 |
 |
 |
|
   |
||||||||
|
|
|||||||
|
Scientific Data Documentation
Electrocardiograms, Ages 20 - 74 years (1982-1984)
DSN: CC37.HSPHANES.EKG ABSTRACT General Information MEASUREMENT AND INTERPRETATION OF ELECTROCARDIOGRAMS Tape Number 6540 Ages 20 Years - 74 Years Version 1 December 1992 The Hispanic Health and Nutrition Examination Survey (HHANES) was conducted from July 1982 through December 1984. The data on the tape documented here are from all three portions of the survey: Mexican Americans Residing in selected counties of Texas, Colorado, New Mexico, Arizona and California Surveyed from July 1982 through November 1983 9,894 persons sampled; 8,554 interviewed; 7,462 examined Cuban Americans Residing in Dade County (Miami), Florida Surveyed from January 1984 through April 1984 2,244 persons sampled; 1,766 interviewed; 1,357 examined Puerto Ricans Residing in the New York City area, including parts of New Jersey and Connecticut Surveyed from May 1984 through December 1984 3,786 persons sampled; 3,369 interviewed; 2,834 examined The following tape characteristics are those of the version of the tape kept at NCHS and of the tape transmitted to the National Technical Information Service for release to users: Tape labels: IBM standard Data set name: HHANES.DU654001 Data set organization: Physical sequential Record format: Fixed block Record length: 1017 Block size: 24408 Number of records: 5815 Data code: EBCDIC This Public Use Data Tape has been edited very carefully. Numerous consistency and other checks were also performed. Nevertheless, due especially to the large number of data items, some errors may have gone undetected. Please bring to the attention of NCHS any errors in the data tape or the documentation. Errata sheets will be sent to people who have purchased the data tapes and corrections will be made to subsequently released data tapes. In publications, please acknowledge NCHS as the original data source. The acknowledgment should include a disclaimer crediting the authors for analyses, interpretations, and conclusions; NCHS should be cited as being responsible for only the collection and processing of the data. In addition, NCHS requests that the acronym HHANES be placed in the abstracts of journal articles and other publications based on data from this survey in order to facilitate the retrieval of such materials through automated bibliographic searches. Please send reprints of journal articles and other publications that include data from this tape to NCHS. Division of Health Examination Statistics National Center for Health Statistics Presidential Building, Room 900 6525 Belcrest Road Hyattsville, MD 20782 Public Use Data Tapes for the Hispanic Health and Nutrition Examination Survey will be released through the National Technical Information Service (NTIS) as soon as the data have been edited, validated and documented. A list of NCHS Public Use Data Tapes that can be purchased from NTIS may be obtained by writing the Scientific and Technical Information Branch, NCHS. Scientific and Technical Information Branch National Center for Health Statistics Presidential Building, Room 1064 6525 Belcrest Road Hyattsville, MD 20782 301-436-8500Cautions for Use of Data Tape CAUTION: BEFORE USING THIS DATA TAPE, PLEASE READ THIS PAGE Read the accompanying description of the survey, "The Plan and Operation of the Hispanic Health and Nutrition Examination Survey", DHHS Publication No. (PHS) 85-1321 before conducting analyses of the data on this tape. Two aspects of HHANES, especially, should be taken into account when conducting any analyses: the sample weights and the complex survey design. Analyses should not be conducted on data combined from the three portions of the survey (Mexican-American, Cuban-American, Puerto Rican). HHANES is a survey of Hispanic households and some of the sample persons included on this tape are not of Hispanic origin. A detailed description of the data codes dealing with national origin or ancestry appears in the NOTES section of this document. Examine the range and frequency of values of a variable before conducting an analysis of data. The range may include unusual or unexpected values. The frequency counts may be useful to determine which analyses may be worthwhile. Language of interview, which may appear several places on this tape, can vary depending on the questionnaire (several used in the survey) and on whether the response was provided by the sample person or by a proxy. For some data items, reference is made to a note. The notes (in a separate section of this document) may be very important in data analyses. Attention to them is strongly urged. For some data items, the number of sample persons with a positive response is very small. In these instances, it may not be possible to produce a reliable population estimate.Introduction and Survey Description The National Center for Health Statistics (NCHS) collects, analyzes, and disseminates data on the health status of Americans. The results of surveys, analyses, and studies are made known primarily through publications and the release of computer data tapes. This document contains details required to guide programmers, statistical analysts, and research scientists in the use of a Public Use Data Tape. From 1960 through 1980 NCHS conducted five population-based, national health examination surveys. Each survey involved collecting data by direct physical examination, the taking of a medical history, and laboratory and clinical tests and measurements. Questionnaires and examination components have been designed to obtain and support analyses of data on certain targeted conditions such as diabetes, hypertension, and anemia. Beginning with the first National Health and Nutrition Examination Survey (NHANES I) a nutrition component was added to obtain information on nutritional status and dietary practices. The numbers of Hispanics in these samples were, however, insufficient to enable adequate estimation of their health conditions. From 1982 through 1984 a Hispanic Health and Nutrition Examination Survey (HHANES) was conducted to obtain data on the health and nutritional status of three Hispanic groups: Mexican Americans from Texas, Colorado, New Mexico, Arizona and California; Cuban Americans from Dade County, Florida; and Puerto Ricans from the New York City area, including parts of New Jersey and Connecticut. The general structure of the HHANES sample design was similar to that of the previous National Health and Nutrition Examination Surveys. All of these studies have used complex, multistage, stratified, clustered samples of defined populations. The major difference between HHANES and the previous surveys is that HHANES was a survey of three special subgroups of the population in selected areas of the United States rather than a national probability sample. A detailed presentation of the design specifications is found in Chapter 5 of "Plan and Operation of the Hispanic Health and Nutrition Examination Survey, 1982-84" (Ref. No. 1). Data collection began with a household interview. Several questionnaires were administered: A Household Screener Questionnaire (HSQ), administered at each selected address, for determining household eligibility and for selecting sample persons. A Family Questionnaire (FQ), administered once for each family containing sample persons, which included sections on family relationships, basic demographic information for sample persons and head of family, Medicare and health insurance coverage, participation in income assistance programs, and housing characteristics. An Adult Sample Person Questionnaire (ASPQ), for persons 12 through 74 years which, depending on age, included sections on health status measures, health services utilization, smoking (20 through 74 years), meal program participation, and acculturation. Information on the use of medicines and vitamins in the past two weeks was also obtained. A Child Sample Person Questionnaire (CSPQ), for sample persons 6 months through 11 years which included sections on a number of health status issues, health care utilization, infant feeding practices, participation in meal programs, school attendance, and language use. Information on the use of medicines and vitamins in the past two weeks was also obtained. At the Mobile Examination Center two questionnaires were administered and an examination performed. An Adult Sample Person Supplement (ASPS), for sample persons 12 through 74 years, which included sections on alcohol consumption, drug abuse, depression, smoking (12 through 19 years), pesticide exposure, and reproductive history. A Dietary Questionnaire (DQ), for persons 6 months through 74 years, by which trained dietary interviewers collected information about "usual" consumption habits and dietary practices, and recorded foods consumed 24-hours prior to midnight of the interview. An examination which included a variety of tests and procedures. Age at interview and other factors determined which procedures were administered to which examinees. A dentist performed a dental examination and a vision test. Technicians took blood and urine specimens and administered a glucose tolerance test, X-rays, electrocardiograms, and ultrasonographs of the gall bladder. Technicians also performed hearing tests and took a variety of body measurements. A physician performed a medical examination focusing especially on the cardiovascular, gastrointestinal, neurological and musculoskeletal systems. The physician's impression of overall health, nutritional and weight status, and health care needs were also recorded. Some blood and urine specimen analyses were performed by technicians in the examination center; others were conducted under contract at various laboratories. Because the HHANES sample is not a simple random one, it is necessary to incorporate sample weights for proper analysis of the data. These sample weights are a composite of individual selection probabilities, adjustments for noncoverage and nonresponse, and poststratification adjustments. The HHANES sample weights, which are necessary for the calculation of point estimates, are located on all data tapes in positions 184-213. Because of the complex sample design and the ratio adjustments used to produce the sample weights, commonly used methods of point and variance estimation and hypothesis testing which assume simple random sampling may give misleading results. In order to provide users with the capability of estimating the complex sample variances in the HHANES data, Strata and Pseudo Primary Sampling Unit (PSU) codes have been provided on all data tapes in positions 214-217. These codes and the sample weights are necessary for the calculation of variances. There are computer programs available designed for variance estimation for complex sample designs. The balanced repeated replication approach (Ref. No. 2) is used in &REPERR and a linearization approach is used in &PSALMS to calculate variance-covariance matrixes. Both routines are available within the OSIRIS IV library (Ref. No. 3). SURREGR (Ref. No. 4) and SUPERCARP (Ref. No. 5) are programs that calculate variance-covariance matrixes using a linearization approach (Ref. No. 6) (Taylor series expansion). Another program, SESUDAAN (Ref. No. 7) calculates standard errors, variances, and design effects. (Note: This version of SESUDAAN should not be used to obtain variances for totals.) SURREGR and SESUDAAN are special procedures which run data under the SAS system (Ref. No. 8). Even though the total number of examined persons in this survey is quite large, subclass analyses can lead to estimates that are unstable, particularly estimates of variances. Con sequently, analyses of subclasses require that the user pay particular attention to the number of sample persons in the subclass and the number of PSU's that contain at least one sample person in the subclass. Small sample sizes, or a small number of PSU's used in the variance calculations, may produce unstable estimates of the variances. A more complete discussion of these issues and possible analytic strategies for examining various hypotheses is presented in Chapter 11 of "Plan and Operation of the Hispanic Health and Nutrition Examination Survey, 1982-84" (Ref. No. 1) and in an earlier NCHS methodology (Series 2) publication (Ref. No. 9). Some users, however, may not have access to the computer programs for estimating complex sample variances or may want to do their preliminary analyses without using them. In addition, variance estimates calculated from HHANES data through use of the programs described previously are likely to be unstable because there were so few sample areas for each portion of HHANES. This instability is not due to there being too few people in the sample but may be due to the fact that the sample was selected from relatively few areas. Therefore, the following discussion is designed to provide an alternative approach to deal with the unavailability of software and the small number of PSU's. The approach is based on using average design effects (Ref. No. 10). The design effect, defined as the ratio of the variance of a statistic from a complex sample to the variance of the same statistic from a simple random sample of the same size, that is, COMPLEX SAMPLE VARIANCE DESIGN EFFECT (DEFF) = ---------------------------------- SIMPLE RANDOM SAMPLE VARIANCE is often used to show the impact of the complex sample design on variances. If the design effect is near 1, the complex sample design has little effect on the variances and the user could consider assuming simple random sampling for the analysis. Some illustrative design effects for HHANES data on this tape are given in the following tables. The design effects in the tables are the average for the age groups usually presented in NCHS Series 11 publications. If the average design effect for a subgroup was less than 1.0 (implying an improvement over simple random sampling), it was coded as 1.0. The following guidelines were used in the calculation of the average design effects: 1. Exclude all persons of non-Hispanic origin, 2. Exclude all estimates for large age ranges, such as all ages combined or 'all adults', and 3. Exclude all estimates where the proportion of the subpopulation with the specific characteristic or condition was zero percent or one hundred percent. Design effects tend to be larger when age groups are combined, just as they are when the sexes are combined, as shown in the tables. The data in the tables give the user an idea of the range in design effects for selected response variables from this data tape. If a response variable is not one shown in the tables take the range into account; it is possible that