• Component Description
• Eligible Sample
• Interview Setting and Mode of Administration
• Quality Assurance & Quality Control
• Data Processing and Editing
• Analytic Notes
• Codebook
  • SEQN - Respondent sequence number
  • DSD010 - Any Dietary Supplements taken?
  • DSD010AN - Any Antacids Taken?
  • DSDCOUNT - Total # of Dietary Supplements taken
  • DSDANCNT - Total # of Antacids Taken
• Appendix 1: Handcard DSQ1
• Appendix 2: Vitamins/Minerals on the "Strength Only" List
• Appendix 3: Created Default Supplements and Antacids
• Appendix 4: Data File Structure and Relationships

National Health and Nutrition Examination Survey

2003-2004 Data Documentation, Codebook, and Frequencies

Dietary Supplement Use 30-Day - File 1, Supplement Counts (DSQ1_C)

Data File: DSQ1_C.xpt

First Published: January 2007

Last Revised: March 2007

Component Description

The Dietary Supplements section (DSQ) provides personal interview data on use of vitamins, minerals, herbals and other dietary supplements. This section also includes supplement name, ingredients, amounts, and serving size.

Eligible Sample

All sample persons (SP). Survey participants over 16 years of age answer for themselves; proxy respondents answer for participants 16 years or younger.

Interview Setting and Mode of Administration

In-person, home interview.

Quality Assurance & Quality Control

Data were routinely examined for discrepancies and erroneous entries. Trained nutritionists reviewed incoming data and matched reported supplements to known supplements from the NHANES database, where possible; sought additional supplement labels if feasible; assigned generic or default supplements as appropriate; transferred or removed inappropriate products; and assigned matching codes (confidence codes).

 

Data Processing and Editing

For exact questions participants are asked, please refer to DSQ codebook. During the household interview, survey participants are asked if they have taken a vitamin, mineral, herbal or any other dietary supplement in the past month. Participants are shown a hand card that lists examples of dietary supplements (Appendix 1: Handcard DSQ1). Those who answer “yes” are asked to show the interviewer the dietary supplement containers of all the products used. For each dietary supplement reported, the interviewer enters the product’s complete name from the container into a computer. If no container is available, the interviewer asks the participant to verbally report the name of the dietary supplement. The interviewer then tries and finds the supplement on a list created by NCHS. If an exact match of the dietary supplement cannot be found, the interviewer is instructed to select “product not on list” from the list. Interviewers can record up to 20 dietary supplements. The manufacturer is either selected from the manufacturer list or entered manually. The same process is also repeated for non-prescription antacids.

Participants are asked how long they have been taking the dietary supplement, how many times in the past 30 days have they taken it and how much do they take daily.

Some prescription supplements are mistakenly recorded in the prescription medicine section rather than the dietary supplement section. Strength and details of the persons’ usage are missing for these supplements. All supplements, including prescription supplements are released in the DSQ section only, not in the medication (RXQ) section.

Information on use of non-prescription antacids was sometimes recorded in the dietary supplement section; other times in the antacid sub-section of the medication section (RXQ). Because of their nutrient content, antacids that contain calcium or magnesium are released with the DSQ data, irrespective of where they were reported. Only these antacids are reported; this is not a complete accounting of all antacids. Data users should also be aware that the section in which an antacid is recorded (e.g. DSQ or RXQ) may not accurately represent the usage of that antacid. For example, some antacids reported with the RXQ (medications) section may have been used as a dietary supplement or vice versa, and some antacids may be used as both medications and calcium supplements. Thus, users are cautioned that analyses of these data to estimate the percentage of antacids used as dietary supplements would not be appropriate.

Matching a reported supplement to known supplement:
Trained nutritionists at NCHS match the product names entered or selected by the interviewer (including prescription supplements and antacids containing calcium or magnesium) to a known product when possible. These matches are made with varying degrees of precision, and a matching code (DSDMATCH) accompanies each match (Appendix 2: Matching Codes). Briefly, the codes are: 1) Exact or near exact match; 2) Probable match; 3) Generic match; 4); Reasonable match; and 5) Default match. In some cases no match can be made with any certainty. These products are coded 6) No match. Products whose names were reported as “Refused” (7777) or “Don’t know” (9999) have matching codes of 7 and 9, respectively.

