1
|
Liu Y, Wang Y, Feng Y, Wall MM. VARIABLE SELECTION AND PREDICTION WITH INCOMPLETE HIGH-DIMENSIONAL DATA. Ann Appl Stat 2016; 10:418-450. [PMID: 27213023 DOI: 10.1214/15-aoas899] [Citation(s) in RCA: 25] [Impact Index Per Article: 3.1] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/19/2022]
Abstract
We propose a Multiple Imputation Random Lasso (mirl) method to select important variables and to predict the outcome for an epidemiological study of Eating and Activity in Teens. In this study 80% of individuals have at least one variable missing. Therefore, using variable selection methods developed for complete data after listwise deletion substantially reduces prediction power. Recent work on prediction models in the presence of incomplete data cannot adequately account for large numbers of variables with arbitrary missing patterns. We propose MIRL to combine penalized regression techniques with multiple imputation and stability selection. Extensive simulation studies are conducted to compare MIRL with several alternatives. MIRL outperforms other methods in high-dimensional scenarios in terms of both reduced prediction error and improved variable selection performance, and it has greater advantage when the correlation among variables is high and missing proportion is high. MIRL is shown to have improved performance when comparing with other applicable methods when applied to the study of Eating and Activity in Teens for the boys and girls separately, and to a subgroup of low social economic status (ses) Asian boys who are at high risk of developing obesity.
Collapse
Affiliation(s)
- Ying Liu
- Department of Biostatistics, Columbia University, 722 West 168th Street, New York, New York 10032, USA,
| | - Yuanjia Wang
- Department of Biostatistics, Columbia University, 722 West 168th Street, New York, New York, 10032, USA,
| | - Yang Feng
- Department of Statistics, Columbia University, 1255 Amsterdam Ave., 10th Floor, MC 4690, New York, New York 10027, USA,
| | - Melanie M Wall
- Department of Biostatistics, Columbia University, 722 West 168th Street, New York, New York 10032, USA,
| |
Collapse
|
2
|
Brittin J, Sorensen D, Trowbridge M, Lee KK, Breithecker D, Frerichs L, Huang T. Physical Activity Design Guidelines for School Architecture. PLoS One 2015; 10:e0132597. [PMID: 26230850 PMCID: PMC4521876 DOI: 10.1371/journal.pone.0132597] [Citation(s) in RCA: 22] [Impact Index Per Article: 2.4] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/06/2014] [Accepted: 06/16/2015] [Indexed: 11/19/2022] Open
Abstract
Increasing children's physical activity at school is a national focus in the U.S. to address childhood obesity. While research has demonstrated associations between aspects of school environments and students' physical activity, the literature currently lacks a synthesis of evidence to serve as a practical, spatially-organized resource for school designers and decision-makers, as well as to point to pertinent research opportunities. This paper describes the development of a new practical tool: Physical Activity Design Guidelines for School Architecture. Its aims are to provide architects and designers, as well as school planners, educators, and public health professionals, with strategies for making K-12 school environments conducive to healthy physical activity, and to engage scientists in transdisciplinary perspectives toward improved knowledge of the school environment's impact. We used a qualitative review process to develop evidence-based and theory-driven school design guidelines that promote increased physical activity among students. The design guidelines include specific strategies in 10 school design domains. Implementation of the guidelines is expected to enable students to adopt healthier physical activity behaviors. The tool bridges a translational gap between research and environmental design practice, and may contribute to setting new industry and education standards.
