29 resultados para Generalized estimating equations
em Deakin Research Online - Australia
Resumo:
In epidemiologic studies, researchers often need to establish a nonlinear exposure-response relation between a continuous risk factor and a health outcome. Furthermore, periodic interviews are often conducted to take repeated measurements from an individual. The authors proposed to use fractional polynomial models to jointly analyze the effects of 2 continuous risk factors on a health outcome. This method was applied to an analysis of the effects of age and cumulative fluoride exposure on forced vital capacity in a longitudinal study of lung function carried out among aluminum workers in Australia (1995-2003). Generalized estimating equations and the quasi-likelihood under the independence model criterion were used. The authors found that the second-degree fractional polynomial models for age and fluoride fitted the data best. The best model for age was robust across different models for fluoride, and the best model for fluoride was also robust. No evidence was found to suggest that the effects of smoking and cumulative fluoride exposure on change in forced vital capacity over time were significant. The trend 1 model, which included the unexposed persons in the analysis of trend in forced vital capacity over tertiles of fluoride exposure, did not fit the data well, and caution should be exercised when this method is used.
Resumo:
Objectives : To evaluate the effectiveness of an intervention to prevent excess weight gain, reduce time spent in screen behaviours, promote participation in and enjoyment of physical activity (PA), and improve fundamental movement skills among children.
Participants : In 2002, 311 children (78% response; 49% boys), average age 10 years 8 months, were recruited from three government schools in low socioeconomic areas of Melbourne, Australia.
Design : Group-randomized controlled trial. Children were randomized by class to one of the four conditions: a behavioural modification group (BM; n=66); a fundamental movement skills group (FMS; n=74); a combined BM/FMS group (BM/FMS; n=93); and a control (usual curriculum) group (n=62). Data were collected at baseline, post intervention, 6- and 12-month follow-up periods.
Results : BMI data were available for 295 children at baseline and 268 at 12-month follow-up. After adjusting for food intake and PA, there was a significant intervention effect from baseline to post intervention on age- and sex-adjusted BMI in the BM/FMS group compared with controls (-1.88 kg m-2, P<0.01), which was maintained at 6- and 12-month follow-up periods (-1.53 kg m-2, P<0.05). Children in the BM/FMS group were less likely than controls to be overweight/obese between baseline and post intervention (adjusted odds ratio (AOR)=0.36, P<0.05); also maintained at 12-month follow-up (AOR=0.38, P<0.05). Compared with controls, FMS group children recorded higher levels and greater enjoyment of PA; and BM children recorded higher levels of PA and TV viewing across all four time points. Gender moderated the intervention effects for participation in and enjoyment of PA, and fundamental movement skills.
Conclusion : This programme represents a promising approach to preventing excess weight gain and promoting participation in and enjoyment of PA. Examination of the mediators of this intervention and further tailoring of the programme to suit both genders is required.
Resumo:
The Generalized Estimating Equations (GEE) method is one of the most commonly used statistical methods for the analysis of longitudinal data in epidemiological studies. A working correlation structure for the repeated measures of the outcome variable of a subject needs to be specified by this method. However, statistical criteria for selecting the best correlation structure and the best subset of explanatory variables in GEE are only available recently because the GEE method is developed on the basis of quasi-likelihood theory. Maximum likelihood based model selection methods, such as the widely used Akaike Information Criterion (AIC), are not applicable to GEE directly. Pan (2001) proposed a selection method called QIC which can be used to select the best correlation structure and the best subset of explanatory variables. Based on the QIC method, we developed a computing program to calculate the QIC value for a range of different distributions, link functions and correlation structures. This program was written in Stata software. In this article, we introduce this program and demonstrate how to use it to select the most parsimonious model in GEE analyses of longitudinal data through several representative examples.
Resumo:
Background Given the importance of physical activity for health and age-related declines in physical activity, understanding influences on related behaviours, such as time outdoors, is crucial. This study aimed to understand individual, social and physical environmental influences on longitudinal changes in urban children’s time outdoors.
Methods The time children spent outdoors in 2001, 2004 and 2006 (aged 5e6 and 10e12 years at baseline) was reported by their parents (n¼421). In 2001, individual, social and physical environmental factors were self-reported by parents. Generalized estimating equations examined longitudinal relationships between baseline predictors and average change in time outdoors over 5 years.
Results Children’s time outdoors significantly declined over time. “Indoor tendencies” inversely predicted time outdoors among younger and older boys, and younger girls. Social opportunities positively predicted time outdoors among younger boys, while “outdoor tendencies” positively predicted time outdoors among older boys. Parental encouragement for activity positively predicted time outdoors among younger and older girls,while lack of adult supervision for active play outdoors after school inversely predicted time outdoors among older girls and older boys.
