Support #396

Health Condition Variables - heart attack and stroke

Added by Lea Samek over 8 years ago. Updated over 8 years ago.

Gundi Knies
Biomarkers and Genetics
Start date:
% Done:



I am currently interested in whether a respondent had any of the provided health conditions in one or more of the 4 waves provided. Therefore, I am using hconds# to check if someone, who has ever been diagnosed with something (hcond#) still has it in the following wave. I also use hcondns# to check if someone was newly diagnosed and still has the health condition.

1. However, when looking at hconds06 (heart attack) and hconds07 (stroke), there are no useful responses provided (only missing and proxy respondent and only wave 1). I am aware that they are not chronic diseases but how can I check if someone experienced a stroke or a heart attack in wave 1 if I only have hcond6/7 (if someone has ever been diagnosed)?

2. Moreover, I wanted to use hcondns and the definition says that these two conditions are left out (I assume for the same reason as above - they are not chronic). However, it appears that both variables are provided in waves 2, 3 and 4 and instead I could not find hcondns11-17 and hcondns9 and hcondns10 are only provided for wave 4. Is that correct or am I missing something?

Thank you in advance.


Updated by Cara Booker over 8 years ago

Hi Lea,

To follow up on what Gundi wrote, here is some additional information regarding these questions.

1. You will only have whether a person has been diagnosed with a heart attack/stroke at wave 1. Respondents were not asked whether they still had either of those conditions.

2. From wave 2 onward the hcondn0 variables include all possible conditions, i.e. the response options are 1-17 to coincide with the different health conditions. It would be unusual for people to have more than 8 of these at once, hence the hcondn01-08, the number was increased to 10 in wave 4. These are then converted to hcondn1-17. Some people were not asked whether they had either a heart attack or stroke in wave 2 so in wave 3 they asked everyone these questions, hconde6-7 and the age at which they were diagnosed, hcondea6-7. From wave 3 the annual event history asks about heart attack and stroke as well as whether they still have the condition.

I hope this clears things up a bit, I am happy to answer any other questions you may have.



Updated by Lea Samek over 8 years ago

Hi Cara,

thank you for your help. Just to clarify and I hope I make sense here:

1. So there is no way in telling whether someone has been diagnosed with a heart attack/stroke 10 years prior to the interview in wave 1 or only a week ago? For my work I need to know whether someone suffers from any of the 17 conditions close to the interview in wave 1, 2, 3, or 4 and I used hconds and hcondns, which worked well for all conditions except heart attack and stroke.

2. I am a bit confused about your explanation because my question refers to hcondnS but you are referring to hcondn. So does that mean that hcondn1-17 refers to/coincides with the specific newly diagnosed diseases while hcondnS1-8 only refers to a number, which has not been converted to the actual 17 conditions. I understand that it is unusual for someone to have more than 8 conditions at once but there should be hcondnS1-17 anyway to represent each condition (otherwise conditions 9) hyperthyroidism, ... 17) clinical depression are left out). If hcondnS1-8 has not been converted, how do I know which conditions 1-8 refer to?

To summarise, in order to know whether respondents suffer from any of the 15 conditions (excluding heart attack and stroke) in any of the 4 waves, would you suggest to use hcond+ hconds and hcondn+ hconds (and exclude hcondns), to check if someone has ever been diagnosed and still is diagnosed and if that person was newly diagnosed with another condition and still suffers from that? For heart attack and stroke, would I only be able to check if someone has ever been diagnosed prior to wave 1 (through hcond), has been newly diagnosed prior to wave 2 (hcondn) or between wave 2 and 3 as well as 3 and 4 (hcondn)?

Thank you very much in advance again.



Updated by Cara Booker over 8 years ago

Hi Lea,

In wave 1 you should be able to know when a person was diagnosed via the age first told they had that condition (hconda variables). These variables are in wave 1 and from wave 3 on but was not asked in wave 2.

The hcondns variables correspond to the condition identified in the hcondno variables. The hcondno variables are coded 1-17 and those numbers are the 17 different conditions. For example if a respondent answers 5 on the hcondno1 that indicates that they have been newly diagnosed with agina. They are then asked if they still have that condition in hcondns1, where they answer yes or no.

You know in wave 1 the condition and age they were first diagnosed and from wave 2 on you know whether someone has been newly diagnosed since they were last interviewed and whether they still have theat condition. From wave 3 on you know for new entrants the condition and age at which they were diagnosed. Also at wave 3 they went back and asked whether respondents who had previously been interviewed had been diagnosed with MI or stroke (c_hconde6, 7) and the age at which they were diagnosed (c_hcondea6, 7).

I think it should be possible for you to know whether respondents suffer from any of the conditions at each wave, but you may need to use some of the information provided at wave three to catch the omissions from wave 2.

I hope this is a bit clearer and apologies if it is not.


Updated by Lea Samek over 8 years ago

Thanks a lot! That is very clear now and will be very helpful.



Updated by Redmine Admin over 8 years ago

  • Status changed from New to Closed
  • % Done changed from 0 to 100

Updated by Gundi Knies over 8 years ago

  • Assignee set to Gundi Knies
  • Target version set to X M

Also available in: Atom PDF