a user could have one of the higher, rather than one of the lower, design effects. Suppose for example, that of the 3,532 Mexican-Americans ages 20-74 years who had an electrocardiogram (ECG), 8.9 percent have an equivocal ECG finding. Suppose, also, that their mean heart rate was 67. Assuming simple random sampling, the variance for the percent is calculated by converting the percent to a proportion and using the standard formula for the variance of a proportion, pq V = __ n This variance (V) multiplied by the design effect (DEFF) provides an estimate of the variance from a complex sample of the same sample size (n). In the example above, (.089)(.911) V = ____________ 3532 = .000029 = variance for a simple random sample Average Design Effects, by Sex, for Selected Variables - Mexican-American Portion Variable Mean or Tape Both Male Female Proportion Positions Sexes ECG Estimate of LV Mass Index x 456-459 1.3 1.1 1.3 Major Abnormality (adjusted) p 461 1.0 1.0 1.0 Minor Abnormality (adjusted) p 462 1.4 1.0 1.3 Diagnostic ECG (adjudicated) p 463 1.5 1.4 1.0 Equivocal ECG (adjudicated) p 464 1.2 1.0 1.2 Heart Rate/min x 513-516 1.3 1.0 1.1 Source: NCHS, HHANES, 1982-84, Tape Number 6540, Version 01 Average Design Effects, by Sex, for Selected Variables - Cuban-American Portion Variable Mean or Tape Both Male Female Proportion Positions Sexes ECG Estimate of LV Mass Index x 456-459 1.4 1.6 1.3 Major Abnormality (adjusted) p 461 1.1 1.0 1.0 Minor Abnormality (adjusted) p 462 1.0 1.0 1.1 Diagnostic ECG (adjudicated) p 463 1.3 1.3 1.0 Equivocal ECG (adjudicated) p 464 1.0 1.0 1.0 Heart Rate/min x 513-516 1.0 1.0 1.1 Source: NCHS, HHANES, 1982-84, Tape Number 6540, Version 01 Average Design Effects, by Sex, for Selected Variables - Puerto Rican Portion Variable Mean or Tape Both Male Female Proportion Positions Sexes ECG Estimate of LV Mass Index x 456-459 1.0 1.0 1.1 Major Abnormality (adjusted) p 461 1.5 1.5 1.1 Minor Abnormality (adjusted) p 462 1.0 1.3 1.3 Diagnostic ECG (adjudicated) p 463 1.1 1.0 1.0 Equivocal ECG (adjudicated) p 464 1.1 1.6 1.1 Heart Rate/min x 513-516 1.0 1.3 1.0 Source: NCHS, HHANES, 1982-84, Tape Number 6540, Version 01 Then, multiplying by the design effect, = (.000022)(1.2) = .000028 = estimated variance for the complex sample In a similar way, the complex sample variance of the mean heart rate is determined by multiplying the simple random sample variance of the mean by the appropriate design effect--in this example, 1.3. The user can then proceed with estimating confidence intervals and testing hypotheses in the usual manner. The user should recognize that this approach does not incorporate the variance covariance matrix. In most cases, this leads to a slight overestimate of the variance because the covariance terms, which are subtracted in the variance of a ratio, in general are positive. Thus, in a borderline case, the null hypothesis would be less likely to be rejected (Ref. No. 11). Alternative or better approaches may exist or be developed. Users who want to suggest such approaches, or who want the latest information should contact the Scientific and Technical Information Branch (address given in the beginning of this documentation).Data Collection and Processing Procedures Data presented in Sections E through H and the family relationships data in Section J were collected on the Household Screener and Family Questionnaires. These interview schedules were administered in sample persons' households. (Data presented in Section K were collected in the Mobile Examination Centers.) Completed interview schedules were reviewed in the Survey's field offices and again at the data processing center of NCHS by clerical editors. The editors checked the forms for completeness, clarity, and compliance with skip patterns, and they coded items such as industry and occupation. At the data processing center the questionnaires were keyed and verified on key-to-disk data entry equipment under the control of programs that checked for valid codes and ranges, compliance with skip patterns and consistency. After being keyed, data were reedited by analysts for reasonableness and consistency and for compliance with instructions for sampling and questionnaire administration. The general tape description format is Tape Position X Item X Counts. The item (field) may be a tape descriptor (e.g., Version Number), a sample person descriptor (e.g. Age at Interview), or a question (e.g., Is sample person covered by Medicare?). Where appropriate, data entries are presented by codes. Frequency counts are given for each code. The counts are included to help the user in planning analyses and in verifying that programs account for all data. The data source is given also (e.g., from Family Questionnaire). In some cases, a note is referenced. The notes contain explanations of the item (e.g., how Poverty Index is calculated). The questionnaire data have undergone many quality control and editing procedures. The responses of sample persons to some questions may appear extreme or illogical. Self-reported data, especially, are subject to a number of sources of variability, including recall and other reporting errors. In the data clean-up process, responses that varied considerably from expected were verified through direct review of the collection form or a copy of it. Such responses may not represent fact, but they are included as recorded in the field. The user must determine if these responses should be included in analyses. Responses to "other" and "specify" were recoded to existing categories, if possible. For responses that could not be recoded, new code categories were created if the information was deemed analytically useful. Caution should be used in interpreting the data from these new categories because there is no way of knowing which other respondents would have selected one of the new categories if given the option. For the adult sample person questionnaires there are three codes for missing information: 7's, 8's, and blanks. In a few questions, 7's were used when the question was not applicable. A code "8", which is labeled as "blank but applicable", is used to indicate that a sample person should have a data value for a particular item but for varying reasons that value is unavailable. Blanks were used to follow skip patterns, i.e., when a question was not supposed to be asked or was not applicable. The "don't know" codes (9, 99, 999) were used only when given as a printed response on the original questionnaire. Copies of the questionnaires, both in English and Spanish, can be found in the plan and operation report for HHANES (Ref. No. 1). Detailed information on interviewing and examination procedures is contained in the household interviewer's manual (Ref. No. 12) and the examination staff procedures manual (Ref. No. 13). These manuals are available upon request from: Division of Health Examination Statistics National Center for Health Statistics Presidential Building, Room 900 6525 Belcrest Road Hyattsville, MD 20782 301-436-7080 The electrocardiographic data presented in Section K give an objective measure of cardiac health status of individuals examined in the HHANES. The electrocardiogram is a relatively insensitive tool for the assessment of cardiac health status. In addition, the individual electrocardiographic measures of cardiac health status have varying specificity in terms of the extent to which they represent abnormality. Despite these limitations, the electrocardiographic variables are carefully and completely defined. Also, an extensive process was used to ensure the accuracy of findings described in this data tape. Data Collection Procedures The procedure for collection of the electrocardiographic data which is summarized below, is described in detail in the Examination Staff Procedures Manual for HHANES, 1982-1984 (Instruction Manual, 15a). Marquette Electronics Micro computer Augmented Cardiograph Digital Recording Electrocardiograph units (MAC-DR) were used for recording the electrocardiograms. These units recorded 10 seconds of ECG data simultaneously for 8 independent components of the 12 standard leads, followed by a calibration signal. The following computer algorithm: III = II - I aVR = -1/2(I + II) aVL = I - 1/2(II) aVF = II - 1/2(I) provided the remaining 4 leads, III, aVR, aVL and aVF from leads I and II. Subsequently, a 20-second rhythm strip was recorded from the Frank XYZ leads. The electrocardiographic data were stored on magnetic tape cassettes after analog to digital conversion. The sampling rate was 250 samples/sec channel for the 12-lead ECG and 500 samples/sec channel for the XYZ leads. These data were subsequently transferred to IBM compatible 9-track digital tapes for processing and permanent storage. Hardcopy tracings including 5.4 seconds from each of the standard leads and 5.4 seconds from the XYZ leads were printed at a paper speed of 25mm/sec. A 35-Hz filter was used to reduce noise on the ECG paper tracing. Frequency response was equal to or exceeded the American Heart Association recommendations (Circulation 52(2):11, 1975). Frequency response was -3db at 0.001 Hz low frequency measured from initial 320 ms of step input response, -3db at 140 Hz high frequency with 5mm amplitude signal. Tracings were recorded by health technicians. Their performance was monitored by a supervisory health technician who reviewed samples of the tracings for poor technique. Field calibration of the electrocardiographic units was performed weekly and included: 1) a self-calibration testing gain, linearity, step response, frequency response and offset; 2) external source calibration testing response of unit to external-generated calibrated signals, and; 3) playback calibration testing the ability to produce an identical copy of a previously recorded electrocardiogram. A staff bioengineer reviewed the results of these tests and ensured continued adequate functioning of the equipment. All electrocardiograms were performed prior to any glucose load. Time since last meal varied because some sample persons had fasted overnight and some had afternoon examinations after lunch. Data Processing Procedures The procedure for measurement, classification and validation of the electrocardiographic variables involved: 1. A computer reading of all electrocardiograms; 2. A manual reading of electrocardiograms with myocardial infarction- associated codes by an expert cardiologist; 3. Manual reading of the latter subset of electrocardiograms by technician coders; 4. Adjudication of differences in the computer reading and the two sets of manual readings by three expert electrocardiographers. The digital electrocardiographic data collected in the field at the mobile examination centers were initially sent to the Dalhousie University Heart Disease Research Centre for computer measurement and coding of the waveforms. The procedures for this computer coding are documented in Section M. These measurements and codes are the source of the "Dalhousie" computer-generated variables which are provided. Electrocardiograms with Dalhousie computer- generated myocardial infarction-associated codes (Minnesota codes 1.1.1- 1.2.5, 1.2.7, 1.2.8-1.3.6, 9.2, 4.1-4.3, 5.1-5.3) were then manually assessed by Cardiologist, Daniel Savage, MD, PhD and coders at the University of Minnesota Electrocardiography Coding Laboratory. Procedures at the Coding Laboratory are documented in Reference 4 (Section M). Finally, differences among the manual readings and the computer readings were adjudicated in a review of individual tracings by three experts (Pentti Rautaharju, MD, PhD, Ronald Prineas, BM, PhD and Daniel D. Savage, MD, PhD). This latter process is the source of the adjudicated readings.References 1. National Center for Health Statistics: Maurer, K.R. and others: Plan and Operation of the Hispanic Health and Nutrition Examination Survey, 1982-84. Vital and Health Statistics. Series 1, No. 19. DHHS Pub. No. (PHS) 85-1321. Public Health Service. Washington. U.S. Government Printing Office. Sept., 1985. 2. National Center for Health Statistics: McCarthy, P.J.: Replication: An Approach to the Analysis of Data from Complex Surveys. Vital and Health Statistics. Series 2, No. 14. PHS Pub. No. 1000. Public Health Service. Washington. U.S. Government Printing Office. Apr., 1966. 3. Survey Research Center Computer Support Group: OSIRIS IV User's Manual, Institute for Social Research, University of Michigan, Ann Arbor, MI, 1979. 4. Holt, M.M.: SURREGR: Standard Errors of Regression Coefficients from Sample Survey Data. Research Triangle Institute, Research Triangle Park, NC, 1977. (Revised Apr., 1982 by B.V. Shah). 5. Hidiroglou, M.A., Fuller, W.A. and Hickman, R.D.: SUPERCARP. Sixth Edition. Survey Section, Statistical Laboratory, Iowa State University, Ames, IA. Oct., 1980. 6. Woodruff, R.S.: A simple Method for Approximating the Variance of a Complicated Estimate. Journal of the American Statistical Association, 66:411-414, 1971. 7. Shah, B.V.: SESUDAAN: Standard Errors Program for Computing of Standardized Rates from Sample Survey Data. RT1/5250/00-01S. Research Triangle Institute, Research Triangle Park, NC. Apr., 1981. 8. Helnig, J.T. and Council, K.A., eds.: SAS Users' Guide: Basics. SAS Institute, Inc. Cary, NC. 1982. 9. National Center for Health Statistics: Landis, J.R., Lepkowski, J.M., Eklund, S.A., and Stehouwer, S.A.: A Statistical Methodology for Analyzing Data from a Complex Survey: The First National Health and Nutrition Examination Survey. Vital and Health Statistics. Series 2, No. 92. DHHS Pub. No. (PHS) 82-1366. Public Health Service. Washington. U.S. Government Printing Office. Sept., 1982. 10. Kovar, M.G. and Johnson, C.: Design Effects from the Mexican- American Portion of the Hispanic Health and Nutrition Examination Survey: A Strategy for Analysts. Proceedings of the section on Survey Research Methods, American Statistical Association, 1986, pp 396-399. 11. Freeman, D.H. and Brock, D.B.: The Role of Covariance Matrix Estimation in the Analysis of Complex Sample Survey Data. In N. Krishnan Namboodiri, ed., Survey Sampling and Measurement. Symposium on Survey Sampling, 2d, University of North Carolina. New York, Academic Press, 1978. 12. National Center for Health Statistics: Instruction Manual Part 15h, Household Interviewer's Manual for the Hispanic Health and Nutrition Examination Survey, 1982-84. Hyattsville, MD. 1986. 