NCHS communicates with many major manufacturing company representatives to determine when various product re-formulations become available. We match reported with known supplements based upon reported supplement name as entered by the interviewer and the product version on the market at that time, if known. Based upon manufacturer advice, we have used a lag time of 5 months after new product market entry in matching recorded products to these new products. Despite these precautions, there is no guarantee that the product taken was not an older or newer product than the one to which it was matched.

NCHS created generic and default dietary supplements and entered these into the supplement database. Reported supplements for which the strength of all ingredients were known were matched to a generic supplement, i.e. one which had no brand name. These were generally single ingredient supplements which included a strength (e.g. vitamin C 500 mg) or multiple vitamins and/or mineral supplements made by a private label manufacturer that was known to us and for which we had a label with identical ingredients and strengths (e.g. brand X all-purpose multivitamin was reported, and we had a label for brand Y, made by the same manufacturer). These matches are coded as 3. When all ingredient strengths were not known, the supplement was matched to a default supplement where possible. Defaults were created for single ingredient and multiple ingredient supplements based on our own data of most frequently reported supplements of that type. These matches were coded as 5. Created default and generic products and the actual products or strengths that were assigned to these defaults are listed in Appendix 3 (Assigned default supplements and antacids).

New variables added to File 1:
DSDANCNT – total number of non-prescription antacids taken.
DSD010AN – Any non-prescription antacids taken?
Antacids have been collected since 1999. DSDANTA variable designates whether a product is an antacid and whether it was recorded as a dietary supplement or in the antacid section.

The NCHS Dietary Supplements Database
NCHS maintains a dietary supplement database. This is not a comprehensive dietary supplements database and only contains dietary supplements and calcium and/or magnesium-containing antacids that have been reported in the NHANES survey since 1999. The ingredient information entered into the database is taken directly from the supplement facts box on the label or carton, if available, or the equivalent from other sources. NCHS nutritionist enter from the label: supplement name; manufacturer; serving size; form of serving size; ingredients and amounts, including an operator to indicate that the amount is less than, more than or equal to the amount.

NCHS does not verify the actual composition of supplements reported: the database is based on label information, not testing. The best information source is the label itself, but when this cannot be obtained, other sources are used. Information from other sources may not always be an accurate reflection of what is actually on the supplement label. This is true for the supplement’s apparent name as well as for the ingredients. The apparent name on the container is most important, since interviewers see the supplement container and record the name as it appears to them. Differences from what appears on the label are particularly noted for information from the Internet (name and ingredients), the PDR (name), and supplement carton (name). In addition, supplements may change the appearance of a label and thus the apparent name without changing the content or may change content with minimal change to the label, or may change both. NCHS attempts to obtain updated labels as they come onto the market, but cannot guarantee complete success. The source of the supplement information (DSDSRCE) is included in the data release.

Ingredients may be listed in various ways, for example: in their elemental form (e.g. calcium); as compounds (e.g. calcium citrate); as plant forms (e.g. extract, concentrate, oil) or plant parts (e.g. root, leaf); as percentages (e.g. beta carotene- % of vitamin A); and as blends or complexes (e.g. bioflavonoid complex; proprietary blend). Suggestions for conversion of compounds are contained in Dietary Supplement Database documentation. These are based upon literature searches, but may contain errors. Users should verify these conversions. Note especially that there are a number of conversions for ferrous sulfate depending upon its form, which was not always specified.

Some supplements contain proprietary blends of ingredients, generally non-nutrients. An amount is specified for the blend but not individual ingredients in the blend. In such cases, the blend and its amount are entered into the database, but the individual ingredients of the blend are listed without any amounts. A few supplement labels list ingredients but no amounts at all, so the amounts are missing.

Product information is released from the in-house Dietary Supplement Database as the NHANES Dietary Supplement Database (NHANES-DSD) which contains detailed information on the dietary supplements and antacids reported by survey participants since NHANES 1999. The NHANES-DSD release consists of three datasets which contain information on products:

Dietary Supplement Product Information (DSPI)
Dietary Supplement Ingredient Information (DSII)
Dietary Supplement Blend Information (DSBI)

The supplement ID numbers (DSDSUPID) located in the Individual Dietary Supplements File can be used to merge with the NHANES-DSD files. For more information on the NHANES-DSD, please refer to the documentation and release files located on the NHANES website.

