ࡱ > g @ bjbjVV r< r< 8 1 * * * * * 4 ^ ^ ^ h D
$ ^ x . s u u u u u u H" $ N u * u * * * * s A G"N M (
0 , H% u | H% A A H% * Q x u u H% : Web-Only Appendix Criteria applied to 24-hour urine samples in the HFUS
A.Table 1: Unweighted Demographic Characteristics Before and After Application of Exclusion CriteriaThis appendix describes the methods undertaken by the investigators to address and correct for incomplete urine collections in the Heart Follow-Up Study (HFUS) both pre and post hoc. Undercollection in 24-hour urine samples is a well-established limitation of this measurement method. A limitation of prior investigations using 24-hour samples is the lack of thorough explanation and disclosure of methods used to assess completeness. The absence of this detail contributes to confusion in interpretation of results and comparison across studies and provides little guidance to researchers conducting similar studies.
The study protocol included criteria designed to identify undercollection during the implementation stage. These criteria were developed based upon review of existing literature, review of protocols requested from investigators of prior landmark studies using 24-hour urine collection , and in consultations with experts. The following criteria were determined a priori and applied during study implementation were as follows:
The first criteria for assessing completeness of the sample involved elements that were built into the 24-hour collection protocol. If a urine collection was <05L (500 ml) or if the participant reported having included the first void of the day in the collection, the participant was asked to redo the collection. Previous studies ADDIN EN.CITE Elliott198815771577157717Elliott, P.Stamler, R.Northwestern University Medical School, Chicago, IL 60611.Manual of operations for "INTERSALT", an international cooperative study on the relation of sodium and potassium to blood pressureControl Clin TrialsControl Clin Trials1S-117S92 Suppl1988/06/01AdultAppointments and SchedulesBlood Pressure/*drug effectsBlood Pressure Determination/standardsClinical ProtocolsClinical Trials as Topic/*methodsFemaleForms and Records Control/standardsHumansMale*Manuals as TopicMiddle AgedPotassium/*pharmacologyQuality ControlSampling StudiesSodium/*pharmacologySpecimen Handling/methods1988Jun0197-2456 (Print)
0197-2456 (Linking)3396367http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=3396367eng1 had also asked participant to redo a collection if hours spanned beyond the collection period (22-26 hours acceptable) or if participant report of missing a urine void. Because the HFUS was population-based and we had limited contact with the participants, we wanted to minimize the samples that we rejected and the number of redos requested. If the participant fell outside of the 22-26 hour range, they were asked but not required to redo the collection. If a participant missed a void, the urine was collected but it was also noted in the record to have missed a void (ultimately, we excluded these participants from our analysis).
The second criteria that was applied related to the laboratory analysis. Researchers at the laboratory assessed the volume of urine and sodium levels and noted any participants where these values seemed incongruous. No samples were flagged at this step.
The third criteria was to adjust all laboratory values to 24-hours. For example if a person had a 23-hour collection, their sodium excretion was multiplied by 24/23. There were n=1775 participants with a 24-hour urine analysis at the start of this investigation. Participants that did not have a collection time were removed from the analysis (n=3) because lab values normalized to 24 hours could not be obtained. Thus, the overall sample at the start of analysis was n=1772 participants.
In May of 2011, a number of high profile articles related to sodium intake stimulated scientific discussion regarding the exclusion criteria used during analysis in studies using 24-hour sodium collection methodology. ADDIN EN.CITE ADDIN EN.CITE.DATA 2-7 Responsive to that concern, we undertook further investigation of relevant literature and engaged in active related discussions with colleagues outside of our institution. We refined the criteria planned for use during the analysis phase to define a complete sample and applied those below.
Normalized urine volume <500 ml, n=16. A urine volume of <500 ml without having missed a urine void is biologically not possible unless the individual is chronically ill. This criteria had previously been used in the INTERSALT study to define a complete 24-hour collection, the largest population-based study of 24-hour urines to date. ADDIN EN.CITE Elliott198815771577157717Elliott, P.Stamler, R.Northwestern University Medical School, Chicago, IL 60611.Manual of operations for "INTERSALT", an international cooperative study on the relation of sodium and potassium to blood pressureControl Clin TrialsControl Clin Trials1S-117S92 Suppl1988/06/01AdultAppointments and SchedulesBlood Pressure/*drug effectsBlood Pressure Determination/standardsClinical ProtocolsClinical Trials as Topic/*methodsFemaleForms and Records Control/standardsHumansMale*Manuals as TopicMiddle AgedPotassium/*pharmacologyQuality ControlSampling StudiesSodium/*pharmacologySpecimen Handling/methods1988Jun0197-2456 (Print)
0197-2456 (Linking)3396367http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=3396367eng1
Self-report missing one urine void, n=55. Self-report of having missed a urine void was also previously used in the INTERSALT study and unequivocally suggests the participants 24-hour urine collection is not complete.