13. National Center for Health Statistics: Instruction Manual Part 15a, Examination Staff Procedures Manual for the Hispanic Health and Nutrition Examination Survey, 1982-84. Hyattsville, MD. 1986.TAPE LAYOUT Tape Position Index TAPE POSITIONS 1-400 contain data categories common to all data tapes: sociodemographic data, family composition, family income, residence and household. Sample weights are also in this set of data. TAPE POSITION 401+ contain data categories unique to this data tape. SOCIODEMOGRAPHIC DATA - SAMPLE PERSON 1-5 Sample Person Sequence Number 6-15 Survey and Tape Indentifiers 16 Examination Status 17 Language of Interview 18-21 Date of Interview 22-25 Date of Examination 26-29 Date of Birth 30-32 Age at Interview 33-38 Age at Examination 39-43 Family Number 44-45 Relationship to Head of Family 46 Sex 47 Race 48-49 National Origin or Ancestry 50-52 Birth Place 53 National Origin Recode 54-56 Education 57 Marital Status 58 Service in Armed Forces 59-69 Work/Occupation/Employment 70-95 Health Insurance/Health Care Support 96-99 Income Assistance/Public Compensation or Support SOCIODEMOGRAPHIC DATA - HEAD OF FAMILY 100 Interview and Examination Status 102-105 Date of Birth 106-108 Age at Inteview 109 Sex 110 Race 111-112 National Origin or Ancestry 113-115 Birth Place 116-118 Education 119 Marital Status 120 Service in Armed Forces 121-131 Work/Occupation/Employment FAMILY COMPOSITION AND INCOME DATA 132-133 Number of People in Family 134-135 Number of Sample People in Family 136-138 Combined Family Income 139-143 Per Capita Income 144-146 Poverty Index 147-162 Income, Food Stamps RESIDENCE AND HOUSEHOLD DATA 163 Size of Place 164 Standard Metropolitan Statistical Area 165-166 Number of People in Household 167-168 Number of Sample People in Household 169-170 Number of Rooms 171 Kitchen Facilities Access 172-183 Heating/Cooling Equipment SAMPLE WEIGHTS 184-189 Examination Final Weight 190-195 Interview Final Weight 196-201 GTT/Ultrasound Weight 202-207 Audiometry/Vision Weight 208-213 Pesticide Weight 214-215 Strata Code 216-217 Pseudo PSU Code FAMILY RELATIONSHIPS 218-400 Data not yet available CONVENTIONAL 12-LEAD ECG DATA 401-404 Tape number 410-415 Summary diagnoses based on computer readings 416-444 Minnesota Code items coded by computer 445-451 Cardiac Infarction/Injury Score 452-460 LV Mass and LV Mass Index estimates 461-482 Minnesota codes after adjudication 513-539 Heart rate, basic ECG interval and mean axis data 542-1017 Amplitude and duration measurements Sociodemographic Data - Sample Person (Pos 1-99) Tape Positions 1-69 Source: Family Questionaire (FQ) Household Screener Questionnaire (HSQ) Position Item Description Counts Source and Code M C P and Notes 1-5 Sample person sequence number 00001-09894 Mexican Americans 3555 - - 10002-12238 Cuban Americans - 907 - 13001-16785 Puerto Ricans - - 1353 6-12 Blank 13 Portion of survey 1 Mexican-American (M) 3555 - - 2 Cuban-American (C) - 907 - 3 Puerto Rican (P) - - 1353 14 Family Questionnaire missing 1 Yes 10 4 5 See Note 1 2 No 3545 903 1348 15 Version number 1 3555 907 1353 16 Examination status 1 Examined 3555 907 1353 See Note 2 2 Not examined 0 0 0 17 Language of interview (Pos 1-400) FQ 1 English 2127 157 561 2 Spanish 1418 746 787 Blank 10 4 5 18-19 Date of interview HSQ 4 01-12 Month 3555 907 1353 20-21 82-84 Year 3555 907 1353 Date of examination From survey control record 22-23 01-12 Month 3555 907 1353 24-25 82-84 Year 3555 907 1353 Date of birth HSQ 2e 26-27 01-12 Month 3555 907 1353 88 Blank but applicable 0 0 0 28-29 08-64 Year 3555 907 1353 88 Blank but applicable 0 0 0 30-31 Age at interview (computed) 20-74 (See next column for units) 3555 907 1353 32 Age at interview units HSQ 2f 1 Years 3555 907 1353 Age at examination (computed) Positions 33-38 are all 0 for non-examined persons 33-34 20-75 Years 3555 907 1353 35-36 00-11 Months 3555 907 1353 37-38 00-30 Days 3555 907 1353 39-43 Family number See Note 3 00002-03527 3555 - - 04005-04922 - 907 - 07003-08584 - - 1353 44-45 What is sample person's relationship HSQ 2b to head of family? Sample person is: See Note 4 01 Head of family living alone (1 family with only 1 member) 143 56 113 02 Head of family with no related persons in household (2+ persons in household) 70 23 23 03 Head of family with related persons in household 1566 368 674 04 Wife of head (husband living at home and not in Armed Forces) 1264 297 290 05 Wife of head (husband living at home and is in Armed Forces) 5 0 0 06 Husband of head (wife living at home and not in Armed Forces) 35 12 37 07 Husband of head (wife living at home and is in Armed Forces) 0 0 0 08 Child of head or head's spouse 277 76 126 09 Grandchild of head or head's spouse 7 0 3 10 Parent of head or head's spouse 57 35 33 11 Other relative (includes ex-spouse, daughter- in-law, etc.) 131 40 54 12 Foster child 0 0 0 46 Sex FQ B-4 1 Male 1572 393 498 2 Female 1983 514 855 47 Observed race FQ B-5 1 White 3446 870 1220 See Note 5 2 Black 30 13 62 3 Other 6 2 27 8 Blank but applicable 41 12 28 9 Not observed 22 6 11 Blank 10 4 5 48-49 Sample person's national HSQ 2c origin or ancestry See Note 6 01 Mexican/Mexicano 940 1 1 02 Mexican-American 2230 0 0 03 Chicano 46 0 0 04 Puerto Rican 7 3 1202 05 Boricuan 0 0 15 06 Cuban 3 796 14 07 Cuban-American 0 69 0 08 Hispano - specify 61 10 20 09 Other Latin-American or other Spanish - specify 25 16 25 00 Other - specify 217 12 76 10 Spanish-American 13 0 0 11 Spanish (Spain) 13 0 0 50-52 In what state or foreign country FQ B-6 was sample person born? 001-115 State/country code 3528 900 1324 888 Blank but applicable 17 3 24 Blank 10 4 5 53 National origin recode See Note 8 "Hispanic" = Mexican-American in Southwest, Cuban-American in Florida, and Puerto Rican in New York City area 1 "Hispanic" 3326 865 1220 2 Not "Hispanic" 229 42 133 54-55 What is the highest grade or FQ B-7 year of regular school sample person has ever attended? 00 Never attended or kindergarten only 141 6 23 01-08 Elementary grade 1312 342 435 09-12 High school grade 1442 292 666 13-16 College 536 228 195 17 Graduate school 69 30 14 88 Blank but applicable 45 5 15 Blank 10 4 5 56 Did sample person finish FQ B-8 that grade/year? 1 Yes 2779 779 1055 2 No 580 108 236 8 Blank but applicable 45 10 34 Blank 151 10 28 57 Is sample person now married, FQ B-9 widowed, divorced, separated or has he or she never been married? 1 Married - spouse in household 2539 622 647 2 Married - spouse not in household 68 17 53 3 Widowed 161 50 66 4 Divorced 209 92 154 5 Separated 149 21 147 6 Never married 403 100 275 8 Blank but applicable 16 1 6 Blank 10 4 5 58 Did sample person ever serve FQ B-11 in the Armed Forces of the United States? 1 Yes 413 27 141 2 No 3125 874 1198 8 Blank but applicable 7 2 9 Blank 10 4 5 59 During the past 2 weeks, did FQ B-12 sample person work at any time at a job or business, not counting work around the house? 1 Yes 2028 581 566 2 No 1507 317 767 8 Blank but applicable 10 5 15 Blank 10 4 5 60 Even though sample person did FQ B-13 not work during those 2 weeks, did he or she have a job or business? 1 Yes 44 12 21 2 No 1462 303 743 8 Blank but applicable 11 7 18 Blank 2038 585 571 61 Was sample person looking for FQ B-14 work or on layoff from a job? 1 Yes 200 39 51 2 No 1306 276 713 8 Blank but applicable 11 7 18 Blank 2038 585 571 62 Which, looking for work or on FQ B-15 layoff from a job or both? 1 Looking 131 30 35 2 Layoff 45 6 8 3 Both 22 2 7 8 Blank but applicable 13 8 19 Blank 3344 861 1284 63-65 What kind of business or FQ B-19 industry does sample person See Note 9 work for? 010-932 Industry code 2238 621 628 990 Blank but applicable 31 10 21 Blank 1286 276 704 66-68 What kind of work was sample FQ B-20 person doing? See Note 9 003-889 Occupation code 2240 622 628 999 Blank but applicable 29 9 21 Blank 1286 276 704 69 Class of worker FQ B-22 1 An employee of a private company, business or individual for wages, salary, or commission 1742 502 501 2 A Federal government employee 72 6 18 3 A State government employee 119 18 17 4 A Local government employee 163 16 56 5 Self-employed in own incorporated business or professional practice 17 12 7 6 Self-employed in own unincorporated business, professional practice or farm 125 66 26 7 Working without pay in family business or farm 1 0 0 8 Blank but applicable 29 10 23 0 Never worked or never worked at a full-time civilian job lasting 2 weeks or more 1 1 1 Blank 1286 276 704 Tape Positions 70-99 Source: Family Questionaire (FQ) Household Screener Questionnaire (HSQ) Position Item Description Counts Source and Code M C P and Notes 70 Is sample person now covered FQ C-2 by Medicare? 1 Covered 258 103 109 2 Not covered 3279 796 1234 8 Blank but applicable 5 3 5 9 Don't know 3 1 0 Blank 10 4 5 71 Is sample person now covered FQ C-3 by the part of Social Security Medicare which pays for hospital bills? 1 Yes 230 96 100 2 No 13 4 4 8 Blank but applicable 14 3 10 9 Don't know 6 3 0 Blank 3292 801 1239 72 Is sample person now covered FQ C-4 by that part of Medicare which pays for doctor's bills? This is the Medicare plan for which he or she or some agency must pay a certain amount each month. 1 Yes 227 98 92 2 No 14 3 11 8 Blank but applicable 14 3 10 9 Don't know 8 2 1 Blank 3292 801 1239 73 Type of Medicare coverage FQ C-5 As shown on Medicare card 2 Medical 2 0 0 3 Card not available 3 0 1 4 Hospital and medical 5 3 0 8 Blank but applicable 14 3 10 Blank 3531 901 1342 HEALTH INSURANCE See Note 10 74 Is sample person covered by FQ C-11 any health insurance plan which pays any part of a hospital, doctor's, or surgeon's bill? 1. Yes 2009 556 560 2 No 1526 343 779 8 Blank but applicable 6 4 9 9 Don't know 4 0 0 Blank 10 4 5 75 Is sample person covered by a FQ C-9 plan that pays any part of hospital expenses? 1 Yes 1979 550 525 2 No 3 3 4 8 Blank but applicable 27 7 35 9 Don't know 6 0 5 Blank 1540 347 784 76 Is sample person covered by a FQ C-10 plan that pays any part of a doctor's or surgeon's bills for operations? 1 Yes 1982 546 523 2 No 9 7 17 8 Blank but applicable 15 7 20 9 Don't know 9 0 9 Blank 1540 347 784 Many people do not carry health FQ C-13/15 insurance for various reasons. See Note 10 Which of these statements describes why sample person is not covered by any health insurance (or Medicare)? (Positions 77-80) 77-78 Main reason 01 Care received through Medicaid or welfare 79 14 289 02 Unemployed, or reasons related to unemployment 180 29 68 03 Can't obtain insurance because of poor health, illness, or age 17 2 9 04 Too expensive, can't afford health insurance 756 163 226 05 Dissatisfied with previous insurance 23 2 2 06 Don't believe in insurance 14 3 4 07 Have been healthy, not much sickness in the family, haven't needed health insurance 94 15 13 08 Military dependent, (CHAMPUS), Veteran's benefits 20 1 11 09 Some other reason--not specified 1 0 2 10 Some other reason-- specified 112 19 37 88 Blank but applicable 49 23 29 Blank 2210 636 663 79-80 Second reason 00 No second reason reported 1102 199 564 01 Care received through Medicaid or welfare 25 10 23 02 Unemployed, or reasons related to unemployment 46 16 12 03 Can't obtain insurance because of poor health, illness, or age 3 1 2 04 Too expensive, can't afford health insurance 81 14 56 05 Dissatisfied with previous insurance 8 1 2 06 Don't believe in insurance 7 1 1 07 Have been healthy, not much sickness in the family, haven't needed health insurance 21 4 3 08 Military dependent, (CHAMPUS), Veteran's benefits 0 0 0 09 Some other reason--not specified 0 0 0 10 Some other reason-- specified 13 6 2 88 Blank but applicable 39 19 25 Blank 2210 636 663 81-87 Blank 88 During the last 12 months, has sample FQ D-6 person received health care which has been or will be paid for by Medicaid? 1 Yes 195 66 385 2 No 3328 832 949 8 Blank but applicable 22 5 14 9 Don't know 0 0 0 Blank 10 4 5 89 Does sample person have a Medicaid card? FQ D-8 1 Yes 198 67 403 2 No 3329 825 931 8 Blank but applicable 18 11 14 9 Don't know 0 0 0 Blank 10 4 5 90 Status of sample person's Medicaid card. FQ D-9 1 Medicaid card seen- current 148 52 285 2 Medicaid card seen- expired 2 0 5 3 No card seen 43 13 103 4 Other card seen 0 0 0 5 Other card seen (specify) 3 0 1 8 Blank but applicable 20 13 23 Blank 3339 829 936 91 Is sample person now covered by any FQ D-11 other public assistance program that pays for health care? 1 Yes 12 2 12 2 No 3527 900 1331 8 Blank but applicable 6 1 5 9 Don't know 0 0 0 Blank 10 4 5 92 Does sample person now receive military FQ D-13 retirement payments from any branch of the Armed Forces or a pension from the Veteran's Administration? Do not in- clude VA disability compensation. 1 Yes 45 2 9 2 No 3492 900 1335 8 Blank but applicable 8 1 4 9 Don't know 0 0 0 Blank 10 4 5 93 Which does sample person receive: the FQ D-14 Armed Forces retirement, the VA pension, or both? 1 Armed Forces 12 0 2 2 Veteran's Administration 24 0 5 3 Both 4 2 1 8 Blank but applicable 13 1 5 Blank 3502 904 1340 94 Is sample person now covered by FQ D-16 CHAMP-VA, which is medical insurance for dependents or survivors of disabled veterans? 1 Yes 20 2 6 2 No 3520 900 1340 8 Blank but applicable 5 1 2 9 Don't know 0 0 0 Blank 10 4 5 95 Is sample person now covered by any other FQ D-18 program that provides health care for military dependents or survivors of military persons? 1 Yes 20 1 5 2 No 3518 901 1336 8 Blank but applicable 7 1 7 9 Don't know 0 0 0 Blank 10 4 5 96 Is sample person included in the AFDC, FQ D-2 "Aid to Families with Dependent Children", assistance program? 1 Yes 87 15 182 2 No 3448 885 1153 8 Blank but applicable 10 2 12 9 Don't know 0 1 1 Blank 10 4 5 97 Does sample person now receive the FQ D-4 "Supplemental Security Income" or "SSI" gold-colored check? 1 Yes 92 43 91 2 No 3441 852 1247 8 Blank but applicable 12 8 10 9 Don't know 0 0 0 Blank 10 4 5 98 Does sample person have a disability FQ D-20 related to his or her service in the Armed Forces of the United States? 1 Yes 48 2 14 2 No 343 20 106 8 Blank but applicable 29 7 30 Blank 3135 878 1203 99 Does sample person now receive FQ D-21 compensation for this disability from the Veteran's Administration? 