OBJECTIVE

Type 2 diabetes often coexists with other conditions that are amenable to pharmacological treatment. We hypothesized that polypharmacy among individuals with type 2 diabetes has increased since 2000.

RESEARCH DESIGN AND METHODS

Using Danish national registries, we established a cohort of all Danish individuals (aged ≥18 years) with type 2 diabetes between 2000 and 2020. We analyzed their medication use and prevalence of varying degrees of polypharmacy (≥5 or ≥10 medications), stratifying by age, sex, number of chronic diseases, and socioeconomic status.

RESULTS

The cohort grew from 84,917 patients in 2000 to 307,011 in 2020, totaling 461,849 unique patients. The number of daily medications used per patient increased from (mean ± SD) 3.7 ± 2.8 (in 2000) to 5.3 ± 3.2 (in 2020). The lifetime risk of polypharmacy was substantial, with 89% (n = 409,062 of 461,849) being exposed to ≥5 medications at some point and 47% (n = 217,467 of 461,849) to ≥10 medications. The increases were driven by an expanding group of medications, with analgesics, antihypertensives, proton pump inhibitors, and statins having the largest net increase. Advanced age, male sex, lower socioeconomic status, and Danish ethnicity positively correlated with polypharmacy but could not explain the overall increase in polypharmacy.

CONCLUSIONS

Medication use and polypharmacy have increased among patients with type 2 diabetes. Although the implications and appropriateness of this increased medication use are uncertain, the results stress the increasing need for health care personnel to understand the potential risks associated with polypharmacy, including medication interactions, adverse effects, and over- and underprescribing.

Polypharmacy, often defined as concurrent treatment with five or more medications (1), places a substantial strain on limited health care resources because it is associated with several adverse health- and medication-related complications (2). Regardless of the negative connotation, polypharmacy is often necessary when treating chronic conditions such as type 2 diabetes, because these patients require a combination of glucose-lowering, antihypertensive, and lipid-lowering medications (3,4). Still, polypharmacy seems to negatively influence both diabetes-specific (e.g., poor glycemic control, risk of hypoglycemia) and health-related (e.g., risk of falls, hospitalization, death) outcomes (5).

We hypothesized that the prevalence of polypharmacy among patients with type 2 diabetes has increased over the past decades for a few reasons. First, fueled by an increasing longevity of the general population (6), the prevalence of multimorbidity amenable to pharmacologic treatment is growing. Second, the past decades have seen an expansion in the available medications for type 2 diabetes and many other diseases (7). These factors likely overshadow the effects of changed diagnostic thresholds and increased preventive and screening efforts that might have decreased the prevalence of polypharmacy by including more early-stage type 2 diabetes cases, which require less treatment (8).

In this study, we describe the trajectories of medication use and polypharmacy from 2000 to 2020 among the Danish population with type 2 diabetes. Focusing on polypharmacy, we examined its correlation with demographic shifts, including changes in age, sex, number of diseases, country of birth, and socioeconomic status. Our goal is to improve the understanding of polypharmacy characteristics among patients exposed to and clinicians and other caregivers dealing with type 2 diabetes to ultimately mitigate the harms associated with polypharmacy.

Study Design and Setting

In this noninterventional, observational cohort study, we identified all adult patients (aged ≥18 years) with type 2 diabetes in Denmark between 2000 and 2020 by crosslinking multiple national registries. The patients entered the cohort at the earliest of three possible dates: their date of type 2 diabetes diagnosis, their 18th birthday, or the study start date. The patients were censored at death or loss to follow-up.

Data Sources

We crosslinked information from the National Patient Registry (9), Danish National Prescription Registry (hereafter, Prescription Registry) (10), Danish National Health Service Registry (11), the Danish National Personal Income Registry (12), the Danish National Education Registry (13), the Civil Registration System (14), and the Clinical Laboratory Information Register (15), using unique identification numbers given to all Danish inhabitants.