Conclusion Individual (indoor and outdoor tendencies) and social factors (social opportunities, parental encouragement and parental supervision) predicted children’s time outdoors over 5 years. Interventions targeting reduced indoor tendencies, increased outdoor play with others, and increased parental encouragement and supervision are warranted.
Resumo:
Objective: To determine the independent contributions of family and neighbourhood environments to changes in youth physical activity and body mass index (BMI) z-score over 5 years.
Methods: In 2001, 2004 and 2006, 301 children (10–12 years at baseline) had their height and weight measured (BMI was converted to z-scores using Centers for Disease Control and Prevention reference charts; see http://www.cdc.gov/growthcharts) and moderate-to-vigorous physical activity (MVPA) assessed using accelerometers. In 2001, parents reported on the home environment (social support, role modelling, rules and restrictions, physical environment) and perceived neighbourhood environment (local traffic, road safety, sporting venues, public transport), and Geographic Information Systems were used to map features of the neighbourhood environment (destinations, road connectivity, traffic exposure). Generalized estimating equations were used to predict average BMI z-score and MVPA over time from baseline home and perceived and objective neighbourhood environment factors.
Results: Among boys, maternal education and heavy traffic were inversely associated, and sibling physical activity, maternal role modelling of MVPA and the presence of dead-end roads were positively associated with MVPA. Having unmarried parents, maternal MVPA role modelling and number of home sedentary items were positively associated with BMI z-score among boys. Among girls, having siblings, paternal MVPA role modelling, physical activity rules and parental physical activity co-participation were positively associated with MVPA. Having unmarried parents and maternal sedentary behaviour role modelling were positively associated, and number of sedentary behaviour rules and physical activity items were inversely associated with BMI z-score among girls.
Conclusion: The home environment seems more important than the neighbourhood environment in influencing children's physical activity and BMI z-score over 5 years. Physical activity and weight gain programmes among youth should focus on parental role modelling, rules around sedentary and active pursuits, and parental support for physical activity. Intervention studies to investigate these strategies are warranted.
Resumo:
Background: Children who participate in regular physical activity obtain health benefits. Preliminary pedometerbased cut-points representing sufficient levels of physical activity among youth have been established; however limited evidence regarding correlates of achieving these cut-points exists. The purpose of this study was to identify correlates of pedometer-based cut-points among elementary school-aged children.
Method: A cross-section of children in grades 5-7 (10-12 years of age) were randomly selected from the most (n = 13) and least (n = 12) ‘walkable’ public elementary schools (Perth, Western Australia), stratified by socioeconomic status. Children (n = 1480; response rate = 56.6%) and parents (n = 1332; response rate = 88.8%) completed a survey, and steps were collected from children using pedometers. Pedometer data were categorized to reflect the sex-specific pedometer-based cut-points of ≥15000 steps/day for boys and ≥12000 steps/day for girls. Associations between socio-demographic characteristics, sedentary and active leisure-time behavior, independent mobility, active transportation and built environmental variables - collected from the child and parent surveys - and meeting pedometer-based cut-points were estimated (odds ratios: OR) using generalized estimating equations.
Results: Overall 927 children participated in all components of the study and provided complete data. On average, children took 11407 ± 3136 steps/day (boys: 12270 ± 3350 vs. girls: 10681 ± 2745 steps/day; p < 0.001) and 25.9% (boys: 19.1 vs. girls: 31.6%; p < 0.001) achieved the pedometer-based cut-points. After adjusting for all other variables and school clustering, meeting the pedometer-based cut-points was negatively associated (p < 0.05) with being male (OR = 0.42), parent self-reported number of different destinations in the neighborhood (OR 0.93), and a friend’s (OR 0.62) or relative’s (OR 0.44, boys only) house being at least a 10-minute walk from home. Achieving the pedometer-based cut-points was positively associated with participating in screen-time < 2 hours/day (OR 1.88), not being driven to school (OR 1.48), attending a school located in a high SES neighborhood (OR 1.33), the average number of steps among children within the respondent’s grade (for each 500 step/day increase: OR 1.29), and living further than a 10-minute walk from a relative’s house (OR 1.69, girls only).
Conclusions: Comprehensive multi-level interventions that reduce screen-time, encourage active travel to/from school and foster a physically active classroom culture might encourage more physical activity among children.
Resumo:
Purpose. Although the family environment is a potentially important influence on children's physical activity (PA), prospective data investigating these associations are lacking. This study aimed to examine the longitudinal relationship between the family environment and PA among youth.