Sex-specific creatinine cutoff: Males <605 mmol, Females <378 mmol creatinine (3 standard deviations below the U.S. mean in INTERMAP study) ADDIN EN.CITE INTERMAP Co-Operative Research Group15951595159526INTERMAP Co-Operative Research Group, 8, n=50. We established the cut-points for creatinine excretion through an internal working group process. This group reviewed published studies ADDIN EN.CITE ADDIN EN.CITE.DATA 1, 9, 10 and invited input from external experts to define our study cutpoints. ADDIN EN.CITE INTERMAP Co-Operative Research Group15951595159526INTERMAP Co-Operative Research Group, 8 Creatinine is a byproduct of muscle breakdown and is therefore 1) excreted at a constant rate and 2) expected to be lower in individuals with less muscle mass (e.g. elderly, females vs. males). Low creatinine was defined using the U.S. creatinine distribution from the INTERMAP study. ADDIN EN.CITE 201118691869186926International Population Study on Macronutrients and BP (INTERMAP) Team - Personal Communication 6/17/1120116/17/118 The study sites for INTERMAP were distributed across the U.S., and included a broad range of participants. Total sample size of the U.S. sample was n=2195. Sex-specific cutoffs were determined as females should have a lower cutoff than males owing to their lower muscle mass. The cutoffs were defined as three standard deviations below the mean and were: for Males <605 mmol creatinine and for Females<378 mmol creatinine. This cutoff therefore addresses the biological differences in body size between men and women and excludes those participants who were at the very bottom of the distribution of the U.S., who have most likely undercollected.
In total, this approach excluded 116 individuals. This total is lower than from the straight sum of individual affected by each criteria because some individuals met more than one of the criteria.
Application of these criteria resulted in a study sample that was similar in demographic characteristics to the overall sample, the implication being the participants that were being excluded were at random and were true undercollectors. There is little existing evidence that undercollection is more prevalent in specific population subgroups. The unweighted breakdown of the overall sample of n=1772 and the resulting sample of n=1656 are shown (Table 1). In both samples, participants were predominantly ages 25-64; female; non-white; and overweight or obese. Mean volume was slightly higher after exclusion criteria were applied.
By considering previous methods used to assess completion and by consulting with other experts in the field, we developed a robust set of completeness criteria that takes into account survey data and biometric information. The process of developing the completeness criteria are described here as model for future published research and to inform the ongoing discussions about the most appropriate approach for determining undercollection.
A.Table 1. Unweighted Demographic Characteristics Before and After Application of Exclusion Criteria
Total (n=1772)After Application of Exclusion Criteria (n=1656)Age Groupn%n%18-24112631026225-4454130550430545-6473741670142465+380216348210SexMale742418692418Female1030582964582RaceNon-Hispanic White642362612370Non-Hispanic Black482272440266Hispanic521294485293Non-Hispanic Asian90518451Other37213521BMIUnderweight, <18541233924Normal, 185-<25560317521315Overweight, 25-<30598338566343Obese, 30+568321526318MeansnMeannMeanHeight (meters)17681671653167Weight (lbs)17677921652792Volume (ml)177215381165615637
References
ADDIN EN.REFLIST 1. Elliott P, Stamler R. Manual of operations for "INTERSALT", an international cooperative study on the relation of sodium and potassium to blood pressure. Control Clin Trials. Jun 1988;9(2 Suppl):1S-117S.
2. Aleksandrova K, Pischon T, Weikert C. Urinary sodium excretion and cardiovascular disease mortality. JAMA. Sep 14;306(10):1083; author reply 1086-1087.
3. Bochud M, Guessous I, Bovet P. Urinary sodium excretion and cardiovascular disease mortality. JAMA. Sep 14;306(10):1084; author reply 1086-1087.
4. Cook NR. Urinary sodium excretion and cardiovascular disease mortality. JAMA. Sep 14;306(10):1085; author reply 1086-1087.
5. de Abreu-Silva EO, Marcadenti A. Urinary sodium excretion and cardiovascular disease mortality. JAMA. Sep 14;306(10):1085-1086; author reply 1086-1087.
6. Labarthe DR, Briss PA. Urinary sodium excretion and cardiovascular disease mortality. JAMA. Sep 14;306(10):1084-1085; author reply 1086-1087.
7. Rebholz CM, He J. Urinary sodium excretion and cardiovascular disease mortality. JAMA. Sep 14;306(10):1083-1084; author reply 1086-1087.
8. Elliott P, Brown I. INTERMAP Co-Operative Research Group, Distributions of creatinine in U.S. males and females in INTERMAP. Personal Communication.2011.
9. Stamler J, Elliott P, Dennis B, et al. INTERMAP: background, aims, design, methods, and descriptive statistics (nondietary). Journal of Human Hypertension. Sep 2003;17(9):591-608.
10. Reinivuo H, Valsta LM, Laatikainen T, Tuomilehto J, Pietinen P. Sodium in the Finnish diet: II trends in dietary sodium intake and comparison between intake and 24-h excretion of sodium. Eur J Clin Nutr. Oct 2006;60(10):1160-1167.
PAGE \* MERGEFORMAT 4
J V U X ] e n
o
T s ƶƦƦƦƚƚzƚggQg +j h? hW CJ OJ QJ UaJ %j h? hW CJ OJ QJ UaJ hv hW CJ H*OJ QJ aJ j hW CJ OJ QJ UaJ hW CJ OJ QJ aJ h? hW 6CJ OJ QJ aJ h? hW 5CJ OJ QJ aJ h? hW CJ OJ QJ aJ hW 5CJ OJ QJ aJ h? hW 5CJ OJ QJ aJ hW 5CJ OJ QJ aJ J K
' % &