1 Yes 31 1 9 2 No 17 1 4 8 Blank but applicable 29 7 31 Blank 3478 898 1309 Sociodemographic Data - Head of Family (Pos 100-131) Source: Family Questionaire (FQ) Household Screener Questionnaire (HSQ) Position Item Description Counts Source and Code M C P and Notes 100 Interview and examination status of See Note 4 head of family 1 Selected as sample person, interviewed on Adult Sample Person Questionnaire, and examined 3158 764 1266 2 Selected as sample person, interviewed on Adult Sample Person Questionnaire, but not examined 120 32 30 3 Selected as sample person, not interviewed, and not examined 98 21 5 4 Not selected as sample person 169 86 47 Blank 10 4 5 101 Blank Date of birth HSQ 2e 102-3 01-12 Month 3535 902 1352 88 Blank but applicable 0 5 1 104-5 00-86, 89-99 Year 3545 904 1352 88 Blank but applicable 10 3 1 106-7 Age at interview 18-95 Years 3555 907 1353 108 Blank 109 Sex FQ B-4 1 Male 2939 719 768 2 Female 606 184 580 Blank 10 4 5 110 Observed race FQ B-5 1 White 3417 862 1218 See Note 5 2 Black 35 17 67 3 Other 4 2 22 8 Blank but applicable 54 18 29 9 Not observed 35 4 12 Blank 10 4 5 111-2 Head of family's national origin HSQ 2c or ancestry. See Note 6 01 Mexican-Mexicano 948 0 2 02 Mexican-American 2180 0 0 03 Chicano 46 0 0 04 Puerto Rican 9 5 1198 05 Boricuan 0 0 14 06 Cuban 4 801 22 07 Cuban-American 0 58 0 08 Hispano--specify 65 14 16 09 Other Latin-American or other Spanish--specify 24 11 16 00 Other--specify 254 18 85 10 Spanish-American 11 0 0 11 Spanish (Spain) 14 0 0 113-5 In what state or foreign country FQ B-6 was head of family born? See Note 7 001-118 State/country code 3509 893 1320 888 Blank but applicable 36 10 28 Blank 10 4 5 116-7 What is the highest grade or year of FQ B-7 regular school head of family has ever attend? 00 Never attended or kindergarten only 138 4 17 01-08 Elementary grade 1406 350 482 09-12 High school grade 1323 256 630 13-16 College 515 236 175 17 Graduate school 88 41 24 88 Blank but applicable 75 16 20 Blank 10 4 5 118 Did head of family finish that FQ B-8 grade/year? 1 Yes 2733 788 1083 2 No 603 90 216 8 Blank but applicable 71 21 32 Blank 148 8 22 119 Is the head of family now married, FQ B-9 widowed, divorced, separated or has he or she never been married? 1 Married--spouse in househld 2741 694 710 2 Married--spouse not in household 56 7 48 3 Widowed 179 41 70 4 Divorced 198 94 156 5 Separated 143 14 178 6 Never married 189 46 183 8 Blank but applicable 39 7 3 Blank 10 4 5 120 Did head of family ever serve in the FQ B-11 Armed Forces of the United States? 1 Yes 780 38 239 2 No 2730 854 1096 8 Blank but applicable 35 11 13 Blank 10 4 5 121 During the past 2 weeks, did head of FQ B-12 family work at any time at a job or business, not counting work around the house? 1 Yes 2529 661 699 2 No 986 230 638 8 Blank but applicable 30 12 11 Blank 10 4 5 122 Even though head of family did not work FQ B-13 during those 2 weeks, did he or she have a job or business? 1 Yes 52 14 15 2 No 934 216 623 8 Blank but applicable 30 12 11 Blank 2539 665 704 123 Was head of family looking for work FQ B-14 or on layoff from a job? 1 Yes 220 46 54 2 No 766 184 583 8 Blank but applicable 30 12 12 Blank 2539 665 704 124 Which, looking for work or on layoff FQ B-15 from a job or both? 1 Looking 115 31 32 2 Layoff 63 9 10 3 Both 40 3 9 8 Blank but applicable 32 15 15 Blank 3305 849 1287 125-7 What kind of business or industry FQ B-19 does head of family work for? See Note 9 010-932 Industry code 2769 705 752 990 Blank but applicable 49 15 22 Blank 737 187 579 128-30 What kind of work was head of family FQ B-20 doing? See Note 9 003-889 Occupation code 2771 705 750 999 Blank but applicable 47 15 24 Blank 737 187 579 131 1 Employee of a private company, FQ B-22 Class of worker business individual for wages, salary, or commission 2155 543 567 2 A Federal government employee 98 3 24 3 A State government employee 118 11 29 4 A Local government employee 180 19 90 5 Self-employed in own incorporated business or professional practice 26 19 10 6 Self-employed in own unincorporated business, professional practice, or farm 201 108 32 7 Working without pay in family business or farm 0 0 0 8 Blank but applicable 39 17 21 0 Never worked or never worked at a full-time civilian job lasting 2 weeks or more 1 0 1 Blank 737 187 579 Family Composition and Income Data (Pos 132-162) Source: Family Questionnaire (FQ) Source: Family Questionaire (FQ) Household Screener Questionnaire (HSQ) Position Item Description Counts Source and Code M C P and Notes 132-3 Number of persons in family (computed) 01-18 Persons 3555 907 1353 134-5 Number of sample persons in family (computed) 01-13 Persons 3555 907 1353 136 Was the total combined family income dur- FQ E-10 ing the past 12 months more or less than $20,000? Include money from jobs, Social Security, retirement income, unemployment payments, public assistance, and so forth. Also include income net from interest, divi- dends, income from business, farm or rent, and any other money income received. 1 $20,000 or more 1195 361 326 2 Less than $20,000 2233 526 1000 7 Refused information 18 1 5 8 Blank but applicable 99 15 17 Blank 10 4 5 137-8 Of those income groups, which best rep- FQ E-11 resents the total combined family income during the past 12 months? Include wages, salaries, and other items we just talked about (in dollars). 01 Less than 1,000 22 7 4 02 1,000 - 1,999 46 6 15 03 2,000 - 2,999 51 14 34 04 3,000 - 3,999 82 20 55 05 4,000 - 4,999 97 21 126 06 5,000 - 5,999 117 32 75 07 6,000 - 6,999 143 26 82 08 7,000 - 7,999 146 31 68 09 8,000 - 8,999 118 26 45 10 9,000 - 9,999 126 33 56 11 10,000 - 10,999 132 46 59 12 11,000 - 11,999 109 31 33 13 12,000 - 12,999 143 39 53 14 13,000 - 13,999 90 21 29 15 14,000 - 14,999 111 17 32 16 15,000 - 15,999 99 23 41 17 16,000 - 16,999 95 22 31 18 17,000 - 17,999 104 21 32 19 18,000 - 18,999 147 20 45 20 19,000 - 19,999 116 34 46 21 20,000 - 24,999 336 101 79 22 25,000 - 29,999 293 61 68 23 30,000 - 34,999 163 44 51 24 35,000 - 39,999 145 48 31 25 40,000 - 44,999 107 32 22 26 45,000 - 49,999 52 27 21 27 50,000 and over 54 34 31 77 Refused information 41 9 25 88 Blank but applicable 260 57 59 Blank 10 4 5 139-43 Per capita income (computed) See Note 11 00083-50000 Dollars 3244 837 1264 88888 Blank but applicable 301 66 84 Blank 10 4 5 144-6 Poverty index (computed) See Note 12 Decimal not shown on tape 0.04-9.78 3244 837 1264 999 Blank but applicable 301 66 84 Blank 10 4 5 147 Did any member of this family receive FQ E-12 any Government food stamps in any of the past 12 months? 1 Yes 619 149 506 2 No 2921 752 840 8 Blank but applicable 5 2 2 Blank 10 4 5 148-9 In how many months of the past 12 FQ E-13 months did any member of this family receive food stamps? 01-12 Months 613 149 504 88 Blank but applicable 11 2 4 Blank 2931 756 845 150 Did this family receive any govern- FQ E-14 ment food stamps last month? 1 Yes 502 120 481 2 No 116 29 25 8 Blank but applicable 6 2 2 Blank 2931 756 845 151-2 In which month did any member of this FQ E-15 family last receive food stamps? 01-12 Month 114 29 25 88 Blank but applicable 8 2 2 Blank 3433 876 1326 153-4 For how many persons were those FQ E-16 food stamps authorized? 01-13 Persons 614 149 505 88 Blank but applicable 10 2 3 Blank 2931 756 845 155-7 What was the total face value of those FQ E-17 food stamps received by this family in that month? 010-520 Dollars 585 147 499 888 Blank but applicable 39 4 9 Blank 2931 756 845 158 Did this family spend more for food FQ E-18 in that month than the value of your food stamps? 1 Yes 539 128 492 2 No 74 21 14 8 Blank but applicable 11 2 2 Blank 2931 756 845 159-61 How much more? FQ E-19 003-880 Dollars 501 120 482 888 Blank but applicable 49 10 12 Blank 3005 777 859 162 Is your family receiving food FQ E-20 stamps at the present time? 1 Yes 474 116 473 2 No 3061 783 869 8 Blank but applicable 10 4 6 Blank 10 4 5 Residence and Household Data (Pos 163-183) Source: Family Questionaire (FQ) Household Screener Questionnaire (HSQ) Position Item Description Counts Source and Code M C P and Notes 163 Size of place See Note 13 1 1 million or more 484 0 1009 2 500,000 - 999,999 423 0 0 3 250,000 - 499,999 450 334 0 4 100,000 - 249,999 91 235 155 5 50,000 - 99,999 568 49 32 6 25,000 - 49,999 386 131 98 7 10,000 - 24,999 352 76 46 8 200 - 9,999 491 52 13 9 Not in a place 310 30 0 164 Standard Metropolitan Statistical Area See Note 13 1 In SMSA, in central city 1771 334 1173 2 In SMSA, not in central city 1332 573 180 4 Not in SMSA 452 0 0 165-6 Number of persons in household HSQ 1a 01-18 Persons 3555 907 1353 167-8 Number of sample persons in household (computed) 01-13 Persons 3555 907 1353 169-70 How many rooms are in this home? Count FQ E-1 the kitchen, but not the bathroom. 01-14 Rooms 3541 902 1346 88 Blank but applicable 4 1 2 Blank 10 4 5 171 Do you have access to complete kitchen FQ E-2 facilities in this home; that is, a kitchen sink with piped water, a refrig- erator and a range or cookstove? 1 Yes 3397 879 1213 2 No 40 7 8 8 Blank but applicable 108 17 127 Blank 10 4 5 172-3 What is the main fuel used for heating FQ E-3 this home? See Note 14 00 No fuel used 214 164 5 01 Oil 2 0 940 02 Natural gas 2888 45 362 03 Electricity 277 682 15 04 Bottled gas (propane) 85 2 0 05 Kerosene 7 2 0 06 Wood 45 3 0 07 Coal 0 0 6 08 Other, not specified 0 0 1 09 Other, specified 6 0 2 88 Blank but applicable 21 5 17 Blank 10 4 5 174-5 What is the main heating equipment FQ E-4 for this home? See Note 14 00 No heating equipment used 214 164 6 01 Steam or hot water with radiators or convectors 19 4 686 02 Central warm air furnace with ducts to individual rooms, or central heat pump 1296 340 96 03 Built-in electric units (permanently installed in wall, ceiling or baseboard) 219 219 32 04 Floor, wall or pipeless furnace 783 30 12 05 Room heaters with flue or vent burning oil, gas, or kerosene 403 12 301 06 Room heaters without flue or vent burning oil, gas, or kerosene 406 4 189 07 Heating stove burning wood, coal or coke 37 0 2 08 Fireplace(s) 37 4 0 09 Portable electric heater(s) 68 95 2 10 Other, not specified 0 0 0 11 Other, specified 53 26 7 88 Blank but applicable 1 5 8 99 Don't know 9 0 7 Blank 10 4 5 176-7 Are any other types of equipment FQ E-5 used for heating this home? See Note 14 00 No other heating equipment used 2877 707 1132 01 Steam or hot water with radiators or convectors 0 0 3 02 Central warm air furnace with ducts to individual rooms or central heat pump 5 9 5 03 Built-in electric units (permanently installed in wall, ceiling or baseboard) 16 0 1 04 Floor, wall or pipeless furnace 5 0 0 05 Room heaters with flue or vent burning oil, gas, or kerosene 12 0 0 06 Room heaters without flue or vent burning oil, gas, or kerosene 15 1 16 07 Heating stove burning wood, coal or coke 42 0 5 08 Fireplace(s) 239 5 4 09 Portable electric heater(s) 95 14 166 10 Other, not specified 4 1 0 11 Other, specified 10 1 1 88 Blank but applicable 15 1 9 Blank 224 168 11 178-9 What is the main fuel used by this FQ E-6 additional equipment? See Note 14 00 No fuel used 1 0 1 01 Oil 0 0 9 02 Natural gas 51 1 10 03 Electricity 116 24 162 04 Bottled gas (propane) 6 0 1 178-9 What is the main fuel used by this FQ E-6 additional equipment? (Cont'd) See Note 14 05 Kerosene 2 0 14 06 Wood 250 5 6 07 Coal 2 0 0 08 Other, not specified 0 0 0 09 Other, specified 5 0 0 88 Blank but applicable 21 2 7 Blank 3101 875 1143 180-1 What is the main fuel used for FQ E-7 cooking in this home? 00 No fuel used 10 4 2 01 Oil 5 0 9 02 Natural gas 2789 163 1236 03 Electricity 639 726 78 04 Bottled gas (propane) 85 7 7 05 Kerosene 0 0 3 06 Wood 0 0 0 07 Coal 0 0 0 08 Other, not specified 0 0 0 09 Other, specified 8 1 0 88 Blank but applicable 9 2 13 Blank 10 4 5 182 Do you have air-conditioning--either FQ E-8 individual room units, a central system or evaporative cooling? 1 Yes 1733 829 347 2 No 1806 73 995 8 Blank but applicable 6 1 6 Blank 10 4 5 183 Which do you have? FQ E-9 1 Individual room unit 779 411 328 2 Central air-conditioning 603 410 10 3 Evaporative cooling 349 3 4 8 Blank but applicable 8 6 11 Blank 1816 77 1000 Sample Weights (Pos 184-217) Source: Family Questionaire (FQ) Household Screener Questionnaire (HSQ) Position Item Description Counts Source and Code M C P and Notes 184-9 Examined final weight 000439-002711 3555 - - 000248-000891 - 907 - 000177-002000 - - 1353 190-5 Interview final weight 000447-002096 3555 - - 000207-000578 - 907 - 000175-001220 - - 1353 GTT/ULTRASOUND, AUDIOMETRY/VISION, PESTICIDE WEIGHTS By design, only some of the persons in the sample were included in the GTT/ultrasound, audiometry/vision, and pesticide components of the survey. Tape positions for those persons not part of these subsamples are BLANK. 196-201 GTT/ultrasound weight 000843-005302 1777 - - 000469-001685 - 449 - 000349-003110 - - 667 Blank 1778 458 686 202-7 Audiometry/vision weight 000870-006283 1778 - - 000454-001600 - 458 - 000343-003123 - - 686 Blank 1777 449 667 208-13 Pesticide Weight 000872-005584 1778 - - 000454-001600 - 458 - 000343-003117 - - 686 Blank 1777 449 667 214-5 Strata code 01-08 3555 907 1353 216-7 Pseudo PSU code 01-02 3555 907 1353 Family Relationships (Pos 218-400) Source: Family Questionaire (FQ) Household Screener Questionnaire (HSQ) Position Item Description Counts Source and Code M C P and Notes 218-400 Blank Data not yet available. Conventional 12-Lead ECG Data (Pos 401-1017) Tape Positions 401-476 Source: Family Questionaire (FQ) Household Screener Questionnaire (HSQ) Position Item Description Counts Source and Code M C P and Notes 401-4 Tape Number 6540 3555 907 1353 405-9 Blank Summary of Presence/Absence of Major and Minor ECG abnormalities according to the Minnesota Code as coded by the Dalhousie Program (positions 410-411) See Note 15 410 Major ECG abnormalities 1 Present 298 75 114 0 Absent 3234 820 1224 Blank 23 12 15 411 Minor ECG abnormalities 1 Present 529 163 288 0 Absent 3003 732 1050 Blank 23 12 15 Likelihood of Myocardial Infarction according to the Minnesota Code (positions 412-413) See Note 16 412 Probable MI - Diagnostic Q Waves 1 Present 72 14 21 0 Absent 3460 881 1317 Blank 23 12 15 413 Possible MI - Equivocal Q Waves or related abnormalities 1 Present 43 12 15 0 Absent 3489 883 1323 Blank 23 12 15 Likelihood of Left Ventricular Hypertrophy according to the Minnesota Code (positions 414-415) See Note 17 414 Probable LVH 1 Present 28 10 18 0 Absent 3505 885 1320 Blank 23 12 15 415 Possible LVH 1 Present 183 69 156 0 Absent 3349 826 1182 Blank 23 12 15 Minnesota Code (MC) Items Coded by Computer See Appendix 1 MC 1 (Q, QS waves and related items) See Note 18 416-7 MC 1 Leadgroup L(I, aVL, V6) 1.0-1.3.1 3532 895 1338 Blank 23 12 15 418-9 MC 1 Leadgroup F(II, III, aVF) 1.0-1.3.6 3532 895 1338 Blank 23 12 15 420-1 MC 1 Leadgroup V(V1-V5) 1.0-1.3.2 3532 895 1338 Blank 23 12 15 Highest code 1.1 8 4 7 in any lead group 68 9 21 Highest code 1.2 35 11 7 Highest code 1.3 3421 871 1303 Highest code 1.0 MC 4 (ST depression) See Note 19 422-3 MC 4 Leadgroup L 3532 895 1338 4.0-4.4.0 23 12 15 Blank 424-5 MC 4 Leadgroup F 3532 895 1338 4.0-4.4.0 23 12 15 Blank 426-7 MC 4 Leadgroup V 3532 895 1338 4.0-4.4.0 23 12 15 Blank Highest code 4.1 3 3 4 in any lead group 33 14 10 Highest code 4.2 19 5 3 Highest code 4.3 9 2 2 Highest code 4.4 3468 871 1319 Highest code 4.0 MC 5 (T wave abnormalities) See Note 20 428 MC 5 Leadgroup L 5.0-5.4 3532 895 1338 Blank 23 12 15 429 MC 5 Leadgroup F 5.0-5.4 3532 895 1338 Blank 23 12 15 430 MC 5 Leadgroup V 5.0-5.4 3532 895 1338 Blank 23 12 15 Highest code 5.1 in any leadgroup 2 2 5 