Design. A 5-year prospective cohort study.
Setting. Nineteen randomly selected public schools in Melbourne, Australia.
Subjects. Families of 5- to 6-year-old (n = 190) and 10- to 12-year-old (n = 350) children.
Measures. In 2001, parents reported their participation in PA, family-based PA, and support and reinforcement for their child's PA. In 2001, 2004, and 2006, moderate to vigorous intensity PA (MVPA) was assessed among youth using accelerometers. Weekend and “critical window” (after school until 6:00 p.m.) MVPA were examined because we hypothesized that the family environment would most likely influence these behaviors.
Analysis. Generalized estimating equations predicted average change in MVPA over 5 years from baseline family environment factors.
Results. Maternal role modeling was positively associated with boys' critical window and weekend (younger boys) MVPA. Paternal reinforcement of PA was positively associated with critical window and weekend MVPA among all boys, and paternal direct support was positively associated with weekend MVPA (older boys). Among girls, maternal coparticipation in PA predicted critical window MVPA, and sibling coparticipation in PA was directly associated with weekend MVPA (younger girls).
Conclusions. Longitudinal relationships, although weak in magnitude, were observed between the family environment and MVPA among youth. Interventions promoting maternal role modeling, paternal reinforcement of and support for PA, and maternal and sibling coparticipation in PA with youth are warranted.
Resumo:
Background: The effectiveness of lifestyle interventions in reducing diabetes incidence has been well established. Little is known, however, about factors influencing the reach of diabetes prevention programs. This study examines the predictors of enrolment in the Sydney Diabetes Prevention Program (SDPP), a community-based diabetes prevention program conducted in general practice, New South Wales, Australia from 2008–2011.
Methods: SDPP was an effectiveness trial. Participating general practitioners (GPs) from three Divisions of General Practice invited individuals aged 50–65 years without known diabetes to complete the Australian Type 2 Diabetes Risk Assessment tool. Individuals at high risk of diabetes were invited to participate in a lifestyle modification program. A multivariate model using generalized estimating equations to control for clustering of enrolment outcomes by GPs was used to examine independent predictors of enrolment in the program. Predictors included age, gender, indigenous status, region of birth, socio-economic status, family history of diabetes, history of high glucose, use of anti-hypertensive medication, smoking status, fruit and vegetable intake, physical activity level and waist measurement.
Results: Of the 1821 eligible people identified as high risk, one third chose not to enrol in the lifestyle program. In multivariant analysis, physically inactive individuals (OR: 1.48, P = 0.004) and those with a family history of diabetes (OR: 1.67, P = 0.000) and history of high blood glucose levels (OR: 1.48, P = 0.001) were significantly more likely to enrol in the program. However, high risk individuals who smoked (OR: 0.52, P = 0.000), were born in a country with high diabetes risk (OR: 0.52, P = 0.000), were taking blood pressure lowering medications (OR: 0.80, P = 0.040) and consumed little fruit and vegetables (OR: 0.76, P = 0.047) were significantly less likely to take up the program.
Conclusions: Targeted strategies are likely to be needed to engage groups such as smokers and high risk ethnic groups. Further research is required to better understand factors influencing enrolment in diabetes prevention programs in the primary health care setting, both at the GP and individual level.
Resumo:
To achieve valid conclusions, studies exploring associations of the built environment with residents' physical activity and health-related outcomes need to employ statistical approaches accounting for clustered data. This article discusses the following main statistical approaches: analysis of covariance, regression models with robust standard errors, generalized estimating equations, and multilevel generalized linear models. The choice of a statistical method depends on the characteristics of the study and research questions. While the first three approaches are employed to account for clustering in the data, multilevel models can also help unravel more substantive issues within a social ecological theoretical framework of health behavior.
Resumo:
BACKGROUND: Information seeking is an important coping mechanism for dealing with chronic illness. Despite a growing number of mental health websites, there is little understanding of how patients with bipolar disorder use the Internet to seek information.
METHODS: A 39 question, paper-based, anonymous survey, translated into 12 languages, was completed by 1222 patients in 17 countries as a convenience sample between March 2014 and January 2016. All patients had a diagnosis of bipolar disorder from a psychiatrist. Data were analyzed using descriptive statistics and generalized estimating equations to account for correlated data.