Collapse
Affiliation(s)
- Jeri Brittin
- University of Nebraska Medical Center, College of Public Health, Department of Health Promotion, Social & Behavioral Health, Omaha, NE, United States of America
| | - Dina Sorensen
- VMDO Architects, Charlottesville, VA, United States of America
| | - Matthew Trowbridge
- University of Virginia School of Medicine, Department of Emergency Medicine, Charlottesville, VA, United States of America
| | - Karen K. Lee
- New York City Department of Health and Mental Hygiene, New York, NY, United States of America
| | - Dieter Breithecker
- Federal Institute for Posture and Mobilisation Support, Wiesbaden, Germany
| | - Leah Frerichs
- University of Nebraska Medical Center, College of Public Health, Department of Health Promotion, Social & Behavioral Health, Omaha, NE, United States of America
- University of North Carolina, Chapel Hill, Center for Health Equity, Chapel Hill, NC, United States of America
| | - Terry Huang
- University of Nebraska Medical Center, College of Public Health, Department of Health Promotion, Social & Behavioral Health, Omaha, NE, United States of America
- City University of New York, School of Public Health, New York, NY, United States of America
| |
Collapse
|
3
|
Pettigrew S, Borys JM, du Plessis HR, Walter L, Huang TTK, Levi J, Vinck J. Process evaluation outcomes from a global child obesity prevention intervention. BMC Public Health 2014; 14:757. [PMID: 25065414 PMCID: PMC4122757 DOI: 10.1186/1471-2458-14-757] [Citation(s) in RCA: 21] [Impact Index Per Article: 2.1] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/09/2013] [Accepted: 07/22/2014] [Indexed: 12/22/2022] Open
Abstract
Background While it is acknowledged that child obesity interventions should cover multiple ecological levels (downstream, midstream and upstream) to maximize their effectiveness, there is a lack of evaluation data to guide the development and implementation of such efforts. To commence addressing this knowledge gap, the present study provides process evaluation data relating to the experiences of groups implementing the EPODE approach to child obesity prevention in various locations around the world. The aim of this exploratory study was to investigate the barriers and facilitators to program implementation in program sites around the world to assist in developing strategies to enhance program outcomes. Methods An online survey that included open-ended questions was distributed to the 25 EPODE programs in operation at the time of the survey (May 2012). The survey items asked respondents to comment on those aspects of program implementation that they found challenging and to suggest areas for future improvement. Eighteen programs representing 14 countries responded to the request to participate in the survey, yielding a 72% response rate. The responses were analyzed via the constant comparative method using NVivo qualitative data analysis software. Results The main concerns of the various EPODE programs were their ability to secure ongoing funding and their access to evidence-based intervention methods and policy advice relating to relationships with third parties. These issues were in turn impacted by other factors, including (i) access to user-friendly information relating to the range of intervention strategies available and appropriate evaluation measures; (ii) assistance with building and maintaining stakeholder relationships; and (iii) assurance of the quality, independence, and transparency of policies and practices. Conclusions The findings are facilitating the ongoing refinement of the EPODE approach. In particular, standardized and tailored information packages are being made available to advise program members of (i) the various evaluation methods and tools at their disposal and (ii) methods of acquiring private partner support. Overall, the study results relating to the types of issues encountered by program members are likely to be useful in guiding the future design and implementation of multi-level initiatives seeking to address other complex and intractable health-related problems.
Collapse
Affiliation(s)
- Simone Pettigrew
- School of Psychology and Speech Pathology, CurtinUniversity, Kent St, Bentley, WA 6102, Australia.
| | | | | | | | | | | | | |
Collapse
|
4
|
Puricelli Perin DM, Frerichs L, Costa S, Ramirez AG, Huang TTK. Perception of childhood obesity and support for prevention policies among Latinos and Whites. J Obes 2014; 2014:328276. [PMID: 25045532 PMCID: PMC4089851 DOI: 10.1155/2014/328276] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 04/01/2014] [Accepted: 06/02/2014] [Indexed: 11/17/2022] Open
Abstract
A cross-sectional survey was administered to Latino and White residents of Omaha, NE, to assess perception of the childhood obesity problem, attribution of responsibility, and support for obesity-related policies. The sample included 40.8% (n = 271) Latinos and 59.2% (n = 393) Whites. Among Latinos, 25% did not see childhood obesity as a problem, compared to 6% of Whites (P < 0.001). This difference persisted after adjusting for age, gender, and education level (odds ratio (OR) 2.10, 95% confidence interval (CI) 1.07-4.14). Latinos were more likely to agree that government was responsible for addressing childhood obesity compared to Whites (OR 2.81, 95% CI 1.82-4.35). Higher support for policy interventions was observed among individuals who perceived childhood obesity as a big problem compared to those who did not, independent of race, sex, age, or education level. The relationship between support for tax-based policies and perception of the childhood obesity problem was mainly evident among Latinos rather than Whites. Despite city-wide efforts to address obesity, differential penetration in community subgroups appears evident. There is room to further engage Latinos in the cause of obesity. Deepening community awareness about the consequences and complexity of childhood obesity can lead to stronger support for childhood obesity policy interventions.