Highest code 5.2 107 35 43 Highest code 5.3 112 22 50 Highest code 5.4 60 22 38 Highest code 5.0 3251 814 1202 MC 9.2 (ST elevation) See Note 21 431 MC 9.2 Leadgroup L 9.2.0 - 9.2.2 3532 895 1338 Blank 23 12 15 432 MC 9.2 Leadgroup F 9.2.0 - 9.2.2 3532 895 1338 Blank 23 12 15 433 MC 9.2 Leadgroup V 9.2.0 - 9.2.2 3532 895 1338 Blank 23 12 15 Highest Code 9.2.2 in any lead group 76 12 37 Other Minnesota Codes 434-5 MC 2 (QRS axis code) 2.0-2.3.0 3532 895 1338 Blank 23 12 1 436-7 MC 3 (Computer code for high-amplitude R waves) 3.0-3.3.2 3532 895 1338 Blank 23 12 15 438-9 MC 6 (A-V conduction) 6.0-6.5.0 3532 895 1338 Blank 23 12 15 440 MC 7 (Ventricular conduction) 7.0-7.6 3532 895 1338 Blank 23 12 15 441 MC 9.1 (Low-amplitude QRS) 9.1.0-9.1.1 3532 895 1338 Blank 23 12 15 442 MC 9.3 (High-amplitude P) 9.3.0-9.3.3 3532 895 1338 Blank 23 12 15 443 MC 9.4 (QRS transition zone) 9.4.0-9.4.2 3532 895 1338 Blank 23 12 15 444 MC 9.5 (High-amplitude T) 9.5.0-9.5.5 3532 895 1338 Blank 23 12 15 Likelihood of Myocardial Infarction and Left Ventricular Hypertrophy according to Dalhousie ECG Program 445-8 Cardiac Infarction/Injury Score for See Note 22 12-Lead ECG Multiplied by 10 -220 - 500 3484 891 1337 Blank 71 16 16 449 Infarction/Injury probable 1 Present 74 27 21 0 Absent 3410 864 1316 Blank 71 16 16 450 Infarction/Injury possible 1 Present 78 24 14 0 Absent 3406 867 1323 Blank 71 16 16 451 Consider Infarction/Injury 1 Present 174 41 54 0 Absent 3310 850 1283 Blank 71 16 16 LV Mass and LVMI Estimates See Note 23 452-5 ECG estimate of LV Mass 40 - 529 3485 892 1328 Blank 70 15 25 456-9 ECG estimate of LV Mass Index 12 - 295 3485 892 1328 Blank 70 15 25 Likelihood of LVH 460 Probable LVH 1 Present 118 47 58 0 Absent 3367 845 1270 Blank 70 15 25 Minnesota Codes after Visual Adjudication of Coding Differences between Computer Coding and Coding by Human Coders Presence/Absence of Major and Minor ECG abnormalities after adjudication See Note 15 461 Major abnormalities 2 By computer only 170 41 71 1 By computer and visual coder 135 35 45 0 absent 3227 819 1222 Blank 23 12 15 462 Minor abnormalities 2 By computer only 106 26 51 1 By computer and visual coder 355 110 181 0 absent 3091 759 1106 Blank 23 12 15 Likelihood of Myocardial Infarction according to the adjudicated Minnesota Code 463 Diagnostic ECG - Diagnostic Q Waves or ST Elevation with Inverted T Waves See Note 16 2 By computer only 27 7 11 1 By computer and visual coder 6 2 2 0 absent 3499 886 1325 Blank 23 12 15 464 Equivocal ECG - Equivocal Q Waves or ST-T abnormalities 2 By computer only 238 57 103 1 By computer and visual coder 88 19 39 0 absent 3206 819 1196 Blank 23 12 15 Adjudicated MC 1 See Note 18 465-6 Leadgroup L 1.0 - 1.3.3 3532 895 1338 Blank 23 12 15 467-8 Leadgroup F 1.0 - 1.3.3 3532 895 1338 Blank 23 12 15 469-70 Leadgroup V 1.0 - 1.3.2 3532 895 1338 Blank 23 12 15 Highest Code 1.1 in any lead group 11 5 8 Highest Code 1.2 60 8 16 Highest Code 1.3 25 9 3 Highest Code 1.0 3436 873 1311 Adjudicated MC 4 See Note 19 471-2 Leadgroup L 4.0-4.4.0 3532 895 1338 Blank 23 12 15 473-4 Leadgroup F 4.0-4.4.0 3532 895 1338 Blank 23 12 15 475-6 Leadgroup V 4.0-4.4.0 3532 895 1338 Blank 23 12 15 Highest code 4.1 in any leadgroup 1 2 3 Highest code 4.2 20 10 8 Highest code 4.3 13 6 0 Highest code 4.4 15 2 1 Highest code 4.0 3483 875 1326 Adjudicated MC 5 See Note 20 Tape Positions 477-633 Source: Family Questionaire (FQ) Household Screener Questionnaire (HSQ) Position Item Description Counts Source and Code M C P and Notes 477 Leadgroup L 5.0-5.4 3532 895 1338 Blank 23 12 15 478 Leadgroup F 5.0-5.4 3532 895 1338 Blank 23 12 15 479 Leadgroup V 5.0-5.4 3532 895 1338 Blank 23 12 15 Highest code 5.1 in any leadgroup 1 2 5 Highest code 5.2 93 29 34 Highest code 5.3 81 19 41 Highest code 5.4 69 26 42 Highest code 5.0 3288 819 1216 Adjudicated MC 9.2 See Note 21 480 Leadgroup L 9.2.0 - 9.2.2 3532 895 1338 Blank 23 12 15 481 Leadgroup F 9.2.0 - 9.2.2 3532 895 1338 Blank 23 12 15 482 Leadgroup V 9.2.0 - 9.2.2 3532 895 1338 Blank 23 12 15 Highest Code 9.2.2 in any lead group 50 6 26 483-512 Blank Heart rate, basic ECG interval and mean axis data 513-6 Heart rate 33-117 beats per min 3530 895 1338 Blank 25 12 15 517-9 PR interval 39-300 msec 3497 893 1329 Blank 58 14 24 520-2 QRS interval 57-187 msec 3497 893 1329 Blank 58 14 24 523-5 QT Interval 303 - 512 msec 3497 893 1329 Blank 58 14 24 526-9 P axis, frontal plane -165 - 156 degrees 3497 893 1329 Blank 58 14 24 530-3 QRS axis, frontal plane -180 - 174 degrees 3531 895 1338 See Note 24 Blank 24 12 15 534-7 T axis, frontal plane -177 - 179 degrees 3497 893 1329 Blank 58 14 24 538 Rhythm Codes 1 Sinus rhythm 2310 641 864 2 Sinus tachycardia 27 8 19 3 Sinus bradycardia 485 108 140 4 Sinus arrhythmia 397 79 169 5 Atrial fibrillation 7 3 1 6 Ventricular ectopic complexes 14 7 13 7 Supraventricular ectopic complexes 7 7 8 8 Wolff-Parkinson-White 8 1 2 Blank 300 53 137 539-41 Blank 12 Lead ECG amplitude and duration measurements 542-5 P amplitude, positive phase, lead II 0-358 microvolts 3496 893 1329 Blank 59 14 24 546-8 P duration, lead II 0-173 msec 3496 893 1329 Blank 59 14 24 549-52 P amplitude, positive phase, lead V1 0-252 microvolts 3488 893 1329 Blank 67 14 24 553-6 P amplitude, negative phase, lead V1 -227-0 microvolts 3488 893 1329 Blank 67 14 24 557-60 Q or QS amplitude, lead I 0-367 microvolts 3495 893 1329 Blank 60 14 24 561-4 Q or QS amplitude, lead II 0-611 microvolts 3495 893 1329 Blank 60 14 24 565-8 Q or QS amplitude, lead III 0-1877 microvolts 3495 893 1329 Blank 60 14 24 569-72 Q or QS amplitude, lead aVL 0-630 microvolts 3495 892 1328 Blank 60 15 25 573-6 Q or QS amplitude, lead aVF 0-1244 microvolts 3495 893 1329 Blank 60 14 24 577-80 Q or QS amplitude, lead V1 0-2233 microvolts 3487 893 1329 Blank 68 14 24 581-4 Q or QS amplitude, lead V2 0-3401 microvolts 3487 892 1329 Blank 68 15 24 585-8 Q or QS amplitude, lead V3 0-3578 microvolts 3491 892 1329 Blank 64 15 24 589-92 Q or QS amplitude, lead V4 0-2345 microvolts 3496 893 1329 Blank 59 14 24 593-6 Q or QS amplitude lead V5 0-680 microvolts 3495 893 1329 Blank 60 14 24 597-600 Q or QS amplitude, lead V6 0-634 microvolts 3496 893 1329 Blank 59 14 24 601-3 Q or QS duration, lead I 0-48 msec 3495 893 1329 Blank 60 14 24 604-6 Q or QS duration, lead II 0-116 msec 3495 893 1329 Blank 60 14 24 607-9 Q or QS duration, lead III 0-128 msec 3495 893 1329 Blank 60 14 24 610-2 Q or QS duration, lead aVL 0-88 msec 3495 892 1328 Blank 60 15 25 613-5 Q or QS duration, lead aVF 0-116 msec 3495 893 1329 Blank 60 14 24 616-8 Q or QS duration, lead V1 0-136 msec 3487 893 1329 Blank 68 14 24 619-21 Q or QS duration, lead V2 0-116 msec 3487 892 1329 Blank 68 15 24 622-4 Q or QS duration, lead V3 0-112 msec 3491 892 1329 Blank 64 15 24 625-7 Q or QS duration, lead V4 0-96 msec 3496 893 1329 Blank 59 14 24 628-30 Q or QS duration, lead V5 0-52 msec 3495 893 1329 Blank 60 14 24 631-3 Q or QS duration, lead V6 0-52 msec 3496 893 1329 Blank 59 14 24 Tape Positions 634-801 Source: Family Questionaire (FQ) Household Screener Questionnaire (HSQ) Position Item Description Counts Source and Code M C P and Notes 634-7 R amplitude, lead I 33-2235 microvolts 3496 893 1329 Blank 59 14 24 638-41 R amplitude, lead II 0-3799 microvolts 3496 893 1329 Blank 59 14 24 642-5 R amplitude, lead III 0-2759 microvolts 3496 893 1329 Blank 59 14 24 646-9 R amplitude, lead aVR 0-569 microvolts 3496 893 1329 Blank 59 14 24 650-3 R amplitude, lead aVL 0-1913 microvolts 3496 892 1328 Blank 59 15 25 654-7 R amplitude, lead aVF 0-3279 microvolts 3496 893 1329 Blank 59 14 24 658-61 R amplitude lead V1 0 - 2293 microvolts 3488 893 1329 Blank 67 14 24 662-5 R amplitude lead V2 0 - 5432 microvolts 3488 892 1329 Blank 67 15 24 666-9 R amplitude lead V3 0 - 4428 microvolts 3492 892 1329 Blank 63 15 24 670-3 R amplitude lead V4 0 - 5759 microvolts 3497 893 1329 Blank 58 14 24 674-7 R amplitude lead V5 45 - 5275 microvolts 3496 893 1329 Blank 59 14 24 678-81 R amplitude lead, V6 32 - 4449 microvolts 3497 893 1329 Blank 58 14 24 682-4 R duration, lead I 12 - 184 msec 3496 893 1329 Blank 59 14 24 685-7 R duration, lead II 0 - 144 msec 3496 893 1329 Blank 59 14 24 688-90 R duration, lead III 0 - 136 msec 3496 893 1329 Blank 59 14 24 691-3 R duration, lead aVR 0 - 108 msec 3496 893 1329 Blank 59 14 24 694-6 R duration, lead aVL 0 - 184 msec 3496 892 1328 Blank 59 15 25 697-9 R duration, lead aVF 0 - 128 msec 3496 893 1329 Blank 59 14 24 700-2 R duration lead V1 0 - 156 msec 3488 893 1329 Blank 67 14 24 703-5 R duration, lead V2 0 - 132 msec 3488 892 1329 Blank 67 15 24 706-8 R duration lead V3 0 - 132 msec 3492 892 1329 Blank 63 15 24 709-11 R duration, lead V4 0 - 160 msec 3497 893 1329 Blank 58 14 24 712-4 R duration, lead V5 12 - 172 msec 3496 893 1329 Blank 59 14 24 715-7 R duration, lead V6 8 - 172 msec 3497 893 1329 Blank 58 14 24 718-21 S amplitude, lead I 0 - 860 microvolts 3496 893 1329 Blank 59 14 24 722-5 S amplitude, lead II 0 - 1211 microvolts 3496 893 1329 Blank 59 14 24 726-9 S amplitude, lead III 0 - 2336 microvolts 3496 893 1329 Blank 59 14 24 730-3 S amplitude, lead aVR 0 - 2529 microvolts 3496 893 1329 Blank 59 14 24 734-7 S amplitude, lead aVL 0 - 1476 microvolts 3496 892 1328 Blank 59 15 25 738-41 S amplitude, lead aVF 0 - 1773 microvolts 3496 893 1329 Blank 59 14 24 742-5 S amplitude, lead V1 0 - 3482 microvolts 3488 893 1329 Blank 67 14 24 746-9 S amplitude, lead V2 0 - 4898 microvolts 3488 892 1329 Blank 67 15 24 750-3 S amplitude, lead V3 0 - 3766 microvolts 3492 892 1329 Blank 63 15 24 754-7 S amplitude, lead V4 0 - 2687 microvolts 3497 893 1329 Blank 58 14 24 758-61 S amplitude, lead V5 0 - 2542 microvolts 3496 893 1329 Blank 59 14 24 762-5 S amplitude, lead V6 0 - 1434 microvolts 3497 893 1329 Blank 58 14 24 766-8 S duration, lead I 0 - 108 msec 3496 893 1329 Blank 59 14 24 769-71 S duration, lead II 0 - 112 msec 3496 893 1329 Blank 59 14 24 772-4 S duration, lead III 0 - 140 msec 3496 893 1329 Blank 59 14 24 775-7 S duration, lead aVR 0 - 120 msec 3496 893 1329 Blank 59 14 24 778-80 S duration, lead aVL 0 - 116 msec 3496 892 1328 Blank 59 15 25 781-3 S duration, lead aVF 0 - 136 msec 3496 893 1329 Blank 59 14 24 784-6 S duration, lead V1 0 - 144 msec 3488 893 1329 Blank 67 14 24 787-9 S duration, lead V2 0 - 144 msec 3488 892 1329 Blank 67 15 24 790-2 S duration, lead V3 0 - 148 msec 3492 892 1329 Blank 63 15 24 793-5 S duration, lead V4 0 - 116 msec 3497 893 1329 Blank 58 14 24 796-8 S duration, lead V5 0 - 104 msec 3496 893 1329 Blank 59 14 24 799-801 S duration, lead V6 0 - 108 msec 3497 893 1329 Blank 58 14 24 Tape Positions 802-1017 Source: Family Questionaire (FQ) Household Screener Questionnaire (HSQ) Position Item Description Counts Source and Code M C P and Notes 802-5 R' amplitude, lead I 0 - 1085 microvolts 3496 893 1329 Blank 59 14 24 806-9 R' amplitude, lead II 0 - 1114 microvolts 3496 893 1329 Blank 59 14 24 810-3 R' amplitude, lead III 0 - 1481 microvolts 3496 893 1329 Blank 59 14 24 814-7 R' amplitude, lead aVR 0 - 509 microvolts 3496 893 1329 Blank 59 14 24 818-21 R' amplitude, lead aVL 0 - 1240 microvolts 3496 892 1328 Blank 59 15 25 822-5 R' amplitude, lead aVF 0 - 1193 microvolts 3496 893 1329 Blank 59 14 24 826-9 R' amplitude, lead V1 0 - 1869 microvolts 3488 893 1329 Blank 67 14 24 830-3 R' amplitude, lead V2 0 - 2317 microvolts 3488 892 1329 Blank 67 15 24 834-7 R' amplitude, lead V3 0 - 701 microvolts 3492 892 1329 Blank 3 15 24 838-41 R' amplitude, lead V4 0 - 2605 microvolts 3497 893 1329 Blank 58 14 24 842-5 R' amplitude, lead V5 0 - 227 microvolts 3496 893 1329 Blank 59 14 24 846-9 R' amplitude, lead V6 0 - 311 microvolts 3497 893 1329 Blank 58 14 24 850-3 J amplitude, lead I - 178 - 157 microvolts 3496 893 1329 Blank 59 14 24 854-7 J amplitude, lead II - 212 - 194 microvolts 3496 893 1329 Blank 59 14 24 858-61 J amplitude, lead III - 150 - 170 microvolts 3496 893 1329 Blank 59 14 24 862-5 J amplitude, lead aVR - 141 - 178 microvolts 3496 893 1329 Blank 59 14 24 866-9 J amplitude, lead aVL - 154 - 105 microvolts 3496 892 1328 Blank 59 15 25 870-3 J amplitude, lead aVF - 144 - 142 microvolts 3496 893 1329 Blank 59 14 24 874-7 J amplitude, lead V1 - 138 - 359 microvolts 3488 893 1329 Blank 67 14 24 878-81 J amplitude, lead V2 - 179 - 438 microvolts 3488 892 1329 Blank 67 15 24 882-5 J amplitude, lead V3 - 228 - 291 microvolts 3492 892 1329 Blank 63 15 24 886-9 J amplitude, lead V4 - 201 - 367 microvolts 3497 893 1329 Blank 58 14 24 890-3 J amplitude, lead V5 - 181 - 284 microvolts 3496 893 1329 Blank 59 14 24 894-7 J amplitude, lead V6 - 176 - 209 microvolts 3497 893 1329 Blank 58 14 24 898-902 Negative T amplitude, lead I - 515 - 0 microvolts 3496 893 1329 Blank 59 14 24 903-907 Negative T amplitude, lead II - 284 - 0 microvolts 3496 893 1329 Blank 59 14 24 908-12 Negative T amplitude, lead III - 376 - 0 microvolts 3496 893 1329 Blank 59 14 24 913-7 Negative T amplitude, lead aVR - 817 - 0 microvolts 3496 893 1329 Blank 59 14 24 918-22 Negative T amplitude, lead aVL - 416 - 0 microvolts 3496 892 1328 Blank 59 15 25 923-7 Negative T amplitude, lead aVF - 210 - 0 microvolts 3496 893 1329 Blank 59 14 24 928-32 Negative T amplitude, lead V1 - 582 - 0 microvolts 3488 893 1329 Blank 67 14 24 933-7 Negative T amplitude, lead V2 - 1149 - 0 microvolts 3488 892 1329 Blank 67 15 24 938-42 Negative T amplitude, lead V3 - 993 - 0 microvolts 3492 892 1329 Blank 63 15 24 943-7 Negative T amplitude, lead V4 - 1376 - 0 microvolts 3497 893 1329 Blank 58 14 24 948-52 Negative T amplitude, lead V5 - 1233 - 0 microvolts 3496 893 1329 Blank 59 14 24 953-7 Negative T amplitude, lead V6 - 859 - 0 microvolts 3497 893 1329 Blank 58 14 24 958-62 Positive T amplitude, lead I 0 - 883 microvolts 3496 893 1329 Blank 59 24 24 963-7 Positive T amplitude, lead II 0 - 795 microvolts 3496 893 1329 Blank 59 14 24 968-72 Positive T amplitude lead III 0 - 569 microvolts 3496 893 1329 Blank 59 14 24 973-7 Positive T amplitude lead aVR 0 - 413 microvolts 3496 893 1329 Blank 59 14 24 978-82 Positive T amplitude lead aVL 0 - 588 microvolts 3496 892 1328 Blank 59 15 25 983-7 Positive T amplitude lead aVF 0 - 643 microvolts 3496 893 1329 Blank 59 14 14 988-92 Positive T amplitude lead V1 0 - 1359 microvolts 3488 893 1329 Blank 67 14 24 993-7 Positive T amplitude lead V2 0 - 1618 microvolts 3488 892 1329 Blank 67 15 24 998-1002 Positive T amplitude lead V3 0 - 1731 microvolts 3492 892 1329 Blank 63 15 24 1003-7 Postitive T amplitude lead V4 0 - 1564 microvolts 3497 893 1329 Blank 58 14 24 1008-12 Positive T amplitude lead V5 0 - 1417 microvolts 3496 893 1329 Blank 59 14 24 1013-7 Positive T amplitude lead V6 0 - 1075 microvolts 3497 893 1329 Blank 58 14 24 NOTES Note 1: Family Questionnaire Missing A Family