RESULTS: 976 (81 % of 1212 valid responses) of the patients used the Internet, and of these 750 (77 %) looked for information on bipolar disorder. When looking online for information, 89 % used a computer rather than a smartphone, and 79 % started with a general search engine. The primary reasons for searching were drug side effects (51 %), to learn anonymously (43 %), and for help coping (39 %). About 1/3 rated their search skills as expert, and 2/3 as basic or intermediate. 59 % preferred a website on mental illness and 33 % preferred Wikipedia. Only 20 % read or participated in online support groups. Most patients (62 %) searched a couple times a year. Online information seeking helped about 2/3 to cope (41 % of the entire sample). About 2/3 did not discuss Internet findings with their doctor.
CONCLUSION: Online information seeking helps many patients to cope although alternative information sources remain important. Most patients do not discuss Internet findings with their doctor, and concern remains about the quality of online information especially related to prescription drugs. Patients may not rate search skills accurately, and may not understand limitations of online privacy. More patient education about online information searching is needed and physicians should recommend a few high quality websites.
Resumo:
Objective. To explore the association between home environmental variables and television (TV) time, and the mediating pathways underlying this association.
Methods. The current study used data from the longitudinal ENDORSE study. Self-reported data was available for 1 265 adolescents (mean age of 12–15 years at baseline) on home environment (availability of a TV in the bedroom, perceived parental modelling, family rules), potential mediators (intention, attitude, perceived behavioural control, subjective norm towards TV viewing) and TV viewing time. Mediation analyses were conducted using General Estimating Equations and mediation effects were calculated as the product-of-coefficients.
Results. Significant overall positive associations were found for the presence of a TV in the bedroom and parental modelling with self-reported TV viewing. Controlling family rules showed an inverse association with reported TV time. Similarly, parental modelling and a TV in the bedroom were significantly positively associated with the Theory of Planned Behaviour variables and habit strength, while family rules showed an inverse association with these potential mediators. In turn, most potential mediators were positively associated with TV viewing. Intention, attitude and habit strength were the strongest mediators in all three associations explaining more than 55% of the overall association. Habit strength alone explained 38.2%–58.0% of the overall associations.
Conclusions. Home and family environmental predictors of TV time among adolescents may be strongly mediated by habit strength and other personal factors. Future intervention studies should explore if changes in home and family environments indeed lead to reductions in TV time through these mediators.
Resumo:
Background : As few longitudinal studies have examined how active transport is associated with physical activity among children and adolescents over time, and how active transport tracks through childhood and adolescence, it is important to understand whether physically active children retain their activity patterns through adolescence. This study aimed to examine (a) tracking of active transport and of moderate-to-vigorous physical activity (MVPA) across childhood and adolescence in two age cohorts; and (b) associations between active transport and MVPA at three distinct time-points, over five years.
Methods : This longitudinal study of two cohorts aged 5-6 years (n = 134) and 10-12 years (n = 201) at baseline (T1), in Melbourne, Australia, gathered follow-up data at three (T2) and five years (T3). Walking/cycling to local destinations was survey-reported; while MVPA was recorded using accelerometers and mean time spent daily in MVPA on week days and on weekends was computed. Tracking of these behaviours was examined over five years using General Estimating Equations. Linear regression analyses were performed to examine associations between active transport and MVPA at each time-point.
Results : Active transport tracked moderately among children (boys, bs = 0.36; girls, bs = 0.51) but not among adolescents. Physical activity tracked moderately (bs value range: 0.33-0.55) for both cohorts. Active transport was not associated with children’s MVPA at any time-point, but was associated with adolescent boys’ MVPA on week days at T1 (B = 1.37 (95% CI: 0.15, 2.59)), at T2 (B = 1.27 (95% CI: 0.03, 2.51)) and at T3 (B = 0.74 (95% CI: 0.01, 1.47)), and with adolescent girls’ MVPA on week days (B = 0.40 (95% CI: 0.04, 0.76)) and on weekends (B = 0.54 (95% CI:0.16, 0.93)) at T3 only.
Conclusion : Active transport was associated only with boys’ MVPA during early adolescence and with boys’ and girls’ MVPA during late adolescence. While active transport should be encouraged among all school-aged children, it may provide an important source of habitual physical activity for adolescent girls, in particular, among whom low and declining physical activity levels have been reported world-wide.
Resumo:
Background
The successful Greater Green Triangle Diabetes Prevention Program (GGT DPP), a small implementation trial, has been scaled-up to the Victorian state-wide ‘Life!’ programme with over 10,000 individuals enrolled. The Melbourne Diabetes Prevention Study (MDPS) is an evaluation of the translation from the GGT DPP to the Life! programme. We report results from the preliminary phase (pMDPS) of this evaluation.