Collapse
Affiliation(s)
- Douglas M. Puricelli Perin
- Center for Global Health and Development, College of Public Health, University of Nebraska Medical Center, Omaha, NE 68198-4341, USA
| | - Leah Frerichs
- Department of Health Promotion and Social and Behavioral Health, College of Public Health, University of Nebraska Medical Center, 984365 Nebraska Medical Center, Omaha, NE 68198-4365, USA
| | - Sergio Costa
- University of Nebraska Medical Center, Omaha, NE 68198-4355, USA
| | - Amelie G. Ramirez
- Institute for Health Promotion Research, University of Texas Health Science Center, San Antonio, TX 78229, USA
| | - Terry T.-K. Huang
- Department of Health Promotion and Social and Behavioral Health, College of Public Health, University of Nebraska Medical Center, 984365 Nebraska Medical Center, Omaha, NE 68198-4365, USA
| |
Collapse
|
5
|
van Koperen MT, van der Kleij RM, Renders CC, Crone MM, Hendriks AMA, Jansen MM, van de Gaar VV, Raat HJ, Ruiter EE, Molleman GG, Schuit JA, Seidell JJ. Design of CIAO, a research program to support the development of an integrated approach to prevent overweight and obesity in the Netherlands. BMC OBESITY 2014. [PMID: 26217497 PMCID: PMC4472617 DOI: 10.1186/2052-9538-1-5] [Citation(s) in RCA: 13] [Impact Index Per Article: 1.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Indexed: 11/10/2022]
Abstract
BACKGROUND The aim of this paper is to describe the research aims, concepts and methods of the research Consortium Integrated Approach of Overweight (CIAO). CIAO is a concerted action of five Academic Collaborative Centres, local collaborations between academic institutions, regional public health services, local authorities and other relevant sectors in the Netherlands. Prior research revealed lacunas in knowledge of and skills related to five elements of the integrated approach of overweight prevention in children (based upon the French EPODE approach), namely political support, parental education, implementation, social marketing and evaluation. CIAO aims to gain theoretical and practical insight of these elements through five sub-studies and to develop, based on these data, a framework for monitoring and evaluation. METHODS/DESIGN For this research program, mixed methods are used in all the five sub-studies. First, problem specification through literature research and consultation of stakeholders, experts, health promotion specialists, parents and policy makers will be carried out. Based on this information, models, theoretical frameworks and practical instruments will be developed, tested and evaluated in the communities that implement the integrated approach to prevent overweight in children. Knowledge obtained from these studies and insights from experts and stakeholders will be combined to create an evaluation framework to evaluate the integrated approach at central, local and individual levels that will be applicable to daily practice. DISCUSSION This innovative research program stimulates sub-studies to collaborate with local stakeholders and to share and integrate their knowledge, methodology and results. Therefore, the output of this program (both knowledge and practical tools) will be matched and form building blocks of a blueprint for a local evidence- and practice-based integrated approach towards prevention of overweight in children. The output will then support various communities to further optimize the implementation and subsequently the effects of this approach.
Collapse
Affiliation(s)
- Marije Tm van Koperen
- Faculty of Earth and Life Sciences, Institute of Health Sciences, VU University Amsterdam, De Boelelaan 1085, 1081 HV Amsterdam, The Netherlands
| | - Rianne Mjj van der Kleij
- Department of Public Health & Primary Care, Leiden University Medical Centre, PO Box 9600, 2300 RC Leiden, The Netherlands
| | - Carry Cm Renders
- Faculty of Earth and Life Sciences, Institute of Health Sciences, VU University Amsterdam, De Boelelaan 1085, 1081 HV Amsterdam, The Netherlands
| | - Matty Mr Crone
- Department of Public Health & Primary Care, Leiden University Medical Centre, PO Box 9600, 2300 RC Leiden, The Netherlands
| | | | - Maria M Jansen
- Caphri research School for Public Health and Primary care, PO Box 616, 6200 MD Maastricht, The Netherlands
| | - Vivian Vm van de Gaar
- Public Health, Erasmus University Medical Center, PO Box 2040, 3000 CA Rotterdam, The Netherlands
| | - Hein Jh Raat
- Erasmus University Medical Center, child and adolescent public health, PO Box 2040, 3000 CA Rotterdam, The Netherlands
| | - Emilie Elm Ruiter
- Department of Primary and Community Care, Academic Collaborative Center AMPHI Integrated Health Policy, Radboud University Medical Centre, PO Box 9101, 6500 HB Nijmegen, The Netherlands
| | - Gerard Grm Molleman
- Department of Primary and Community Care, Academic Collaborative Center AMPHI Integrated Health Policy, Radboud University Medical Centre, PO Box 9101, 6500 HB Nijmegen, The Netherlands
| | - Jantine Aj Schuit
- National Institute for Public Health and the Environment (RIVM), PO Box 1 3720, BA Bilthoven, The Netherlands
| | - Jacob Jc Seidell
- Faculty of Earth and Life Sciences, Institute of Health Sciences, VU University Amsterdam, De Boelelaan 1085, 1081 HV Amsterdam, The Netherlands
| |
Collapse
|