This package contains tools for processing the NHANES 2003-2004 and 2005-2006 waves' accelerometry data as well as the resulting processed data. Specifically, the code contained herein will transform the "long format" accelerometry data, which is available for download from the CDC, to the 1440+ format with one row per subject-day, calculate wear/non-wear flags on the tranformed data, merge NHANES data from non-accelerometry sources, and process national mortality data linked to NHANES participants. In addition to code for processing the 2003-2006 accerlerometry data, this package contains code to: process and merge NHANES data across multiple "raw" (.xpt) data files which are downloaded from the NHANES website; and process linked mortality data for the NHANES 2003-2006 participants.
This primary contents of this package are: 1) four data procesing functions and two functions helpful in analyzing NHANES data; 2) processed accelerometry, demographic, lifestyle, comorbiditiy, and mortality data for the NHANES 2003-2006 waves; and 3) raw data used to create the processed data in the package, excluding the ``raw" accelerometer data due to package file size limitaitons. These three sets of content are described in more detail below.
Data processing functions: The data processing functions used here can be used to reproduce all the data files in this package. See the package vignette on processing the NHANES accelerometry data for details on how exactly to reproduce all data in the package using the four functions below (create-all-processed-data). In general, these functions work by specifying the NHANES waves which you wish to process, as well as additional function-specific parameters. The NHANES waves are specified using the NHANES naming convention of "Name"_*.ext where "Name" denotes file name, \* corresponds to the letter associated with a particular NHANES wave (for example, _C and _D correspond to the 2003-2004 and 2005-2006 waves, respectively), and .ext indicates file extension.
process_accel: The function process_accel will take the long-format accelerometry data which is downloaded directly from the CDC website, and transform it to the 1440+ format
with one row per subejct-day, and seven rows per subject. This can either 1) process data that has been downlaoded locally; or 2) downlaod the data from the CDC.
process_flags: The function process_flags will take the 1440+ format accelerometry data which has been created by the process_accel function.
This function uses the weartime function from the accelerometry package to estimate weartime.
See weartime for details on weartime algorithm parameters.
process_covar: The function process_covar is intented to make retrieving and merging data across many different .xpt files easy.
Specificlly, the fucntion will take a vector of variable names associated with NHANES data and search all SAS xport (.xpt) files that match
the NHANES naming convention in a directory for the variables specified.
This function has the option to not specify any variables, and instead return every variable across all files.
process_mort: The function process_mort will take the raw .dat mortality files linked to NHANES participants (downloadable from the CDC) and
transform them into a format compatable with analyses.
Data analysis functions:
reweight_accel The function reweight_accel will calculate 2- and 4-year survey weights for NHANES 2003-2004 and 2005-2006 waves on a dataset containing participants
from both waves. The function has the option to return both adjusted and unadjusted survey weights. This function calculates both interview and examination survey weights.
The adjusted survey weights are calculated based on a missing completely at random assumption within age, gender, and ethinicity strata (see reweight_accel for additional details).
exclude_accel The function exclude_accel will identify days which meet minimum wear-time criteria and are deemed as "good" by on the basis of NHANES supplied
data qualify flags (PAXSTAT) and device calibration indicator (PAXCAL).
The processed data included in the package are broadly divided into four categories: 1) activity count data; 2) wear/non-wear flags; 3) additional NHANES data; 4) linked mortality data.
Activity count data
PAXINTEN_C Accelerometry data for the NHANES 2003-2004 wave
PAXINTEN_D Accelerometry data for the NHANES 2005-2006 wave
Wear/non-wear flags
Flags_C Wear/non-wear flags estimated using the process_flags function applied to the 2003-2004 accelerometry data (PAXINTEN_C)
Flags_D Wear/non-wear flags estimated using the process_flags function applied to the 2005-2006 accelerometry data (PAXINTEN_D)
Additional NHANES data
Covariate_C Processed versions of commonly used demographic, lifestyle, and comordbiditiy data for the NHANES 2003-2004 wave
Covariate_D Processed versions of commonly used demographic, lifestyle, and comordbiditiy data for the NHANES 2005-2006 wave
Linked mortality data
Mortality_2011_C Processed mortality data for the NHANES 2003-2004 wave. Mortality data corresponds to the 2011 release.
Mortality_2011_D Processed mortality data for the NHANES 2005-2006 wave. Mortality data corresponds to the 2011 release.
Finally, this package also contains raw NHANES data (excluding accelerometry) and the 2011 publicly linked mortality data. These data are not publicly facing and therefore do not have
documentation. However, all variables contained in the "raw NHANES data" files listed below are the process_covar function. In addition, the
process_mort uses the raw "linked mortality data" to create the processed mortality data included in the package.
Raw NHANES data
ALQ_C.XPT and ALQ_D.XPT Alcohol consumption questionnaire data for the 2003-2004 and 2005-2006 waves, respectively.
BMX_C.XPT and BMX_D.XPT Body measurement data for the 2003-2004 and 2005-2006 waves, respectively.
BPX_C.XPT and BPX_D.XPT Blood pressure data for the 2003-2004 and 2005-2006 waves, respectively.
DEMO_C.XPT and DEMO_D.XPT Demographic data for the 2003-2004 and 2005-2006 waves, respectively.
DIQ_C.XPT and DIQ_D.XPT Diabetes questionnaire data for the 2003-2004 and 2005-2006 waves, respectively.
MCQ_C.XPT and MCQ_D.XPT Medical conditions questionnaire data for the 2003-2004 and 2005-2006 waves, respectively.
PFQ_C.XPT and PFQ_D.XPT Physical function questionnaire data for the 2003-2004 and 2005-2006 waves, respectively.
SMQ_C.XPT and SMQ_D.XPT Smoking status questionnaire data for the 2003-2004 and 2005-2006 waves, respectively.
Linked mortality data
NHANES_2003_2004_MORT_2011_PUBLIC.dat Raw mortality data for the 2003-2004 wave
NHANES_2005_2006_MORT_2011_PUBLIC.dat Raw mortality data for the 2005-2006 wave
National Center for Health Statistics. Office of Analysis and Epidemiology, Public-use Linked Mortality File, 2015. Hyattsville, Maryland. (Available at the following address: http://www.cdc.gov/nchs/data_access/data_linkage/mortality.htm
Centers for Disease Control and Prevention (CDC). National Center for Health Statistics (NCHS). National Health and Nutrition Examination Survey Data. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. Available at: https://wwwn.cdc.gov/nchs/nhanes/default.aspx. Accessed Oct. 10, 2018.
Choi, Leena et al. “Validation of Accelerometer Wear and Nonwear Time Classification Algorithm.” Medicine and science in sports and exercise 43.2 (2011): 357–364. PMC. Web. 10 Oct. 2018.
National Cancer Institute. Risk factor monitoring and methods: SAS programs for analyzing NHANES 2003-2004 accelerometer data. Available at: http://riskfactor.cancer.gov/tools/nhanes_pam. Accessed Oct. 10, 2018.
Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, Mcdowell M: Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc 2008; 40: 181-188.