Methods
The pMDPS is a randomised controlled trial with 92 individuals aged 50 to 75 at high risk of developing type 2 diabetes randomised to Life! or usual care. Intervention consisted of six structured 90-minute group sessions: five fortnightly sessions and the final session at 8 months. Participants underwent anthropometric and laboratory tests at baseline and 12 months, and provided self-reported psychosocial, dietary, and physical activity measures. Intervention group participants additionally underwent these tests at 3 months. Paired t tests were used to analyse within-group changes over time. Chi-square tests were used to analyse differences between groups in goals met at 12 months. Differences between groups for changes over time were tested with generalised estimating equations and analysis of covariance.
Results
Intervention participants significantly improved at 12 months in mean body mass index (−0.98 kg/m2, standard error (SE) = 0.26), weight (−2.65 kg, SE = 0.72), waist circumference (−7.45 cm, SE = 1.15), and systolic blood pressure (−3.18 mmHg, SE = 1.26), increased high-density lipoprotein-cholesterol (0.07 mmol/l, SE = 0.03), reduced energy from total (−2.00%, SE = 0.78) and saturated fat (−1.54%, SE = 0.41), and increased fibre intake (1.98 g/1,000 kcal energy, SE = 0.47). In controls, oral glucose at 2 hours deteriorated (0.59 mmol/l, SE = 0.27). Only waist circumference reduced significantly (−4.02 cm, SE = 0.95).
Intervention participants significantly outperformed controls over 12 months for body mass index and fibre intake. After baseline adjustment, they also showed greater weight loss and reduced saturated fat versus total energy intake.
At least 5% weight loss was achieved by 32% of intervention participants versus 0% controls.
Conclusions
pMDPS results indicate that scaling-up from implementation trial to state-wide programme is possible. The system design for Life! was fit for purpose of scaling-up from efficacy to effectiveness.
Resumo:
The aim of this study was to investigate the convergent and predictive validity of two skill tests that examine the ability of golfers to hit accurate approach-iron shots. Twenty-four high-level golfers (handicap = 2.6 ± 1.7) performed the Nine-Ball Skills Test (assesses the ability to shape/control ball trajectory with high accuracy) and the Approach-Iron Skill Test (assesses the ability to hit straight shots from varying distances with high accuracy). Participants then completed at least eight rounds of tournament golf over the following 90 days and reported an indicator of approach-iron accuracy (per cent error index). A moderate correlation (r = 0.50, P < 0.05) was noted between scores for both tests. Generalised estimating equations, using two covariates (lie of the ball and distance to hole), were used to determine model fit and the amount of variance explained for tournament per cent error index. Results showed that the Approach-Iron Skill Test was the slightly stronger predictor of on-course per cent error index. With both test scores considered together, a minimal amount of additional variance was explained. These findings suggest that either of the tests used individually or combined may be used to predict tournament approach iron performance in high-level golfers.
Resumo:
Abstract
Background: The home food environment is an important setting for the development of dietary patterns in childhood. Interventions that support parents to modify the home food environment for their children, however, may also improve parent diet. The purpose of this study was to assess the impact of a telephone-based intervention targeting the home food environment of preschool children on the fruit and vegetable consumption of parents.
Methods: In 2010, 394 parents of 3 – 5 year – old children from 30 preschools in the Hunter region of Australia were recruited to this cluster randomised controlled trial and were randomly assigned to an intervention or control group. Intervention group parents received four weekly 30-minute telephone calls and written resources. The scripted calls focused on; fruit and vegetable availability and accessibility, parental role-modelling, and supportive home food routines. Two items from the Australian National Nutrition Survey were used to assess the average number of serves of fruit and vegetables consumed each day by parents at baseline, and 2-, 6-, 12-, and 18-months later, using generalised estimating equations (adjusted for baseline values and clustering by preschool) and an intention-to-treat-approach.
Results: At each follow-up, vegetable consumption among intervention parents significantly exceeded that of controls. At 2-months the difference was 0.71 serves (95% CI: 0.58-0.85, p < 0.0001), and at 18-months the difference was 0.36 serves (95% CI: 0.10-0.61, p = 0.0067). Fruit consumption among intervention parents was found to significantly exceed consumption of control parents at the 2-,12- and 18-month follow-up, with the difference at 2-months being 0.26 serves (95% CI: 0.12-0.40, p = 0.0003), and 0.26 serves maintained at 18-months, (95% CI: 0.10-0.43, p = 0.0015).
Conclusions: A four-contact telephone-based intervention that focuses on changing characteristics of preschoolers’ home food environment can increase parents’ fruit and vegetable consumption.
(ANZCTR12609000820202)