Participants

We included all patients with type 2 diabetes aged ≥18 years, using a previously validated method (16). First, we identified all patients with either type 1 or 2 diabetes by identifying the following events: 1) redemption of a glucose-lowering medication (Anatomical Therapeutic Chemical [ATC] group A10 in the Prescription Registry); 2) diagnosis with diabetes in hospital (ICD-10 code E10 or E11 in the National Patient Registry); 3) HbA1c measurements ≥6.5% (≥48 mmol/mol) (NPU27300 in the Clinical Laboratory Information Register); and 4) diabetes-specific services by podiatrist or general practitioner (code 54 or 80 plus subcode 0131 or 0132 in the Danish Health Service Registry). To avoid misclassification, we used the following exclusion criteria: 1) all events in 1 year before and after pregnancy or diagnosis of gestational diabetes, due to the risk of misclassification as gestational diabetes; 2) patients who only had one of the mentioned events; 3) patients who were female, aged <40 years, and only redeemed metformin (this indicates polycystic ovary syndrome); and 4) patients who were only treated with glucagon-like-peptide 1 receptor agonists approved only for body weight reduction and not reimbursed in Denmark (liraglutide tradename Saxenda; Novo Nordisk, Plainsboro, NJ).

We then excluded patients with type 1 diabetes, defined as patients who either only used insulin and had a diagnosis of type 1 diabetes, or had a majority of type 1 diagnoses while being prescribed insulin within 180 days from onset of diabetes, and the prescribed glucose-lowering medications primarily constitute of insulin. The remaining patients were classified as having type 2 diabetes. A flowchart of the algorithm is provided in Supplementary Fig. 1.

Medication Exposure

The Prescription Registry contains all redeemed (filled) prescriptions since 1996 (10). We excluded prescriptions without available quantity data (n = 592; 0.02‰), as well as those intended for topical and vaginal use, chewing gums, ear and eye drops, flushing fluids, mouthwashes, nasal sprays, and medicinal toothpaste. When tallying medications used per patient, we grouped them by their ATC classification at the fifth level (i.e., defining the exact chemical substance) (17). Fixed-dose combinations, such as metformin and dapagliflozin (ATC code A10BD15), were counted as a single medication. We calculated the daily number of medications for each patient by determining the coverage period for each medication group. To calculate the coverage period, we divided the redeemed amount for each prescription by the patient’s daily dosage and added this figure to the prescription’s redemption date. Because the Prescription Registry only contains the quantity of medication redeemed per prescription, without details on the exact dosage regimen, we adopted a previously established method (18,19) for estimating daily dosages. This method estimates the daily dosage on the basis of the cumulative quantity redeemed from the previous six or fewer prescriptions, divided by the number of days since the redemption of the first of these prescriptions, thus allowing for the estimation of patient-specific and time-varying dosage regimens.

For new treatments, without prior prescription history, we defaulted to the World Health Organization’s defined daily dose (20). A treatment was also considered new if the previous coverage period had expired and the last redemption was more than 6 months ago. However, a 30-day stock was allowed (21), meaning patients could be uncovered for 30 days between the calculated coverage periods and still be considered in treatment, despite the last redemption being more than 6 months ago. Furthermore, days spent in hospital were considered grace periods, meaning they were not counted against the coverage period, because medications are provided during hospital admissions in Denmark.

Polypharmacy

We defined two degrees of polypharmacy: concurrent treatment with ≥5 or ≥10 medications. Daily, annual, and lifetime prevalences of polypharmacy were calculated. Annual prevalence was defined as individuals exposed to polypharmacy at some point during that year. Similarly, lifetime prevalence was defined as exposure during the study period (years 2000–2020). Daily prevalences are presented graphically using locally estimated scatterplot smoothing (22). Unsmoothed graphics are available in Supplementary Fig. 2.

Medication Categories Used by Patients Exposed to Polypharmacy

When analyzing medication types used, we grouped the medications into clinically meaningful categories and subcategories. We defined these categories guided by the ATC system to aid the interpretability and clinical relevance of the results. The full description of the categories is available in Supplementary Fig. 1.

Covariates

We gathered information about sex, age, ethnicity, gross household income, level of education, and number of chronic diseases. Age was categorized into the following groups: 18–64, 65–74, or ≥75 years. Chronic diseases were identified using ICD-10 codes. The number of chronic diseases includes the diagnosis of diabetes and was categorized into one to two diseases, three to six, or more than six. A detailed list of the definitions and algorithm for identifying chronic diseases are available in Supplementary Table 2. We used the country of birth as a surrogate for ethnicity. We categorized this into Denmark, Western, and non-Western countries, according to Statistics Denmark’s official classifications (23). Gross household income was retrieved per year and categorized into quartiles based on data from the full Danish population provided by Statistics Denmark. The level of education was categorized according to the International Standard Classification of Education.

