improved glucose tolerance and lipid profile and reduced adiposity associated with a HF diet.

Sign in / Sign up

Export Citation Format

Share Document