SN:2836 Dietary and Nutritional Survey of British Adults, 1986-1987 This dataset was deposited with the Archive as six SPSS files, containing summary information, and four SIR databases. The SIR databases consist of a number of SIR records, each containing specific types of information, nutrient quantities for example. These SIR records are extracted by the Archive to create files which correspond to these records. The last one or two digits of a filename indicate which SIR record the data was extracted from. For example, the file 2836a1 contains data from SIR record 1. ----------------------------------------------------------------------------- Useful background information There were ca 2200 Adults with satisfactorily completed 7-day food diaries (out of some 2600 with usable home interview questionnaires). They were spread fairly evenly over a whole year to provide a seasonally unbiased typical and representative year for the sample, with ca 550 in each wave of three months. As there was no intention of following INDIVIDUALS through the year, there is no association whatsoever other than matching digits between respondents 10103 and 20103. The serial number is just that - i.e. serial, but it has some structure. 2-01-03 means the third person sample in cluster number 1 in wave number 2. Users should also recognise that towns or districts represented by the area codes - which can range from 1 to 30 - changed from wave to wave. That is, the district numbered area 1 in wave one is DEFINITELY NOT the same as that numbered 1 in wave 2. We sample areas for our NDNS "without replacement". ------------------------------------------------------------------------------ Additional information follows: Record 1 For variable DAYOFWK the following coding applies: 1 = Sunday 2 = Monday 3 = Tuesday 4 = Wednesday 5 = Thursday 6 = Friday 7 = Saturday 8 = An extra weekend day 9 = An extra weekday day 8 and 9 are for unspecified weekend and weekday days. ------------------------------------------------------------------------------ Record 3: Exyyyy These are binary (o, 1) flags to indicate an unusually high level of the nutrient or quasi-nutrient. Zero is the normal condition. It is evident that each Exyyyy corresponds to a "nutrient" yyyy. Our contractors, OPCS, ran a number of checks on the diary records after data entry, which included comparison of apparent nutrient intake against prior expectations of normal behaviour. These would inevitably be occasionally exceeded for valid reasons. For example, if the edit/validate procedure finds a "too high" level of IRON according to some criterion which is NOT IN THIS DATASET the error condition should be negated if EXIRON = 1. (NOTE: The depositor thinks that not much can be done with this information other than to be aware of the original purpose of these flags). ------------------------------------------------------------------------------ Record 6 SEX 1 Male 2 Female ORIGIN Ethnic group by appearance 1 Asian 2 Coloured 3 White 4 Not seen EMPLOY Informant employ status 1 Full time 2 Part time 3 None ------------------------------------------------------------------------------ Record 11 HB Haemat HCT Haemoglobin (g/d1) MCV Mean Corpuscular Volume MCHC Mean Corpuscular Haemoglobin Conc. (g/d1) B12 Vitamin B12 (mg/1) RCF Red Cell Folate (mg/1) VITB2 Erythrocyte Glutathone Reductose Activ C VITB6 Vitamin B6 ------------------------------------------------------------------------------ Record types 9 and 15 In record 9, "Daily Tot", all the totals (labelled Dt1 total nutrient x from food + supplement - 7 days a1 average nutrient x from food + supplement - (t1/7) t2 total nutrient x from food only - 7 days a2 average nutrient x from food only (t2/7) t3 total nutrient x from food only, away from home only - 7 days ------------------------------------------------------------------------------ Record 18: QGTnnz & QGAnnz where nn = 2 digit number z = 1 letter Refer to appendix H in the first Report, Gregory J et al, HMSO 1990. OPCS's schema description mentions (a) "new foodgroups" and (b) "food[s] not in new foodgroup." (a) "New foodgroup" was their understanding - at the time - of what Appx H calls "sub(sidiary) foodgroups". These are "daughters" of the "main" food groups, numbered nn=1, 51. They were identified by tag letters A, B, C and so on. (b) Sub-food groups tags "R" represent foods not otherwise assigned. "R" was also used if the group and sub group were identical. The two series start as follows: Food Groups: 1 = Pasta, Rice & other miscellaneous cereals 2 = White bread Sub(sidiary) Food Groups 1A = Pasta ---> QGT/AO1A 1B = Rice ---> QGT/AO1B 1R = Other Cereals ---> QGT/AO1R 2R = White Bread ---> QGT/AO2R The ordering of variables in Record 18 is nearly consistent with that in Appx.H, except that the Diet Soft Drinks were labelled "46B" for some reason - and "Other S/D" were "46A". (Users should be wary of using the "foodgroup" identifier "FOODGROUP" mentioned in Record 3's schema as depositors believes some values were re-assigned or over ridden by SIR procedures with food-code-dependent special handling which is not apparent in the data and not 100% consistent with this variable). The values assigned to QGT/Annz relate to 7-day total and average food consumption. Depositor indicates that OPCS - again through special-cased procedure coding - allowed for dilution for some beverages and soft drinks - for example, a factor such as 5 would be applied for a mention of orange squash concentrate whereas, for a mention of orange squash as drunk, the quantity reported would remain unaltered. ----------------------------------------------------------------------------- Date and Time variables When data is extracted from the SIR databases variables declared as DATE or TIME variables are treated in a special way. DATE variables are stored as integers. This integer is the date according to the Gregorian calendar, which is the number of days since October 15, 1582. For example, 2nd November 1986 would be stored as 147577. TIME variables are also stored as integers. The number stored is the number of seconds since midnight. For example 20:42:00 would be stored as 74520. The DATE and TIME variables in this study are as follows: DATE VARIABLES Record 1 DATE Record 5 DATEINT PQ8DATE PQ18DATE PQ22STAD PQ22ENDD PQ23DATE Record 10 COLLST TIME VARIABLES Record 5 PQ11TIME PQ22STAT PQ22ENDT Record 7 STARTWK1 ENDWORK1 STARTWK2 ENDWORK2 STARTWK3 ENDWORK3 STARTWK4 ENDWORK4 STARTWK5 ENDWORK5 STARTWK6 ENDWORK6 STARTWK7 ENDWORK7 STARTWK8 ENDWORK8 STARTWK9 ENDWORK9 STRTWK10 ENDWRK10 STRTWK11 ENDWRK11 STRTWK12 ENDWRK12 STRTWK13 ENDWRK13 Record 14 URINPER ------------------------------------------------------------------------------