Data Editing

When a variable was modified globally, as part of the editing process, the third letter in the variable name was changed from a Q (i.e. DSQ) to a D (i.e. DSD). Variables that were obtained or generated from an external database (i.e. dietary supplement database) will also have the letter “D” in the third position of the variable name.

Specific Variables and edits

DSDANTA is a created variable that indicated whether an antacid was reported in the DSQ section or the Antacid section.
0= Non-antacid supplement
1= Antacid reported with dietary supplements (DSQ)
2= Antacids reported with medications (RXQ), reported in Antacid section

Note: For a few participants, the same antacid was recorded in both questionnaire sections. In these instances, the antacid was considered to be in the DSQ section and coded as 1.

DSDMTCH: (matching code): confidence codes (see Appendix 2). Supplements are recorded in the household interview with varying degrees of accuracy and completeness, so we have a system to of how certain that we are matching a supplement recorded during the interview with the actual supplement label. Briefly, the codes are: 1) Exact or near exact match; 2) Probable match; 3) Generic match; 4); Reasonable match; and 5) Default match. In some cases no match can be made with any certainty. These products are coded 6) No match. Products whose names were reported as “Refused” (7777) or “Don’t know” (9999) have matching codes of 7 and 9, respectively.

Note: Analysts should be aware that for default matches and matches that chose between several similarly named supplements, there is less certainty that the ingredients and ingredient amounts in the supplement assigned exactly match those in the supplement actually taken. Additionally, NCHS cannot guarantee in any case that the matched product was the exact product taken or even that any product actually was taken, as these data are self-reported.

DSDCOUNT: (Total number of supplements taken).
This is the count given to all supplements reported in the DSQ or RXQ section, as well as antacids that were reported in the DSQ section(DSDANTA=1).

Antacids recorded in the antacid section to not contribute to this count (DSDANTA=2).

Non-supplements that were reported in the DSQ section do not count.
Products whose names were entered as 7777 or 9999, that is, don’t know or refused, are counted as supplements, since there is no clear evidence that they are not and they were reported by the participant as a supplement.

DSD010 (derived variable) that indicates any dietary supplement use.

Participants who took a product coded as a supplement in the last 30 days (DSDCOUNT ≥1) are coded as 1. Supplements that were mistakenly recorded as RXQ antacid or prescription medicine are moved to the DSQ section and are counted as supplements.

Participants who reported an antacid containing calcium or magnesium in the last 30 days that was recorded only in RXQ antacid section (DSDANTA=2), and did not take any dietary supplement are coded as 2.
Participants who did not take any product determined to be a dietary supplement in the last 30 days (DSDCOUNT =0) are coded as 2. Prescription medicines and analgesics that are mistakenly recorded in the DSQ section are moved to their appropriate sections for data release.

Note: If a respondent “refused” or answered “don’t know” to DSQ010, the codes 7777 or 9999, respectively, are used in both DSD010 and DSDCOUNT and all other variables are coded as missing. If a respondent “refused” or answered “don’t know” to a supplement name, the name of the variable is 7777 or 9999 and matching codes are 7 and 9, respectively. Other information on use of the supplement given by the participant is retained.

DSDANCNT – total number of non-prescription antacids taken. Count of non-prescription antacids taken per respondent.

DSD010AN – Any non-prescription antacids taken? Indicates if any non-prescription antacids taken.

DSD070 denotes if the dietary supplement container was seen during the household interview. Containers are seen approximately 88% of the time. This enables interviewers to collect more complete and accurate supplement names compared to supplement containers that are not seen. For containers that are not seen the interviewer relies on the participants’ memory for the name of the supplement taken.

Note: A more precise name for a supplement can be recorded by the interviewer, and thus a more precise match to a known supplement can be made, when the interviewer sees the supplement container rather than recording the participant’s report of the supplement name (for example, multivitamin/multiminerals are often reported as multivitamins). In general, this is reflected in the matching code, but analysts should be aware that precision is greater when a container is seen.

DSD090 (How Long Supplement Taken) is converted to days for release by multiplying years by 365; months by 30.4; and weeks by 7.

DSD103 (Days Supplement Taken, Past 30 Days) is released as recorded by interviewer. This information is missing for dietary supplement data that was recorded in the RXQ medication section.

