Data extraction description
Data access and user on the Research Machine
We do not need access to a research machine, we have a project at DST.
Treatment site
We want the data to be transferred to our project on the DST research server.
Population
To identify familial relations through individuals’ life course trajectory, the population requested covers all individuals, who live or lived in Denmark from 1 January 1970 to 31 December 2021 (~8.5 million people). The family linkage and structure makes the construction of the population complicated, which we will do ourselves. For that we need the entire population because we need data on not just the individual, but also the family (parents and siblings). Denmark Statistics has given us the population and data we requested.
Data connection
Justification for the requested data
Population:
- We require data on all individuals, including those without diabetes, in order for our models to have a comparison to use and to identify risk factors. We also need the data on the individuals before they get diabetes until their birth in order to study life course effects on diabetes risk and eventual prevention.
IVF treatments and cancer diagnoses on all persons who have lived and lived in Denmark in the period 1970-2021, including persons who are not diabetic patients or in a family relationship to a diabetic patient:
Over the last decades, increasing trends in the incidence of specific types of cancers, obesity, and type 2 diabetes has been observed globally. Lifestyle factors such as the type of diet that is consumed, and the quantity of physical activity that is performed are some of the drivers in the development of obesity, which is a strong independent risk factor for type 2 diabetes and the link to certain types of cancers such as liver, pancreas, breast, among others. Furthermore, shared biological pathways between diabetes and cancer include inflammation, metabolic and genetic anomalies. In summary, a broad body of literature shows a strong link between diabetes and cancer. In this project, we would investigate how diabetes and cancer (and other diseases) cluster in families and how this disease aggregation in the family affects the individual risk of developing different chronic disease profiles (e.g. diabetes or cancer, or both). In short, data on cancer diagnoses on all persons who have lived and lived in Denmark from the period 1970-2021 is necessary, as this population will serve as the reference group when relevant.
For example, if we want to estimate the risk of diabetes for 50-year-old male individuals with a paternal history of cancer and diabetes, the reference group would be the 50-year-old male individuals with no paternal history of cancer and diabetes. It would be possible to investigate the independent effect of paternal cancer history on diabetes risk among those individuals with no paternal history of diabetes. A second example of a reference population is when estimating mortality rates of diabetes and cancer using transition models. Where individuals enter the cohort without any disease and then move into the next state that could be either diabetes only, cancer only, or death, then move to diabetes with cancer and then transition to death.
Both CAR and IVF:
- Diseases following the diabetes diagnosis includes not only the classical diabetes related diseases, but also less studied diseases like reproductive impairment and cancers. Therefore, we request access to several registers to identify relevant factors related to family and early life determinants on the development, management, and care of diabetes and for development of diabetes and diseases following diabetes.
DRG og DAGS grupperet LPR (DRG)
Hospital treatment is a crucial cost when it comes to diabetes care. It is important to analyse all costs (and not only those with an acute diagnosis of diabetes) as diabetes may be related to a general change in and therefore the consumption of hospital care.
The years are extended to 2004 in order to calculate charlson co-morbidity index with a 10-year look back.
Landspatientregisteret psykiatri (LPR)
- Treatment in a psychiatric hospital may be influenced by a possible diabetes diagnosis and should therefore be included to provide the full picture.
Plejehjemsadresser (PLH)
- People with diabetes may be at increased risk of needing to live in a care home, particularly in the context of diabetes-related conditions. We need nursing home names in order to make the most detailed calculation of the cost of living in nursing home.
Register for Udvalgte Kroniske Sygdomme (RUKS)
- RUKS is used to compare the identified diabetes population and to compare morbidity between the identified diabetes population and the diabetes-free population.
Lægemiddelstatistiskregisteret hos Danmarks Statistisk (LMDB)
- Diabetes patients can have a considerable consumption of medicines both directly related to diabetes, but also other medicines.
Laboratoriedatabasen (LAB)
- HbA1c measurements from the laboratory database researcher table are desired to examine incidence and progression of diabetes in the whole population, especially from general practice which is only available to a limited extent. Data unbiased by value are desired as diagnostic criteria have changed over time. The remaining measurements (cholesterol, eGFR, CRP, Urat, etc.) are desired to examine the incidence and progression of complications of the disease. Information on the laboratory and the applicant is requested in order to take into account possible differences in measurements and results among laboratories and geographically. Reference ranges are requested to examine responses indicated to be in the normal range and any differences between laboratories and over time.
- The relevant NPU/DNK codes are listed in the table below.