Questionnaire was to be completed for each eligible family in a household with sample persons. However, a few Family Questionnaires are missing. Data records for sample persons in families with missing questionnaires are flagged with a code = 1, and all family data are blank. Data records for sample persons in families with a Family Questionnaire are flagged with a code = 2. During the Mexican-American portion of the HHANES survey, a Family Questionnaire continuation booklet containing sample person information was lost for one sample person. Therefore, the sociodemographic data for this sample person are missing. The reference person, family composition, income, residence, and household data for this person were obtained from another person in the household. Note 2: Examination Status Not all sample persons consented to come to a Mobile Examination Center to participate in the examination phase of the survey. In certain rare instances (less than 0.1%), sample persons who came to the Mobile Examination Centers did not participate in sufficient components of the examination to be considered as "examined". This data field contains code = 1 for those persons who participated fully in the examination phase, and code = 2 for those who did not come to the examination center or who did not satisfactorily complete the examination. Note 3: Family Number In HHANES, all household members who were related by blood, marriage, or adoption were considered to be one "family". All sample persons in the same family unit have the same computer-generated family unit code. Note 4: Head of Family Relationship of Sample Person to Head of Family (Pos. 44-45) Each family containing sample persons has a designated "head of family", and the relationship of each sample person to the head of his or her family is coded in tape position 44-45. The first three categories of this variable describe the "head" of three different kinds of families. Code '01' identifies sample persons who lived alone (i.e., "head" of one-person families, no unrelated individuals living in the household). Code '02' identifies sample persons who lived only with unrelated persons. Code '03' identifies sample persons who were "heads" of families containing at least one other person (whether or not the household included additional families unrelated to the sample person). Sociodemographic Data (Pos. 100-131) This data tape includes some sociodemographic data about the head of each sample person's family (Section F). Because there can only be one "head" per family, the data in this section (positions 100-131) are the same for all sample persons in the same family (i.e., with the same family number codes in positions 39-43). If the sample person is the head of his or her family, the data in positions 100-131 are the same as in the corresponding positions in Section E. Note 5: Observed Race "Race" was observed by the interviewer for all sample persons actually seen. Rules for classification of observed race were consistent with those used in the NHANES II and the National Health Interview Survey at that time. The categories were coded as follows: White - Includes Spanish origin persons unless they are definitely Black, Indian or other nonwhite. Black - Black or Negro Other - Race other than White or Black, including Japanese, Chinese, American Indian, Korean, Eskimo Note 6: National Origin or Ancestry The value for national origin or ancestry is based on Item 2c in the Household Screener Questionnaire and was reported by the household respondent for all household members. In the Mexican-American portion of the survey, if "other Latin-American or other Spanish" (code 9) or "Other" (code 0) was recorded and the specified origin was "Spanish-American" or "Spanish (Spain)", a code of 10 or 11, respectively, was assigned. In all three portions of the survey, if more than one category was reported, the first appropriate "Hispanic" code, if any, was assigned (codes 1, 2, 3, 8, 10 or 11 in the Mexican-American portion; codes 6 or 7 in the Cuban-American portion; codes 4 or 5 in the Puerto Rican portion). If none of these codes was recorded, the first category entered was coded. Note 7: Codes for States and Foreign Countries Code State or Foreign Country 001 Alabama 002 Alaska 004 Arizona 005 Arkansas 006 California 008 Colorado 009 Connecticut 010 Delaware 011 District of Columbia 012 Florida 013 Georgia 015 Hawaii 016 Idaho 017 Illinois 018 Indiana 019 Iowa 020 Kansas 021 Kentucky 022 Louisiana 023 Maine 024 Maryland 025 Massachusetts 026 Michigan 027 Minnesota 028 Mississippi 029 Missouri 030 Montana 031 Nebraska 032 Nevada 033 New Hampshire 034 New Jersey 035 New Mexico 036 New York 037 North Carolina 038 North Dakota 039 Ohio 040 Oklahoma 041 Oregon 042 Pennsylvania 044 Rhode Island 045 South Carolina 046 South Dakota 047 Tennessee 048 Texas 049 Utah 050 Vermont 051 Virginia 053 Washington 054 West Virginia 055 Wisconsin 056 Wyoming 060 American Samoa 093 Canada 061 Canal Zone 062 Canton and Enderbury Islands 091 Central America 095 Costa Rica 063 Cuba 064 Dominican Republic 065 El Salvador 062 Enderbury Islands 087 Germany 066 Guam 068 Guatemala 069 Haiti 070 Jamaica 090 Japan 067 Johnston Atoll 080 Mexico 088 Honduras 071 Midway Islands 081 Nicaragua 096 Palestine 097 Austria 098 Lebanon 099 Chile 100 Philippines 101 Brazil 102 Holland 103 Colombia 082 Panama 072 Puerto Rico 092 Saudi Arabia 083 Spain 094 Taiwan 089 Turkey 084 Uruguay 085 Venezuela 073 Ryukyu Islands, Southern 074 Swan Islands 075 Trust Territories of the Pacific Islands (includes Caroline, Mariana and Marshall Island groups) 076 U.S. miscellaneous Caribbean Islands (includes Navassa Islands, Quito Sueno Bank, Roncador Cay, Serrana Bank and Serranilla Bank) 077 U.S. miscellaneous Pacific Islands (includes Kingman Reef, Howland, Baker & Jarvis Islands, and Palmyra Atoll) 086 United States 078 Virgin Islands 079 Wake Island 104 Azores 105 Peru 106 England 107 Vietnam 108 Italy 109 Ecuador 110 North America 111 Surinam 112 Argentina 113 Portugal 114 Trinidad 115 Egypt 116 Sudan 117 British Honduras 118 China 888 Blank but applicable Note 8: Origin Recode In the HHANES, if any household member was identified as "Hispanic" (as defined below), all household members, regardless of origin, were eligible to be selected as sample persons. The national origin recode specifies whether a sample person is considered to be "Hispanic" or "not Hispanic" for purposes of analysis. "Hispanic" is defined as: Mexican-American residing, in selected counties of Texas, Colorado, New Mexico, Arizona, and California; Cuban-American, residing in Dade County (Miami), Florida; or Puerto Rican, residing in the New York City area, including parts of New Jersey and Connecticut. The recode was assigned as follows: A. Southwest portion 1) If the original national origin or ancestry code on the Household Screener Questionnaire was 1, 2, 3, 8, 10, or 11, then National origin recode = 1; 2) If national origin or ancestry was 4, 5, 6, 7, 9, or 0 but the person specified Mexican/Mexicano, Chicano, or Mexican-American self- identification on the Adult Sample Person Questionnaire (question M10), or the person was the biological child of a household member with Recode equal to 1 (as determined by questions A-1-A-11 on the Family Questionnaire), then National origin recode = 1; 3) In all other cases, National origin recode = 2. B. Dade County, Florida portion 1) If the original national origin or ancestry code was 6 or 7, then National origin recode = 1; 2) In all other cases, National origin recode = 2; C. New York City area portion 1) If the original national origin or ancestry code was 4 or 5, then National origin recode = 1; 2) If national origin or ancestry was 1, 2, 3, 6, 7, 8, 9 or 0 but the person specified Boricuan or Puerto Rican self-identification on the Adult Sample Person Questionnaire (question M10), or the person was the biological child of a household member with Recode equal to 1 (as determined by questions A-1/A-11 on the Family Questionnaire), then National origin recode = 1; 3) In all other cases, National origin recode = 2. The national origin recode may be used in analysis in one of two ways: a) Selecting on Recode = 1 will restrict analysis to "Hispanics" only. In this case, in the Southwest portion of the survey, the weighted estimates by age and sex will approximately equal the U.S. Bureau of the Census population estimates of the number of Mexican Americans and a small proportion of other Hispanics assumed to be Hispano in the five Southwest States (Arizona, California, Colorado, New Mexico, and Texas) at the midpoint of the Mexican-American portion of HHANES--March 1983. The weighted estimates of Cuban Americans represents an independent estimate of the number of Cuban Americans in Dade County at the midpoint--February 1984. The weighted estimates of Puerto Ricans represents an independent estimate of the number of Puerto Ricans in the sample counties in New York, New Jersey, and Connecticut at the midpoint of the Puerto Rican portion--September 1984. b) Using Recode greater than 0, that is, all sample persons, will include "Hispanic" and "not Hispanic" persons and the Southwest weighted estimates by age and sex will overestimate the U.S. Bureau of the Census population estimates of Mexican Americans and other Hispanics by about 4.5 percent. In Dade County, using Recode greater than 0 will increase the weighted estimates by about 5.3 percent over that for Cuban Americans only, using Recode greater than 0 for the New York area will increase the weighted estimates by about 9.2 percent over that for Puerto Ricans only. Note 9: Industry and Occupation Code Family Questionnaire questions B-12 through B-15 (see page 117 or 139 of Ref. No. 1 in Section C) identified sample persons 17 years old or older who were in the labor force working for pay at a job or business or who worked without pay in a family business or farm operated by a related member of the household without receiving wages or salary for work performed. Questions B-17 through B-22 provided a full description of sample persons' current or most recent job or business. The detail asked for in these questions was necessary to properly and accurately code each occupation and industry. Interviewers were trained to define a job as a definite arrangement for regular work for pay every week or every month. This included arrangements for either regular part-time or regular full-time work. If a sample person was absent from his or her regular job, worked at more than one job, was on layoff from a job or was looking for work during the two-week reference period, interviewers were trained to use the following criteria to determine the job described: a) If a sample person worked at more than one job during the two-week reference period or operated a farm or business and also worked for someone else, the job at which he or she worked the most hours was described. If the sample person worked the same number of hours at all jobs, the job at which he or she had been employed the longest was entered. If the sample person was employed at all jobs the same length of time, the job the sample person considered the main job was entered. b) If a sample person was absent from his or her regular job all of the two-week reference period, but worked temporarily at another job, the job at which the sample person actually worked was described, not the job from which he or she was absent. c) If a sample person had a job but did not work at all during the two-week reference period, the job he or she held was described. d) If a sample person was on layoff during the two-week reference period, the job from which he or she was laid off, regardless of whether a full- time or part-time job, was described. e) If a sample person was looking for work or waiting to begin a new job within 30 days of the interview, the last full-time civilian job which lasted two consecutive weeks or more was described. The 1980 census of population Alphabetical Index of Industries and Occupations was used in the coding of both industry and occupation. This book has Library of Congress Number 80-18360, and is for sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 for $3.00. Its Stock Number is 003024049-2. Note 10: Health Insurance In the Health Insurance section of the Family Questionnaire, up to three separate health insurance plans could be reported for a family. Each sample person could have been covered by any combination of the three, or by none at all. In order to simplify the health insurance coverage data, the information on all reported plans was combined to a single variable for each sample person, i.e., whether or not the person is covered by any plan (position 74). For all persons covered by at least one plan, information on the type of coverage is then indicated; position 75 specifies whether any of the sample person's plans pays hospital expenses and position 76 specifies whether any of the sample person's plans pays doctor's or surgeon's bills. For all sample persons who were not covered by Medicare or any health insurance plan, the reasons for not being covered were ascertained. Positions 77-78 contain the main or only reason reported. For persons with one or more additional reasons, the first (lowest) code entered on the questionnaire was coded in positions 79-80. Note 11: Per Capital Income Per capita income was computed by dividing the total combined family income by the number of people in the family. Note 12: Poverty Index The poverty index is a ratio of two components. The numerator is the midpoint of the income bracket reported for each family in the Family Questionnaire (E-11). Respondents were asked to report total combined family income during the 12 months preceding the interview. The denominator is a poverty threshold which varied with the number of persons in the family, the adult/child composition of the family, the age of the reference person, and the month and the year in which the family was interviewed. Poverty thresholds published in U.S. Bureau of the Census reports* are based on calendar years and were adjusted to reflect differences caused by inflation between calendar years and 12-month income reference periods to which question E-11 referred. Average Consumer Price Indexes for all Urban consumers (CPI-U) for the calendar year for which the poverty thresholds were published (see table below) and for the 12 months representing the income reference period for the respondent were calculated. The percentage difference between these two numbers represents the inflation between these two periods and was applied to the poverty threshold appropriate for the family (based on the