Statistical Analysis

Continuous covariates are presented as means (SD) for normally distributed data, and median (interquartile range) for skewed data. Categorical covariates are presented as numbers and frequencies with 95% CIs. Differences in patient characteristics were analyzed using Kruskal-Wallis rank sum test or χ2 test. The mean number of medications used per year was calculated using Poisson regression with the logarithm of patient-days as offset, thus allowing for adjusting for age, country of birth, education, gross household income, and sex. The correlation between the 15 most used medication categories from 2020 was examined. Missing data were categorized as unknown, and patients with missing data were excluded from the stratified analyses but included in regression analyses. All data management, statistical analyses, and figures were performed in R, version 4.1.3.

Ethics

Ethical approval is not required for retrospective register-based studies in Denmark. This study has been registered with the Capital Region of Denmark’s Knowledge Center for Data Reviews (approval no. P-2021-220).

Data and Resource Availability

The data supporting this study’s findings are available from Statistics Denmark. Restrictions apply to the availability of these data and are therefore not publicly available.

Cohort Characteristics

The cohort included 461,849 patients, collectively contributing with 4,234,333 person-years (1.5 billion patient-days). An inclusion flowchart is available in Supplementary Fig. 1. The cohort steadily increased from 84,917 patients alive in 2000 to 307,011 in 2020. Patient characteristics stratified by year are available in Supplementary Table 3. In short, patients in 2020 were older, more often male, less frequently born in Denmark, and had a higher level of education than patients in 2000. Moreover, multimorbidity increased linearly since 2000. In 2000, 52% of the population had two or more chronic diseases (including type 2 diabetes), which increased to 76% in 2020. The mean number of chronic diseases for each year is available in Supplementary Fig. 3.

Number of Medications Used

Figure 1 shows the daily prevalence of polypharmacy, both as the absolute number of patients and the proportion of patients alive each day. In 2000, 53% (n = 44,798 of 84,917) of the population had been exposed to one of the two defined degrees of polypharmacy. In 2020, this had increased to 76% (n = 231,970 of 307,011) in 2020. This is illustrated in Fig. 2, which shows the percentage of the population exposed to a specific number of medications and polypharmacy at some point that year. When followed throughout the study period, 89% (n = 409,062 of 461,849) had used ≥5 medications at some point, and 47% (n = 217,467 of 461,849) had used ≥10 medications. Figure 2 also shows that there has been a decline in the usage of ≥14 medications since 2010.

Figure 1

Daily population size and prevalence of polypharmacy, defined as concurrent treatment with ≥5 and ≥10 medications. Absolute number of patients (left). Percentage of patients (right).

Figure 1

Daily population size and prevalence of polypharmacy, defined as concurrent treatment with ≥5 and ≥10 medications. Absolute number of patients (left). Percentage of patients (right).

Close modal
Figure 2

Percentage of the population using a specific number of medications (bars on the x-axis) and exposed to various degrees of polypharmacy (colors) at least once that year. (The data consist of daily use aggregated per year. The area under the curve can exceed 100% because one individual can contribute to multiple bars on the x-axis for each year).

Figure 2

Percentage of the population using a specific number of medications (bars on the x-axis) and exposed to various degrees of polypharmacy (colors) at least once that year. (The data consist of daily use aggregated per year. The area under the curve can exceed 100% because one individual can contribute to multiple bars on the x-axis for each year).

Close modal

Contrary to the linear increase in chronic diseases from 2000 to 2020, the number of medications used primarily increased from 2000 to 2010. From 2000 to 2020, the mean increased from 3.7 (SD 2.8) to 5.3 (SD 3.2), corresponding to an unadjusted increase of 43 (95% CI 43–43%). A visual presentation of the mean number of medications used is available in Supplementary Fig. 3. Adjusting for age, country of birth, education, gross household income, and sex did not change the estimate vastly (46%; 95 CI 46–46%). The mean number of glucose-lowering medications used increased from 0.9 (SD 0.7) to 1.1 (SD 1.0), and the mean number of antihypertensives increased from 0.4 (SD 0.6) to 0.7 (SD 0.6).

Types of Medications Used

Figure 3 shows which medications the patients were using in 2000 and 2020. Despite distribution changes in subgroups, glucose-lowering medications remained the most used medications throughout the study period. The use of preventive cardiovascular medications (i.e., antihypertensives, antithrombotic medications, and statins) and analgesics increased, and the use of anxiolytics decreased. Correlations between the use of different medication classes are available in Supplementary Fig. 4. For example, there was a correlation between the use of laxatives and opioids.

