MPP Diabetes endpoint report 2017

Malmö Preventive Project 
Department of Clinical Sciences 
Skåne University Hospital, Malmö 
Lund University

Report by
Anders Dahlin 
Peter Nilsson

Malmö Preventive Project - Diabetes Endpoints

End of follow-up: 31 December 2016 Report: 12 April 2018

General information

An individual with diabetes can be characterized as either prevalent or incident with respect to diabetes onset in relation to the time of baseline study. Hence, individuals with prevalent diabetes can not be incident. Only the first recorded diabetes event of an individual has been used in this report. 

Prevalent: date of diagnosis ≤ date of entry 
Incident: date of diagnosis > date of entry 

In total, 15 different sources of data were used to identify individuals with diabetes (table 5). All individuals in the Diabetes 2000 (except certain with missing data), the ANDIS (Alla Nya Diabetiker I Skåne) and the NDR (Swedish National Diabetes Register) registries were considered as diabetics, as well as individuals with at least two HbA1c-values >=6,0% (not on the same day) in the HbA1c register at Clinical Chemistry, Malmö. In the National Patient Register (inpatient care and outpatient care) and the Swedish Cause of Death Register from the National Board of Health and Welfare individuals with the ICD10 codes E10-E14 and O244-O249 (and corresponding ICD7-9 codes), were treated as diabetics. The ATC code A10 was used to identify diabetics in the Prescribed Drug Register. From the MPP and MDC cohorts, seven different study generated data sources from baseline, rescreening and subcohort studies have been used to identify diabetics. The MPP endpoint update takes into account MDC data, and vice versa. Briefly, individuals with fasting blood glucose >=6.5, or glucose >=11 mmol/L at 120 minutes OGTT, or fasting plasma glucose >=7 mmol/L, or have reported intake of A10 drugs, or have responded “Yes” on the question “Do you have diabetes?” (or similar) have been considered as diabetics. Elevated glucose levels in the MDC CV rescreening study must have been verified through fasting plasma glucose and/or OGTT tests in a subsequent examination in order to confirm new diabetes. 

Type of diabetes are reported in the Diabetes 2000 and NDR registries, but the type has also been possible to assign through the ICD codes in the National Patient Register and the Swedish Cause of Death Register. Diabetics who have taken oral anti-diabetes drugs only have been classified as Type 2 diabetics. An individual can have information about diabetes from more than one source but only the information associated with the first recorded event has been used regardless of completeness or reliability. Contradictory information about diabetes status or type in subsequent events has been neglected. Some of the 15 diabetes sources contain multiple information (e g both questionnaire and lab data) on diabetes. In order to choose one of these within source the following selection order has been applied: 1) questionnaire, 2) drugs, 3) fasting plasma glucose, and 4) glucose at 120 minutes OGTT. Year at diabetes onset was only available in the Diabetes 2000 and the NDR registries. Information on diabetes onset year was taken into account, together with data from 15 sources in determining the timing of prevalent or incident diabetes.

Amendments and implementation of new procedures

* The year when the endpoint update was made, usually the year after last follow-up year.

Cohort

Total: 33 346 
Men: 22 444 
Women: 10 902

Time of follow-up 

Follow-up time from date of entry to death, lost to follow-up or to 31 December 2016.

Individuals with prevalent or incident diabetes by sex

Number of individuals with diabetes until 31 Dec 2016.

Individuals with incident diabetes by year of diabetes onset or first recorded event 

Number of individuals with incident diabetes until 31 December 2016 by year of diabetes onset or first recorded event.

Individuals with prevalent or incident diabetes by source and type of diabetes 

Number of individuals with prevalent or incident diabetes (recorded in the first event) until 31 December 2016 by source of data and type of diabetes.

1) The Diabetes 2000 Registry
2) The Swedish National Diabetes Register (NDR) 
3) The HbA1c register at Clinical Chemistry, SUS, Malmö 
4) The Hospital Discharge Register (also The National Inpatient Register, IPR) (Slutenvårdsregistret) 
5) The National Patient Register – Outpatient Care (Spec Öppenvårdsregistret)
6) The Cause-of-death Register (Dödsorsaksregistret)
7) The Swedish Prescribed Drug Register (Läkemedelsregistret)
8) MPP baseline screening (1974-92) (type of source: questionnaire, fB-glucose, OGTT)
9) MPP 6-year rescreening (1981-89) (questionnaire, fB-glucose, OGTT)
10) MPP rescreening (2002-06) (questionnaire, list of drugs, fP-glucose)
11) MDC baseline screening (1991-96) (questionnaire, list of drugs)
12) MDC baseline screening cardiovascular cohort (1992-94) (fB-glucose)
13) MDC 5-year rescreening (1997-2001) (questionnaire, list of drugs)
14) MDC cardiovascular rescreening (2007-12) (questionnaire, list of drugs, fP-glucose, OGTT)
15) The ANDIS (Alla Nya Diabetiker I Skåne) Registry