characteristics listed above). For example, for a family interviewed in November 1983, the 1982 poverty threshold was updated to reflect inflation by multiplying by the percent change in the average CPI-U for the 12-month reference period, which would have been November 1982 through October 1983, over the calendar year January through December 1982, in this example. To compute poverty indexes, the midpoint of the total combined family income bracket was divided by the updated poverty threshold. Members of families with incomes equal to or greater than poverty thresholds have poverty indexes equal to or greater than 1.0 and can be described as "at or above poverty"; those with incomes less than the poverty threshold have indexes less than 1.0 and can be described as "below poverty". Poverty thresholds used were computed on a national basis only. No attempt was made to adjust these thresholds for regional, State, or other variations in the cost of living. None of the noncash public welfare benefits such as food stamp bonuses were included in the income of the low income families receiving these benefits. U.S. Bureau of the Census, Current Population Reports, Series P-60, No. 138, "Characteristics of the Population Below the Poverty Level: 1981", U.S. Government Printing Office, Washington, D.C., March 1983. U.S. Bureau of the Census, Current Population Reports, Series P-60, No. 144, "Characteristics of the Population Below the Poverty Level: 1982", U.S. Government Printing Office, Washington, D.C., March 1984. Average Consumer Price Index, all Urban consumers (CPI-U), U.S. city average, 1981-84 Y E A R ________________________________________________ MONTH 1981 1982 1983 1984 January 260.5 282.5 293.1 305.2 February 263.2 283.4 293.2 306.6 March 265.1 283.1 293.4 307.3 April 266.8 284.3 295.5 308.8 May 269.0 287.1 297.1 309.7 June 271.3 290.6 298.1 310.7 July 274.4 292.2 299.3 311.7 August 276.5 292.8 300.3 313.0 September 279.3 293.3 301.8 October 279.9 294.1 302.6 November 280.7 293.6 303.1 December 281.5 292.4 303.5 Average 272.4 289.1 298.4 Source: U.S. Department of Labor, Bureau of Labor Statistics Note 13: Size of Place and SMSA Codes for size of place and SMSA were obtained from U.S. Bureau of the Census summary tape files (STF 1B). A place is a concentration of population. Most places are incorporated as cities, town, villages or boroughs, but others are defined by the U.S. Bureau of the Census around definite residential nuclei with dense, city- type street pat terns, with, ideally, at least 1,000 persons per square mile. The boundaries of Census defined places may not coincide with civil divisions. A Standard Metropolitan Statistical Area (SMSA) is a large population nucleus and nearby communities which have a high degree of economic and social integration with that nucleus. Generally, an SMSA includes one or more central cities, all urbanized areas around the city or cities, and the remainder of the county or counties in which the urbanized areas are located. SMSAs are designated by the Office of Management and Budget. The same place size and SMSA codes were assigned to all persons in the same segment (for the definition of segments see Ref. No. 1 in Section C). In a few cases segments were divided by place boundaries. In these cases codes were assigned after inspecting segment maps. If the segment was predominantly in one place, then the place code for that place was used. If the segment was approximately evenly divided, the code for the larger place was used. Note 14: Home Heating Questions E-3 through E-6, pertaining to the main fuel and equipment used for heating the home, appear to have codes which are inconsistent. It has been verified that these are codes that were recorded on the original document; that is, codes that appear inconsistent were not incorrectly keyed. Note 15: Major and Minor ECG Abnormalities Minnesota Code Comments MAJOR ABNORMALITIES Major Q, QS waves 1.1 or 1.2 except 1.2.8 Highest code in any leadgroup ST depression 4.1 or 4.2 Negative T waves 5.1 or 5.2 Complete AV block 6.1 Not coded by computer WPW pattern 6.4 Artificial pacemaker 6.8 Not coded by computer Ventricular conduction 7.1 or 7.2 or 7.4 DEFECTS Atrial fibrillation/flutter 8.3 Rhythm code 05 ST elevation 9.2 MINOR ABNORMALITIES Minor Q waves 1.2.8 or 1.3 High R waves 3.1 or 3.3 Any 3.1 or 3.3 code Minor ST codes 4.3 or 4.4 Minor T wave codes 5.3 or 5.4 Prolonged PR interval 6.3 RR' in V1 or V2 7.3 or 7.5 Left anterior fascicular block 7.7 Note 16: Criteria for Diagnostic and Equivocal ECGs Probable MI - Diagnostic Q waves - one of: Minnesota Code 1.1.1 through 1.2.7 for Q and QS Patterns. Minnesota Code 1.2.8 through 1.3.6 for Q and QS Patterns and a major or T-wave inversion (Minnesota Code 5.1 or 5.2) Possible MI - Equivocal Q waves: Q and QS Patterns 1.2.8 through 1.3.6 Note 17: Likelihood of Left Ventricular Hypertrophy (LVH) According to the Minnesota Code Probable LVH: Code 3.1 with code 5.1 or 5.2 or 5.3 Possible LVH: Code 3.1 without code 5.1 and 5.2 and 5.3, OR Any code 3.3 Note 18: Range of Values for MC 1 The range indicated here does not correspond to the actual severity levels of MC 1. See Appendix 1. Note 19: Range of Values for MC 4 The range indicated here does not correspond to the actual severity levels of MC 4. See Appendix 1. Note 20: Range of Values for MC 5 The range indicated here does not correspond to the actual severity levels of MC 5. See Appendix 1. Note 21: Range of Values for MC 9.2 Minnesota Code 9.2 categorizes ST elevation into two levels; 9.2.0 (ST elevation absent), and 9.2.2 (ST elevation present). Note 22: Cardiac Infraction/Injury Score (CIIS) This ECG coding scheme was developed as a measure of the likelihood of myocardial infarction on a continuous scale. The following thresholds for the score define the likelihood of infarction in a decreasing order: Probable infarction CIIS > or = 20 Possible infarction 15 < or = CIIS < 20 Consider infarction 10 < or = CIIS < 15 These thresholds correspond to the estimated specificity levels of 98%, 95% and 90%. Reference: Rautaharju PM, Warren J, Jain U, Wolf HK and Nielsen CL. Cardiac Infarction/Injury Score: An electrocardiographic coding scheme for ischemic heart disease. Circulation 64:249-258,1981. Note 23: Coefficients of Regression Equation used for ECG Estimat COEFFICIENTS FOR THE REGRESSION EQUATION USED FOR ECG ESTIMATION OF LEFT VENTRICULAR MASS (LVM) AND LEFT VENTRICULAR MASS INDEX (LVMI) White and Black Men Variables LVM LVMI R amplitude in V5 (microvolts) 0.0217 0.0100 Q or S amplitude in V1 (microvolts)* 0.0338 0.0203 Q or S amplitude in III (microvolts)* 0.0600 0.0287 Negative T amplitude in V6 (microvolts) 0.3158 0.1819 Positive T amplitude in aVR (microvolts) -0.2958 -0.1482 QRS duration (msec.) 1.8204 1.0485 Intercept -58.5098 -36.4290 White Women Variables LVM LVMI R amplitude on aVL (microvolts) 0.0320 -- R amplitude in V5 (microvolts) 0.0233 0.0178 Q or S amplitude in V5 (microvolts)* 0.0693 0.0528 Q or S amplitude in I (microvolts)* -0.1545 -0.1128 Positive T amplitude in V1 (microvolts) 0.1122 0.1075 Negative T amplitude in aVF (microvolts) -- 0.1701 Positive T amplitude in V6 (microvolts) -0.1236 -0.0939 Intercept 134.7777 88.4357 Black Women Variables LVM LVMI R amplitude in aVL (microvolts) -- 0.0216 R amplitude in I (microvolts) 0.0507 -- (R amplitude in V6 + S amplitude in V2) (microvolts) 0.0235 0.0184 R amplitude in V1 -0.0507 -- R amplitude in V2 (microvolts) -- -0.0143 Q or S amplitude in V6 (microvolts)* -0.0980 -0.0693 Negative T amplitude in aVL (microvolts) -- 0.199 Negative T amplitude in I (microvolts) 0.5225 -- QRS duration (msec.) 1.8478 0.7460 Intercept -90.7136 -22.3064 * whichever is larger The following limits for LVMI are taken to indicate the presence of probable left ventricular hypertrophy: Males Females Probable LVH LVMI > 131g/m2 LVMI > 110g/m2 Note 24: Frontal Plane QRS Axis The algorithm used for QRS axis determination provides a more accurate estimation of the mean frontal plane axis than the approximation used in Minnesota Code 2 according to the conventional visual measurement. The algorithm used for the QRS axis determination is also used for P and T axis calculation. Values of QRS integrals (net QRS 'areas', A) determined from the six limb leads are used for the mean frontal plane QRS axis calculation. Three separate axis angle (ANG) values are calculated from three pairs of limb leads. The lead vectors of these three pairs of leads are assumed to be orthogonal according to the Einthoven's equilateral triangle approximation, and the relative strength of the lead vectors of leads aVR, aVL and aVF are assumed to be 3/2 times the lead vector strengths of leads I, II and III. Consequently, the augmented unipolar limb leads are scaled by factor 1.16 in these pairwise calculations of the three angles ANG(1), ANG(2) and ANG(3). ANG(1) = ARCTG (1.16 x A(aVF), A(I)), ANG(2) = ARCTG (A(II), 1.16 x A(aVF)), ANG(3) = ARCTG (1.16 x A(aVR), A(III)) + 120 In case the three values are reasonably consistent, the final mean frontal plane axis is taken as the mean value of these three separate angle determinations. Several inconsistency checks are performed, and if abnormally large discrepancies are found, the angle is termed 'undetermined'. QRS axis values can be used to identify abnormal axis deviations. Please refer to Appendix 2, Category 2. QRS Axis (criteria for Category 2). NOVACODE ECG MEASUREMENT AND CLASSIFICATION PROGRAM P.M. Rautaharju Introduction The Novacode ECG measurement and classification program incorporates the classic Minnesota Code (1) and two more recent ECG classification modules designed for improved estimation of the likelihood of myocardial infarction (MI) (2) and left ventricular hypertrophy (LVH) (3). This document describes these two new ECG classifiers. It also contains the first published report which documents the Minnesota Code ECG wave measurements and ECG codes using logic operators and a glossary of symbols and concise definitions of the variables used in ECG coding. This was done in order to reduce the ambiguities in traditional Minnesota Code definitions for the fairly complex visual ECG wave measurement and classification "rules" when expressed in everyday English (4). There have been substantial technological improvements in ECG acquisition since the introduction of the Minnesota Code nearly three decades ago and the development of its guidelines for ECG wave identification and measurement. These guidelines were devised for visual ECG measurement of ECG complexes primarily from single-channel ECG tracings. The "majority rule" was introduced in order to cope with beat-to-beat waveform variations caused, for instance, by respiration and muscle noise. Computer processing of ECGs has considerably enhanced ECG signal quality. It has also introduced the use of a single representative signal-averaged ECG complex for wave measurements instead of beat-by-beat measurements from single-channel ECGs as traditionally done in visual ECG coding. Another recent major change has been the introduction of simultaneous acquisition of all independent components of the standard 12-lead ECG, a new procedure which was used, for instance, in HHANES. This has greatly improved the precision and accuracy of ECG interval measurements, and it has subsequently reduced the variance of ECG interval measurements compared to single-channel measurements used, for instance, in NHANES I. These technological improvements have also necessitated certain revisions in ECG wave definitions and measurement procedures which deviate from the old Minnesota Code guidelines. These differences between the traditional visual Minnesota Code measurements and measurements performed by the Novacode program will be discussed in this document. A record library composed in 1015 visually coded ECGs was initially available when the Novacode ECG program development was initiated over a decade ago. An iterative program performance optimizing scheme was employed to minimize the discrepancies between computer and visual measurements. Thus, the choice for various thresholds and other program parameters in various wave detection algorithms was made iteratively within a wide range of possible values. The subsequent program development was performed in connection with the Multiple Risk Factor Intervention Trial (MRFIT) which required a parallel visual and computer ECG coding of nearly 100,000 ECGs (5). Key Characteristics of the Novacode Program The basic ECG measurement module of the Novacode program was initially developed for processing of both rest and exercise ECGs (6-8). The measurement program was probably the first ECG program which relied on measurements made from representative complexes obtained after clustering and selective averaging instead of making measurements from individual complexes on a beat-to-beat basis. This has resulted in an excellent noise tolerance. A recent revision with an advanced baseline drift correction algorithm has further improved the robustness of the program. QRS Wave Detection A. Initial search interval for QRS waves QRS wave detection is performed on each set of simultaneously recorded ECG leads by detecting the earliest and the latest sign of ventricular excitation. These time coherent demarcation lines are called reference lines C(QRS onset) and D(QRS offset) See figure 1. According to the Minnesota Code measurement rules, the begin ning of the Q wave or the initial R wave in any given individually considered lead does not necessarily coincide with time line C, and a refined wave detection algorithm is needed (for excluding so called isoelectric segments) to satisfy the initial QRS wave measurements for the Minnesota Code. The initial search interval for QRS wave detection extends from C-16 msec to D+16 msec. B. Differences between visual and computer measurements The Minnesota Code definitions for measurement and classification were formulated as instructions to human coders. In spite of serious attempts at consistent, clear and explicit definitions when the Minnesota Code was initially developed, numerous ambiguities became evident when a systematic effort was made to formulate the definitions following rigid rules of logic. These ambiguities in measurement and coding partially explain the relatively poor repeatability in visual coding of some of the categories. There are several minor but important differences from visual measurement rules introduced in the Novacode program. In view of the importance of these differences in practical applications, they will be listed next and discussed in some detail. 