Figure 3

The most frequently used medications in 2000 and 2020. Categorized in main and subcategories and arranged by most used medications in 2020. Both main and subcategories show the percentage of the full population. Because only some subcategories are shown, the sum of the percentages in the subcategories will not always equal the main categories’ percentages. Categories used by <5% of all patients, as well as antibiotics, antimycotics, vitamins, and minerals, were excluded. The categories’ definitions are available in Supplementary Fig. 1. DPP-4, dipeptidyl peptidase 4; GLP1–RA, glucagon-like peptide 1 receptor agonists; NSAID, nonsteroidal anti-inflammatory drug; SGLT-2, sodium–glucose transport protein 2; SSRI, selective serotonin reuptake inhibitor; TCA, tricyclic antidepressant.

Figure 3

The most frequently used medications in 2000 and 2020. Categorized in main and subcategories and arranged by most used medications in 2020. Both main and subcategories show the percentage of the full population. Because only some subcategories are shown, the sum of the percentages in the subcategories will not always equal the main categories’ percentages. Categories used by <5% of all patients, as well as antibiotics, antimycotics, vitamins, and minerals, were excluded. The categories’ definitions are available in Supplementary Fig. 1. DPP-4, dipeptidyl peptidase 4; GLP1–RA, glucagon-like peptide 1 receptor agonists; NSAID, nonsteroidal anti-inflammatory drug; SGLT-2, sodium–glucose transport protein 2; SSRI, selective serotonin reuptake inhibitor; TCA, tricyclic antidepressant.

Close modal

Stratified Analyses

Figure 4 shows the prevalence of polypharmacy stratified by different patient characteristics. The prevalence has increased in all strata, and parallel lines indicate there has been a similar increase in the subgroups of each stratum. The prevalence of polypharmacy was higher in patients with advanced age, multiple chronic diseases, male sex, Danish ethnicity, low level of education, and low gross household income.

Figure 4

Stratified daily prevalence of polypharmacy, defined as concurrent use of ≥5 (solid) and ≥10 (dashed) medications.

Figure 4

Stratified daily prevalence of polypharmacy, defined as concurrent use of ≥5 (solid) and ≥10 (dashed) medications.

Close modal

This study shows that since 2000, the number of Danish patients with type 2 diabetes has increased alongside the number of medications used by each patient. Consequently, the number of patients exposed to polypharmacy has increased by factors of 7 and 11 for concurrent use of at least 5 and 10 medications, respectively. Furthermore, 9 of every 10 patients had been exposed to polypharmacy (at least five medications) at some point in the study period. Neither the changes in population size nor demographics explain the observed increase in polypharmacy. Although the number of chronic diseases has increased, the stale prescribing rate seen in the past 10 years has resulted in a decreasing ratio between the average number of medications used and the average number of chronic diseases. Still, the prevalence of polypharmacy also increased in patients with two or fewer chronic diseases. In fact, the relative increase in prevalence of polypharmacy (both >5 and >10 medications) is larger in this patient group compared with the patient group with five or more chronic diseases (Supplementary Fig. 5).

Use of medications such as analgesics, antihypertensives, proton pump inhibitors, and statins had the largest net increase, whereas use of anxiolytics has been vastly reduced. As expected, the most used medications in 2020 were glucose-lowering and preventive cardiovascular medications, such as statins and antihypertensives. However, polypharmacy is becoming the norm in patients with type 2 diabetes, and most patients are exposed to analgesics, diuretics, antithrombotics, and other cardiovascular medications.

This increasing trend toward polypharmacy not only poses a potential strain on the health care system but also underscores the need for future health care workers to understand the complexities of polypharmacy, including knowledge about interactions between medications as well as certain medication–disease combinations.

Strength and Limitations

The main limitation of these data resides in evaluating medication use from registry data, because medication adherence and exact dosages are unknown. However, the Danish Prescription Registry is of high quality (24) and based on redeemed and not issued prescriptions, which increases the likelihood of actual medication consumption (25). We estimated the dosages using each patient’s previously redeemed amounts, thereby increasing the precision for chronically used medications. However, the precision might be lower when estimating intermittent (as-needed) medications, occasionally leading to both overestimations and underestimations of actual use. Although we stratified and adjusted for multiple confounders, there remains a risk for residual confounding.

