ADaM BDS of Changes in Weight and/or Body Mass Index (BMI)

Illustrates Domains

ADEFF

Illustrates Variables

ABLFL

Illustrates Data Structures

BDS

Content

This example includes abbreviated ADaM datasets together with metadata to demonstrate possible ways to construct analysis datasets to analyze vital signs data in pediatric type 1 diabetes (T1D) subjects. The variables presented are those that may be of specific interest to analysis of changes in weight and/or body mass index (BMI) for an observational study in the period following initial diagnosis of T1D. In this hypothetical case, subjects are enrolled within 2 months of being diagnosed with T1D. After the enrollment visit, there are no study visits; vital signs data is collected at each healthcare visit. The researchers want to analyze changes in BMI at 30, 60, and 90 days following diagnosis.

The following example includes sample data along with dataset, variable, and value/parameter-level metadata. The source metadata for the variables and values are provided for illustrative purposes and are not intended to represent standard derivation logic. Within the Source/Deriviation/Comment column there is additional text intended to provide further discussion. This additional text would not be present in an actual Define-XML document. ADVSBMI is an example of an ADaM BDS dataset that could be used to support analysis of BMI and BMI percentile over time.

advsbmi.xpt

Rows 101-108:	Illustrate data for the baseline and 30-day timepoints for subject XYZ-001-001. (Day 60 and day 90 timepoints are not displayed; this would make the example expansive with little added value.) For this subject, VSSEQ is populated for BMI and BMIAPCTL because these were collected on the CRF. The CHG variable is populated for the 30-day timepoints.
Rows 109-116:	Illustrate data for the baseline and 30-day timepoints for subject XYZ-002-004. (Day 60 and day 90 timepoints are not displayed; this would make the example expansive with little added value.) For this subject, VSSEQ is not populated for BMI and BMIAPCTL because these were not collected on the CRF. The BMI was derived from the corresponding weight and height values as described in the Parameter Level Metadata for the BMI PARAMCD. The BMIAPCTL was derived using the BMI value along with the AGEMTPT and SEXN variable to identify the percentile from the US Centers for Disease Control and Prevention's BMI-for-age percentile growth charts.

Rows 101-108:

Illustrate data for the baseline and 30-day timepoints for subject XYZ-001-001. (Day 60 and day 90 timepoints are not displayed; this would make the example expansive with little added value.) For this subject, VSSEQ is populated for BMI and BMIAPCTL because these were collected on the CRF. The CHG variable is populated for the 30-day timepoints.

Rows 109-116:

Illustrate data for the baseline and 30-day timepoints for subject XYZ-002-004. (Day 60 and day 90 timepoints are not displayed; this would make the example expansive with little added value.) For this subject, VSSEQ is not populated for BMI and BMIAPCTL because these were not collected on the CRF. The BMI was derived from the corresponding weight and height values as described in the Parameter Level Metadata for the BMI PARAMCD. The BMIAPCTL was derived using the BMI value along with the AGEMTPT and SEXN variable to identify the percentile from the US Centers for Disease Control and Prevention's BMI-for-age percentile growth charts.

advsbmi.xpt

Row	STUDYID	USUBJID	AAGE	AAGEU	SEX	SEXN	BRTHDT	T1DDXDT	VSDTC	ADT	ATPT	ATPTN	AGEYTPT	AGEMTPT	PARAM	PARAMCD	AVAL	ABLFL	BASE	CHG	VSSEQ
101	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2018-12-27	27DEC2018	Baseline	0	11.49	137.92	Height (m)	HEIGHT	1.48	Y	1.48		3
102	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2018-12-27	27DEC2018	Baseline	0	11.49	137.92	Weight (kg)	WEIGHT	38.3	Y	38.3		1
103	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2018-12-27	27DEC2018	Baseline	0	11.49	137.92	Body Mass Index (kg/m2)	BMI	17.5	Y	17.5		5
104	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2018-12-27	27DEC2018	Baseline	0	11.49	137.92	BMI-for-Age Percentile	BMIAPCTL	52	Y	52		6
105	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2019-01-28	28JAN2019	30 Days	30	11.58	138.97	Height (m)	HEIGHT	1.49		1.48	0.01	13
106	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2019-01-28	28JAN2019	30 Days	30	11.58	138.97	Weight (kg)	WEIGHT	39.1		38.3	0.8	11
107	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2019-01-28	28JAN2019	30 Days	30	11.58	138.97	Body Mass Index (kg/m2)	BMI	17.6		17.5	0.1	15
108	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2019-01-28	28JAN2019	30 Days	30	11.58	138.97	BMI-for-Age Percentile	BMIAPCTL	50		52	-2	16
109	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2019-02-27	27FEB2019	60 Days	60	11.66	139.96	Height (m)	HEIGHT	1.49		1.48	0.01	23
110	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2019-02-27	27FEB2019	60 Days	60	11.66	139.96	Weight (kg)	WEIGHT	40.5		38.3	2.2	21
111	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2019-02-27	27FEB2019	60 Days	60	11.66	139.96	Body Mass Index (kg/m2)	BMI	18.2		17.5	0.7	25
112	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2019-02-27	27FEB2019	60 Days	60	11.66	139.96	BMI-for-Age Percentile	BMIAPCTL	47		52	-5	26
113	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2019-03-30	30MAR2019	90 Days	90	11.75	140.98	Height (m)	HEIGHT	1.50		1.48	0.02	33
114	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2019-03-30	30MAR2019	90 Days	90	11.75	140.98	Weight (kg)	WEIGHT	42.1		38.3	3.8	31
115	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2019-03-30	30MAR2019	90 Days	90	11.75	140.98	Body Mass Index (kg/m2)	BMI	18.7		17.5	1.2	35
116	TD011	001-001	11.49	YEARS	M	1	30JUN2007	12DEC2018	2019-03-30	30MAR2019	90 Days	90	11.75	140.98	BMI-for-Age Percentile	BMIAPCTL	42		52	-10	36