1. Baseline: The Minnesota Code measurement rules suggest that QRS and ST amplitude measurements be made "by measuring the vertical distance between the upper edge of the trace at the beginning of the QRS interval and the upper edge of the trace at the apex of the deflection" (9). (Negative deflections are measured following the lower edge of the tracing.) This baseline definition is valid also for all ST measurements. However, the precise definition of the 'upper edge of the trace at the beginning of QRS interval' is missing. The Novacode program defines the baseline as a smoothed 18 msec. segment centered at 16 msec before the beginning of QRS. This choice, instead of the beginning of QRS, resulted in improved stability and reduced vulnerability to possible minor errors in defining the QRS onset. The Novacode program uses this common baseline for amplitude measurements for all Minnesota Code categories with the exception of Code 5. The inertia of tradition, rather than any sound scientific or technical reason, compelled us to accept, as an interim compromise, the TP baseline for T amplitude measurements. 2. Majority Rule: The Novacode program makes measurements from the average complex of the members of the majority cluster formed on the basis of QRS duration, amplitude and RR interval data. The selection of a 'model com plex' rather than the average would perhaps be closer to the majority rule followed by visual codes. On the other hand, the choice of the average complex results in a substantially improved consistency and repeatability particularly if the record quality is borderline. There are two important deviations from the majority rule in visual coding which may result in discrepancies in comparison with the measurements made from the average complex. Firstly, the R wave amplitudes for Code 3 are most commonly measured from the 'second to the last good beat' in a given lead group. Secondly, the presence of an initial R wave exceeding 25microvolts in any QRS complex rules out the Q and QS waves in that lead. 3. Threshold for Initial R Wave Amplitude: The threshold for the detection of initial R waves from the averaged QRS complex was reduced to 20 microvolts from the 25 microvolts used by visual coders for single beat measurements. This improved the agreement between visual and computer coding of Q waves. 4. Code 2. Frontal Plane QRS Axis: The coding categories have been rearranged into mutually exclusive groups. A new category is included for borderline left axis deviation. 5. CODE 4. J Amplitude, ST Slope: The classic definition of the J point well illustrates the difficulties encountered in the attempts to formulate rapid logic rules for computer coding of ECGs. "The term 'ST-T junction' or 'J' (RS-T junction) should be used to indicate the point or shoulder which marks the end of the QRS complex, the point when steep slopes of the QRS deflections are more or less abruptly replaced by the more gradual slopes which precede or comprise the first limb of the T wave (9)." This definition is further qualified by the following rules: "If more than one change of slope makes J point determination arbitrary, choose the latest one, or if ambiguous, do not code." "If there is no clear S-T demarcation, include the whole length of the line as S-T coding 4.3." An analysis of the logic followed by visual coders in identifying J amplitude measurement from single-lead ECG complexes for Code 4 reveals a variety of subjective choices depending on combinations of relative changes in the linearity, slope and amplitude values in the proximity of the apparent end of QRS. In addition, the choice of the ST depression measurement point is different, if the ST segment is 'U-shaped' or 'bell-shaped' without a clearly delineated inflection point suggesting the presence of a definable J point. In fact, three different J points may be chosen to designate the end of QRS, ST depression and ST elevation in a given lead. For instance, in case of a bell-shaped ST segment, the visual coders compare the amplitude of the J point with the amplitude of the inflection point in the beginning of the T wave or T wave nadir amplitude. If J amplitude is larger than that of the latter two amplitudes the whole ST slope is considered negative and the maximum point of the ST segment is measured for coding of ST depression. Our attempts to define functionally the logic for these diverse definitions of the J point were not entirely successful, to a large extent because of the difficulty in reproducing visual measurements for optimization tests. Therefore, instead of J amplitude, it was decided to select the maximum ST amplitude within the initial part of the ST segment in the interval from QRS end +6 msec. to QRS end + 25 msec. for coding of ST depressions. Best agreement for ST slope measurements was reached by identifying the minimum 50 msec. slope in the interval from QRS end + 25 msec. to one-third of the total ST-T segment length. Finally, a bell-shaped ST segment is considered to have a negative slope for classification purposes if T wave nadir (amplitude of the first negative wave in the ST-T complex) is more negative than the J amplitude. Admittedly, these decision rules are still more complex than desirable. However, they were able to reproduce with adequate accuracy visual coding in categories 4.1 and 4.2. The most recent version of the Novacode ECG program is described in detail in (10). Fig. 1. Representative average complexes of the eight simultaneously recorded ECG leads. Time demarcation lines A to E identify common reference time points for the beginning and end of atrial depolarization (A-B), ventricular depolarization (C-D) and the end of ventricular depolarization (E). References 1. Blackburn H, Keys A, Simonson E, Rautaharju PM and Punsar S: The electrocardiogram in population studies: A classification system. Circulation 21:1160-1175, 1960. 2. Rautaharju PM, Warren J, Jain U, Wolf HK and Nielsen CL: Cardiac Infarction Injury Score: An electrocardiographic coding scheme for ischemic heart disease. Circulation 64:249-256, 1981. 3. Rautaharju PM, LaCroix AZ, Savage DD, Haynes S, Madans JH, Wolf HK, Hadden W, Keller J, Cornoni-Huntly J: Electrocardiographic estimate of left ventricular mass vs. radiographic cardiac size and the risk of cardiovascular disease mortality in the epidemiologic follow-up study of the First National Health and Nutrition Examination Survey. Am J Cardiol 62: 59-68, 1988. 4. Prineas RJ, Crow RS and Blackburn H: The Minnesota Code Manual of Electrocardiographic findings. Standards and Procedures for Measurement and Classification. John Wright. PSG Inc. Boston, Bristol, London, 1982. 5. Rautaharju PM, Broste SK, Prineas RJ, Eifler WJ, Crow RS and Furberg CD: Quality control procedures for the resting electrocardiogram in the Multiple Risk Factor Intervention Trial. J. Controlled Clinical Trials 7:46s-65s, 1986. 6. Wolf HK, MacInnis PJ, Stock S, Helppi RK and Rautaharju PM: The Dalhousie Program: A comprehensive analysis program for rest and exercise electrocardiograms. In: Computer Application on ECG and VCG Analysis. Chris Zyweitz and B. Schneider, Eds. North-Holland Publishing Co., Amsterdam - London, 1973; pp231-240. 7. Wolf HK, MacInnis PJ, Stock S, Helppi RK and Rautaharju PM: Computer analysis of rest and exercise electrocardiograms. Comput Biomed Res 5:329-346, 1972. 8. Wolf HK and Rautaharju PM: An on-line program for acquisition and analysis of resting and exercise electrocardiograms. In: Vectorcardiography I. I. Hoffman, Ed. North-Holland Publishing Co., 1971, pp231-236. 9. Rose GA and Blackburn H (1968): Cardiovascular Survey Methods. World Health Organization Monograph Series No. 56. 10. Rautaharju PM, MacInnis PJ, Warren JW, Wolf HK, Rykers PM, Calhoun HP: Methodology of ECG interpretation in the Dalhousie Program: NOVACODE ECG classification procedures for clinical trials and population health surveys. Methods of Information in Medicine 29:362-374, 1990. *For Figure 1 "Representative average complexes of the eight simultaneously recorded ECG leads" contact User Support at (404)332-4569. APPENDIX 1 - GLOSSARY OF SYMBOLS AND DEFINITIONS Glossary of Symbols and Definitions used for Computer Coding of the Minnesota Code 1. Definition of logical operators ~ Denotes logical negation. Thus, if ~ C is true, then C is false. + Denotes logical disjunction, OR. Thus, if C + D is true, then either C or D is true or both are true. C + D is false only if both C and D are false. . Elevated period between symbols denotes logical conjunction AND, implying that both C and D are true if C D is true. If either C or D is false or both are false, then C D is false. *For symbols associated with and examples of extended logical conjunction and disjunction please contact User Support at (404)332-4569. 2. Definitions of global variables common to all ECG waves. Baseline (L) A reference amplitude for ECG voltage and time integral measurements. This reference amplitude is taken as the smoothed value at 16 msec. preceding time demarcation line C (QRS onset). Smoothing is done by fitting a quadratic polynomial spanning 18 msec. and centered at C-16 msec. by the method of least squares. Time Time coherent reference points for all simultaneously recorded demarcation ECG leads. These demarcation points identify the beginning lines and end of the P and QRS waves and the end of the T wave, and they are used for global ECG interval measurements (See figure 1). Line A Time demarcation point denoting the beginning of atrial excitation (onset of P wave) Line B Time demarcation point denoting the latest sign of atrial excitation (end of P wave) Line C Time demarcation point denoting the earliest sign of ventricular excitation (QRS complex) Line D Time demarcation point denoting the latest sign of ventricular excitation (end of QRS or the J point) Line E Time demarcation point denoting the latest sign of ventricular repolarization (end of T wave) 3. Symbols for codable ECG waves INTDEFL Intrinsicoid deflection; time interval from the beginning of QRS to the peak of the largest codable positive QRS wave in lead L PL Denotes existence of a codable P wave in lead L PNEGL Absolute amplitude of negative phase of the P wave in lead L PPOSL Absolute amplitude of positive phase of the P wave in lead L QL Denotes the existence of a codable Q wave in lead L QAMPL Absolute amplitude of Q wave in lead L QDURL Duration of Q wave in lead L QRL Ratio of QAMP to RAMP in lead L QRSDURL Duration of QRS in lead L QSL Existence of a codable QS wave in lead L RL Existence of a codable R wave in L RAMPL Absolute amplitude of the largest codable positive QRS wave (R or R') in lead L R1L Absolute amplitude of first codable positive QRS wave in lead L R2L Absolute amplitude of R' wave in lead L (largest positive wave following R1, if any) RSL Ratio of RAMP to SAMP in lead L RTL Ratio of RAMP to TAMP in lead L R'L Existence of a codable R' wave in lead L SL Existence of a codable S wave in lead L SAMPL Absolute amplitude of S wave in lead L STJL Amplitude of J point (QRS offset) in lead L STML Average amplitude of the initial 3/8th of ST-T in lead L STSL Slope of least squares line fitted in the window (D + 10 msec.) to (D + 76 msec.) in lead L SLMINL Minimum of all 50 msec. forward slopes calculated in the ST segment excluding the initial 25 msec. in lead L SL00L Initial 50 msec. forward slope of ST segment at time point D in lead L SL25L 50 msec. forward slope of ST segment at (D + 25 msec.) in lead L SL50L 50 msec. forward slope of ST segment at (D + 50 msec.) in lead L STMAXL Maximum amplitude of the initial 3/8th portion of the smoothed ST-T segment in lead L STMINL Minimum amplitude of the smoothed ST segment in lead L 4. Special definitions for ST-T waves for computer coding C(t) Convexity function; a generating function used for identification of positive, negative, flat and multiphasic ST-T waves. The convexity function is computed from the smoothed ST-T complex S(t) according to the following formula: C(t) = 1/2{S(t - 50) + S(t + 50)} - S(t) for all sample points t within the ST-T complex of each lead AW1L Amplitude of first wave of ST-T wave train in lead L AW2L Amplitude of second wave of ST-T wave train in lead L AW3L Amplitude of third wave of ST-T wave train in lead L TFLATL T wave which does not meet criteria for Positive (P) or negative (N) T waves in lead L TNEGL Minimum amplitude (A) of negative wave within terminal 5/8th of ST-T in lead L with A < -30 microvolts and convexity C > 30 microvolts TPOSL Maximum amplitude of a positive wave within terminal 5/8th of ST-T in lead L with amplitude A > 5 microvolts and convexity C < - 30 microvolts 5. Logical variable for ST-T wave identification FFFL Logical variable denoting that no ST-T waves were detected in lead L FFNL Logical variable denoting detection of only one wave which was negative FFPL Logical variable denoting detection of only one wave which was positive FPNL Logical variable denoting detection of two waves which were in temporal order positive and negative FNPL Logical variable denoting detection of two waves which were in temporal order negative and positive NPNL Logical variable denoting detection of three waves which were in temporal order negative, positive and negative PNPL Logical variable denoting detection of three waves which were in temporal order positive, negative and positive 6. Symbols for measurement units V Volt mV millivolt V x 10-3 microvolt V x 10-6 sec. second msec. millisecond; sec. x 10-3 APPENDIX 2 - MINNESOTA CODE, 1982 VERSION Category 1. Q and QS Patterns Data only available on hard copy. Please contact User Support at (404)332-4569 for further information. Category 2. QRS Axis Data only available on hard copy. Please contact User Support at (404)332-4569 for further information. Category 3. High Amplitude R Waves and Related Items Data only available on hard copy. Please contact User Support at (404)332-4569 for further information. Category 4. ST Abnormalities Data only available on hard copy. Please contact User Support at (404)332-4569 for further information. Category 5. T wave abnormalities Data only available on hard copy. Please contact User Support at (404)332-4569 for further information. Category 6. A-V Conduction Data only available on hard copy. Please contact User Support at (404)332-4569 for further information. Category 7. Ventricular Conduction Defections Data only available on hard copy. Please contact User Support at (404)332-4569 for further information. Category 8. Arrhythmias Data only available on hard copy. Please contact User Support at (404)332-4569 for further information. Category 9. Miscellaneous Codes Data only available on hard copy. Please contact User Support at (404)332-4569 for further information.