Prevalence of Polypharmacy and Inappropriate Pharmacotherapy

Previous studies have shown that the pooled prevalence of polypharmacy in patients with type 2 diabetes is ∼65% but varies widely depending on the definition of polypharmacy, study population, and study year (5). In this study, with a full national population of patients with type 2 diabetes, we showed that 57% of the population with type 2 diabetes in 2020 fulfilled the most commonly used definition of polypharmacy (concurrent treatment with at least five medications) (1) on any given day of the year. Interestingly, 76% met this definition at least once in 2020, and 89% met it at some point in the study period. These results highlight that the prevalence of polypharmacy depends on the definition and methods used to evaluate it.

Increasing degrees of polypharmacy are strongly associated with inappropriate pharmacotherapy (2). Therefore, using a numeric threshold to define polypharmacy can be a convenient method for identifying patients who might benefit from medication optimization interventions, such as medication reviews. It is noteworthy that previous studies suggest the risk of inappropriate medications at a certain degree of polypharmacy might be lower in patients with type 2 diabetes than in the general population (26,27). This notion is important to consider when interpreting the high prevalence of polypharmacy and attempting to use polypharmacy as a proxy for inappropriate medication use by patients with type 2 diabetes. For that reason, we also included a higher threshold for polypharmacy (concurrent treatment with ≥10 medications), sometimes referred to as severe or excessive polypharmacy (1).

Medication Categories

The availability of new medications, particularly for chronic conditions, has significantly contributed to the increase in polypharmacy by introducing more possibilities for combination therapy. For instance, in glucose-lowering medications, a shift from sulphonylureas to a wider array of medications and a net increase in medication use is evident. Likewise, although the use of acetylsalicylic acid has declined, likely due to a combination of evidence and updated recommendations against primary prophylaxis (28), the use of clopidogrel and direct factor Xa inhibitors has increased, resulting in a net increase in the use of antithrombotic medications. Collectively, preventive cardiovascular medications (i.e., antihypertensives, HMG-CoA reductase inhibitors [statins], and antithrombotic medications) used according to evidence-based guidelines (29–31) seem to be the main contributors to the increased prevalence of polypharmacy.

In recent decades, several medications have been discouraged due to various adverse events, leading to notable shifts in prescribing patterns. For instance, anxiolytics (32) and nonsteroidal anti-inflammatory drugs (33) use have been vastly reduced. Interestingly, despite the reduction in nonsteroidal anti-inflammatory drug use, there has been an overall increase in analgesic use. Notably, paracetamol use has increased dramatically: 50% of the total population with type 2 diabetes and 75% of those exposed to ≥10 medications used it in 2020. Additionally, the use of gabapentinoids (gabapentin and pregabalin) has increased 10-fold. This likely reflects an increasing use for diabetes-associated pain conditions, but it may also pose a risk for adverse effects from the central nervous and cardiovascular system (34).

The increased awareness of adverse events associated with certain medications might also, at least in part, explain the increased use of laxatives. These are often used to mitigate gastrointestinal side effects commonly induced by opioids (35) and could lead to electrolyte disturbances and affect cognitive function that are also treatable (36). This trend showcases a prescribing cascade (37), a phenomenon where the side effects of one medication necessitate the use of additional medications.

Stratified Analyses

Our stratified analysis reveals a relatively uniform increase in polypharmacy across most, but not all, demographic subgroups. Naturally, the number of chronic diseases seems to partially drive the prevalence of polypharmacy. Still, this increase is also evident in patients with fewer chronic diseases, indicating that factors beyond disease burden—like the increased selection of available medications and changes in clinical guidelines—are contributing to the increase in polypharmacy. Although certain characteristics were associated with an increased prevalence of polypharmacy, they could not explain the increase seen across the population. Patients with lower education levels, low household incomes, and those born in Denmark were more likely to be exposed to polypharmacy compared with their counterparts. These disparities have remained relatively constant over time, except for a widened span between gross household income quartiles. Further studies are needed to explore these observations.

Conclusion

Since 2000, the number of Danish patients with type 2 diabetes has increased fourfold, and their medication use has almost doubled. Consequently, polypharmacy is becoming the norm, with almost 9 of every 10 patients having used five or more medications concurrently at some point. Additionally, more than half of the patients have used 10 or more medications concurrently at some point. This increase in polypharmacy could not be explained by changes in population size or demographics.

The overall implication and appropriateness of this increased medication use are uncertain and likely patient specific. However, the findings stress the increasing need for health care personnel to understand the complexities of multimorbidity and risks associated with polypharmacy, which include increased risk of medication interactions, adverse effects, and over- and underprescribing.

