Changed the prediabetes incidence from 9/1000 to 6.2/1000

wp3.qmd

@@ -24,7 +24,7 @@ WP3 establishes such a cohort across Steno Centers with long-term academic and s
 Academically, we describe a core protocol, which will be carried out within the DP-Next budget and timeframe, and an extended protocol, which describes our ambitions for deeper phenotyping and longer follow-up.  Connection and coordination will be sought with new Danish phenotyping initiatives currently underway or in the planning process (e.g. PRECISE).
 
 ### Who will be included:
-In the core protocol we will identify a cohort of ~1050 individuals aged 40 to 50 years who have recently (within the past 12 months) had a routine HbA1c measurement in the pre-diabetic range (42-47 mmol/mol) and are not using any glucose lowering medication. This age range was chosen to maximise life-long impact on diabetes risk. Sample size calculations use our recent analysis of HbA1c defined (pre)diabetes in Denmark [@Nicolaisen2023], showing that 25% to 40% of Danish residents aged 40-49 had at least 1 HbA1c measurement in 2018. In our target age range pre-diabetes incidence was 9/1000 person-years, and the subsequent 5-year cumulative incidence of type-2 diabetes was 31.5%. The 5-year cumulative incidence was recalculated to 23.3% for a follow-up time of 3.5 years using an exponential model assuming a constant hazard rate over the follow-up period. This is equivalent to assuming a diabetes incidence rate of 72.9/1000 person-years. Logistic regression based power calculations show that a cohort of 1050 individuals with a follow-up time of 3.5 years gives us 0.9 power at a 0.05 significance level to detect an odds ratio of at least 1.267 per standard deviation in a continuous exposure variable.
+In the core protocol we will identify a cohort of ~1050 individuals aged 40 to 50 years who have recently (within the past 12 months) had a routine HbA1c measurement in the pre-diabetic range (42-47 mmol/mol) and are not using any glucose lowering medication. This age range was chosen to maximise life-long impact on diabetes risk. Sample size calculations use our recent analysis of HbA1c defined (pre)diabetes in Denmark [@Nicolaisen2023], showing that 25% to 40% of Danish residents aged 40-49 had at least 1 HbA1c measurement in 2018. In our target age range pre-diabetes incidence was 6.2/1000 person-years, and the subsequent 5-year cumulative incidence of type-2 diabetes was 31.5%. The 5-year cumulative incidence was recalculated to 23.3% for a follow-up time of 3.5 years using an exponential model assuming a constant hazard rate over the follow-up period. This is equivalent to assuming a diabetes incidence rate of 72.9/1000 person-years. Logistic regression based power calculations show that a cohort of 1050 individuals with a follow-up time of 3.5 years gives us 0.9 power at a 0.05 significance level to detect an odds ratio of at least 1.267 per standard deviation in a continuous exposure variable.
 
 ### Recruitment:
 Participants will be recruited in Odense (350), Aarhus (350), Aalborg (150), Greenland (50) and the Faroe Islands (50). We intend to include 100 individuals of Greenlandic ancestry, 50 living in Denmark and 50 living in Greenland. Participants will be recruited based on data from routine HbA1c checks, accessed through the LABKA registers. Consequently the clinical decisions prompting a HbA1c measurement will form part of the selection process, reflecting current clinical practice.