DSD122Q/U (Quantity and dosage form of supplement taken daily) is released as recorded by interviewer. This information is missing for dietary supplement data that was recorded in the RXQ medication section.

 

Analytic Notes

The 2-year sample weights (WTINT2YR, WTMEC2YR) should be used for all NHANES 2003–2004 analyses. There are no 4-year weights in this file. The 4-year weights were provided with the NHANES 2001–2002 release file because there were some transition issues related to the use of 1990 Census and 2000 Census information. Detailed instructions for linking earlier datasets (1999–2000 and 2001–2002) are provided in the NHANES Analytic Guidelines. Analysts are encouraged to review the NHANES Analytic Guidelines provided with the data release files to determine the appropriate analytic methodology.

Using self-reported data

NHANES data are self-reported and recorded by interviewers, and thus may contain inconsistencies or errors. Unless a data entry is clearly mis-recorded and the correct response can be confidently identified, possible errors or inconsistencies are not edited. Users will note that some records may indicate a person had been taking a supplement for a short time, e.g. only one day (DSD090), but report a frequency of taking it that is longer (DSD100), e.g. twice a week. Such apparent inconsistencies are not edited, as there is no obvious basis for editing. Users are advised to assess the data and edit it as deemed appropriate for the analyses being undertaken.

Deriving nutrient estimates from dietary supplement data
Dietary supplement and antacid intake data (DSQ) refer to the past 30 days. The data necessary for computation of total nutrient intake are: personal usage data (DSQ files 1& 2); dietary supplement composition data (DSQ files DSBI, DSPI, DSII)

The analyst must combine these files: File 1 portrays a participant’s overall use of supplements; file 2 refers to the person’s usage of one particular supplement; and files DSBI, DSPI, and DSII refer to the content of one specific supplement.

DSQ codebooks, documentation, and data are described in the questionnaire section of this data release.

Dietary supplement data were reported as times per month in 1999-2000 and days in the past month in 2001-2002. For participants who took one supplement each day, one time per day, daily nutrient intake from supplements can be estimated directly from the label ingredients.

Intake of multiple supplements with the same nutrient(s), multiple use of the same supplement on the same day, and nutrients in blends must be taken into account in nutrient calculations. Nutrient names and the quantity units need to be synchronized and nutrient amounts from all these calculations must then be summed. Some nutrient amounts are for a nutrient compound (generally a foreign-made product or an antacid) and these must be converted to a nutrient amount.

If one or more supplements were taken less than daily, estimation of supplement nutrient intake could be calculated for a month, a daily average, highest possible daily intake, or lowest non-zero intake.

Combining nutrient estimates from dietary supplement plus food- recall data
To estimate total dietary nutrient intake, nutrients from diet, supplements, and antacids should be combined. Because of different data collection, referent time periods, and release systems, these data require some manipulation and assumptions to combine. Consequently, the resultant total may not provide a very accurate estimate of daily total nutrient intake nor is a variance estimate directly available. This would need to be calculated using special programs developed for this purpose (e.g. C-Side).

Deriving nutrient quantity from dietary supplement requires extracting the nutrient content of each supplement, as described above. Dietary recall data (DRX) are released as total combined nutrient amount from all foods reported on the one day dietary recall; no derivation is needed (e.g. calcium quantities from all food sources are combined and released as a single one day quantity). RXQ codebooks, documentation, and data are located in the exam section of the data release.

Because of the different time referents for dietary supplement and food intake, an analyst must decide upon the most useful way to combine these data to estimate intake for a given purpose. For example, if supplement use was reported to be daily, the nutrient value from the supplement could be simply added to the recall nutrient value, although the time frames do not overlap. When some or all supplements were taken less than daily (1999-2000) or less than 30 times a month (2001-2002), the nature of the analysis and the data itself should guide the decision about nutrient calculation.

For instance, if a supplement were taken nearly monthly, an analyst could:
1) assume intake to be daily;
2) calculate monthly intake based on intake frequency;
3) calculate average daily value from the monthly value; or
4) analyze data only from people who took supplements daily

The latter would mean that the sample would no longer be representative of all supplement users, but should represent daily supplement users.

Because of the data differences, combining nutrient intake from these different variables to estimate total nutrient intake requires thoughtful consideration of the analytic goal and methods, and deserves accurate description of methods, assumptions, and weaknesses in any presentation of results.

Codebook and Frequencies