ADVSBMI Dataset Metadata

Dataset Name	Dataset Description	Dataset Location	Dataset Structure	Keys	Class	Documentation
ADVSBMI	BMI Analysis Dataset	advsbmi.xpt	One record per subject per parameter per analysis timepoint	STUDYID, USUBJID, PARAMCD, ATPT	BASIC DATA STRUCTURE	XYZ Protocol and SAP

ADVSBMI Variable Metadata

Variable Name	Variable Label	Type	Length/Display Format	Codelist/Controlled Terms	Source/Derivation/Comment
STUDYID	Study Identifier	text	$5		ADSL.STUDYID
USUBJID	Unique Subject Identifier	text	$7		ADSL.USUBJID
AAGE	Analysis Age	float	(6.2)		ADSL.AAGE
AAGEU	Analysis Age Units	text			ADSL.AAGEU
SEX	Sex	text		(SEX)	ADSL.SEX
SEXN	Sex (N)	integer		1, 2	ADSL.SEXN
BRTHDT	Date of Birth	integer	date9.		ADSL.BRTHDT
T1DDXDT	Date of Type 1 Diabetes Diagnosis	integer	date9.		ADSL.T1DDXDT
VSDTC	Date/Time of Measurements	text	ISO 8601		VS.VSDTC
ADT	Analysis Date	integer	date9.		Numeric version of date portion of VSDTC
ATPT	Analysis Timepoint	text		Baseline 30 Days 60 Days 90 Days	If ADT <= ADSL.RFICDT, set to "Baseline"; else if ADSL.RFICDT < ADT <= ADSL.T1DDXDT + 45, set to "30 days"; else if ADSL.T1DDXDT + 45 < ADT < = ADSL.T1DDXDT+ 75, set to "60 days"; else if ADSL.T1DDXDT+ 75 < ADT <=ADSL.T1DDXDT+ 105, else set to "90 days"; otherwise, set to missing.
ATPTN	Analysis Timepoint Number	integer		0 (= Baseline) 30 (= 30 Days) 60 (= 60 Days) 90 (= 90 Days)	Assigned based on value of ATPT. Provided to facilitate plotting.
AGEYTPT	Age at Analysis Timepoint (years)	float	(6.2)		Derived: Age in years (in 2 decimal points) calculated between BRTHDT and ADT as ROUND(((ADT – BRTHDT) / 365.25),0.01).
AGEMTPT	Age at Analysis Timepoint (months)	float	(6.2)		Derived: Age in months (in 2 decimal points) calculated between BRTHDT and ADT as ROUND(12*((ADT – BRTHDT) / 365.25),0.01).
PARAM	Parameter	text		Height (m), Weight (kg), Body Mass Index (kg/m2), BMI-for-Age Percentile	Assign based on PARAMCD using controlled terminology.
PARAMCD	Parameter Code	text		HEIGHT, WEIGHT, BMI, BMIAPCTL	Set PARAMCD=VS.VSTESTCD where VSTESTCD in ("HEIGHT, WEIGHT, BMI, BMIAPCTL")
AVAL	Analysis Value	float			See Parameter-Level Metadata.
ABLFL	Baseline Record Flag	text		Y	Set to "Y" for the record with ADT closest and prior to T1DDXDT with a non-missing AVAL.
BASE	Baseline Value	float			AVAL where ABLFL="Y"
CHG	Change from Baseline	float			AVAL - BASE
VSSEQ	Sequence Number	integer			When the record is based on a single record in VS, set to VS.VSEQ where VS.VSTESTCD= PARAMCD and VS.VSDTC = ADT. Otherwise missing.

Parameter Value List - ADVSBMI [AVAL]

Variable	Where	Type	Length/DisplayFormat	Source/Derivation Comment
AVAL	PARAMCD in ("WEIGHT","HEIGHT")	float	(6.2)	Set to VS.VSTRESN where VS.VSTESTCD = PARAMCD
AVAL	PARAMCD="BMI"	float	(5.1)	After deriving ADaM data for parameters HEIGHT and WEIGHT from SDTM records of a given VS.VSDTC, check for a VS.VSTESTCD="BMI" at the same VS.VSDTC. If there exists a record in VS with VS.VSTESTCD="BMI" for the same VSDTC as parameters HEIGHT and WEIGHT and VS.VSSTRESN is non-missing, set AVAL = VS.VSSTRESN else, if AVAL is non-missing for both parameters HEIGHT and WEIGHT for the same VS.VSDTC, set AVAL = (the value of AVAL for PARAMCD = "WEIGHT" for the same VSDTC)/(the square of the value of AVAL for PARAMCD="HEIGHT" at the same VSDTC) Otherwise, missing.
AVAL	PARAMCD="BMIAPCTL"	integer	(2.0)	If there is a VS record where VS.VSTESTCD = "BMIAPCTL" and VS.VSDTC = ADT, set AVAL to VS.VSSTRESN; else use the AGEMTPT, SEXN and BMI to look up the BMIAPCTL value in the available reference tables (may be captured in the protocol, SAP, or external references) that define the percentile value associated with the combination of variable values.