Prepared by: Ben Schöttker, German Cancer Research Center, Clinical Epidemiology and Ageing research, Heidelberg, Germany
Contents:
- Summary of CRP assessment in cohorts
- Harmonized CRP variable
- Availability of eligible data in cohorts for the harmonized CRP variable
- Additional comments
- Missing information
- References
1. Summary of CRP assessment in cohorts
Details of collected information about CRP measurement from the CHANCES cohorts are displayed in the attached Excel table.
Ten cohorts can provide data (ESTHER, HAPIEE CZ, MORGAM-FINRISK, MORGAM-Belfast, MORGAM-Scotland, DD, NHS, RES, SENECA and Tromsö). Repeated measurements are available for ESTHER, RES and Tromsø.
Cohorts show the following common ground:
- CRP measurement took place from samples stored at at least -70°C or have been measured immediately (< 48h) from taken blood samples.
- Turbidimetric immunoassays were applied.
- Subjects with CRP measurements reflect random samples of general populations. Exceptions are DD (only subjects with dementia) and the NHS (CRP for subjects with heart disease and diabetes plus controls).
Cohorts vary in the following:
- Measurements from serum (6 cohorts), plasma (2 cohorts) and full blood (1 cohort) [not known in 1 cohort].
- Apparatus used for measurements and calibrators applied (in most cohorts the calibrator of the assay is not known).
2. Harmonized CRP variable
The following CRP variable has been defined in document Definitions of variables:
3. Availability of eligible data in cohorts for the harmonized CRP variable
- Codable for all ten cohorts.
Table 7 – Number of available CRP measurements in CHANCES cohorts for general population samples
| Cohort | Contact with data | n | Re-contact with data | n |
|---|---|---|---|---|
| ESTHER | BL | 9,690 | 5-year FUP 8-year FUP |
~ 5,000 ~ 5,000 |
| HAPIEE CZ | BL | 6,658 | ||
| RES | BL | total cohort | FU2 | ? |
| SENECA | BL | 831 | ||
| Tromsø | 4 | 6,739 | 5 6 |
5,856 7,211 |
| MORGAM Scotland | BL | 15,000 | ||
| MORGAM FINRISK | BL | 8,000 | ||
| MORGAM Belfast | BL | 2,500 | ||
| Total | ~ 49,418 + RES | ~ 10,856* + RES |
Note: DD was not listed as it does not reflect a random sample of a general population. NHS is not listed because CRP was only measured for nested case control studies for subjects with heart disease and diabetes.
* Only the re-contact of each cohort was counted that was closest to the baseline measurement with respect to follow up duration.
4. Additional comments
- Although the type of assay is always the same, the calibrations of the assays vary as well as the pre-analytics. The influence of these differences is uncertain. Parallel analyses in individual cohorts are recommended but pooled analyses might be possible as well.
- The DD is not a general population sample and can not be pooled with other cohorts.
- The NHS can not be used for general population analyses because vitamin D was only measured for nested case control studies for subjects with hypertension, colon cancer, breast cancer and colorectal adenoma. The control-subjects might be usable. However, they are not random samples of the total study population.
5. Missing information
- Apparatus used and assay type in ESTHER FUPs, RES FUP2 and Tromsø 4 and 6?
- Sample storage conditions in DD and Tromso 5 and 6?
- Serum, plasma or whole blood used in DD?
- CRP assessed in total population in RES at FUP2?