This article contains supplementary material online at https://doi.org/10.2337/figshare.25612791.

See accompanying article, p. 2104.

This article is featured in a podcast available at diabetesjournals.org/journals/pages/diabetes-core-update-podcasts.

Funding. The study is funded by the Capital Region of Denmark.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. K.S.J. and M.B.C. conceived of the study. K.S.J. managed the data, performed the statistical analyses and wrote the first draft of the manuscript. All authors contributed to the study design, had access to the data, were involved in interpreting the results, reviewed the manuscript, and approved the final version. K.S.J. and M.B.C. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

1.
Masnoon
N
,
Shakib
S
,
Kalisch-Ellett
L
,
Caughey
GE.
What is polypharmacy? A systematic review of definitions
.
BMC Geriatr
2017
;
17
:
230
2.
Davies
LE
,
Spiers
G
,
Kingston
A
,
Todd
A
,
Adamson
J
,
Hanratty
B.
Adverse outcomes of polypharmacy in older people: systematic review of reviews
.
J Am Med Dir Assoc
2020
;
21
:
181
187
3.
ElSayed
NA
,
Aleppo
G
,
Aroda
VR
, et al.;
American Diabetes Association
.
9. Pharmacologic approaches to glycemic treatment: Standards of Care in Diabetes—2023
.
Diabetes Care
2023
;
46
(
Suppl. 1
):
S140
S157
4.
ElSayed
NA
,
Aleppo
G
,
Aroda
VR
, et al.;
American Diabetes Association
.
10. Cardiovascular disease and risk management: Standards of Care in Diabetes—2023
.
Diabetes Care
2023
;
46
(
Suppl. 1
):
S158
S190
5.
Remelli
F
,
Ceresini
MG
,
Trevisan
C
,
Noale
M
,
Volpato
S.
Prevalence and impact of polypharmacy in older patients with type 2 diabetes
.
Aging Clin Exp Res
2022
;
34
:
1969
1983
6.
Carstensen
B
,
Rønn
PF
,
Jørgensen
ME.
Prevalence, incidence and mortality of type 1 and type 2 diabetes in Denmark 1996-2016
.
BMJ Open Diabetes Res Care
2020
;
8
:
e001071
7.
Johansson
KS
,
Sonne
DP
,
Knop
FK
,
Christensen
MB.
What is on the horizon for type 2 diabetes pharmacotherapy? An overview of the antidiabetic drug development pipeline
.
Expert Opin Drug Discov
2020
;
15
:
1253
1265
8.
Knudsen
JS
,
Knudsen
SS
,
Hulman
A
, et al
.
Changes in type 2 diabetes incidence and mortality associated with introduction of HbA1c as diagnostic option: a Danish 24-year population-based study
.
Lancet Reg Health Eur
2022
;
14
:
100291
9.
Schmidt
M
,
Schmidt
SAJ
,
Sandegaard
JL
,
Ehrenstein
V
,
Pedersen
L
,
Sørensen
HT.
The Danish National Patient Registry: a review of content, data quality, and research potential
.
Clin Epidemiol
2015
;
7
:
449
490
10.
Kildemoes
HW
,
Sørensen
HT
,
Hallas
J.
The Danish National Prescription Registry
.
Scand J Public Health
2011
;
39
(
Suppl.
):
38
41
11.
Andersen
JS
,
Olivarius
NdeF
,
Krasnik
A.
The Danish National Health Service Register
.
Scand J Public Health
2011
;
39
(
Suppl.
):
34
37
12.
Baadsgaard
M
,
Quitzau
J.
Danish registers on personal income and transfer payments
.
Scand J Public Health
2011
;
39
(
Suppl.
):
103
105
13.
Jensen
VM
,
Rasmussen
AW.
Danish education registers
.
Scand J Public Health
2011
;
39
(
Suppl.
):
91
94
14.
Schmidt
M
,
Pedersen
L
,
Sørensen
HT.
The Danish Civil Registration System as a tool in epidemiology
.
Eur J Epidemiol
2014
;
29
:
541
549
15.
Arendt
JFH
,
Hansen
AT
,
Ladefoged
SA
,
Sørensen
HT
,
Pedersen
L
,
Adelborg
K.
Existing data sources in clinical epidemiology: laboratory information system databases in Denmark
.
Clin Epidemiol
2020
;
12
:
469
475
16.
Isaksen
AA
,
Sandbæk
A
,
Bjerg
L.
Validation of register-based diabetes classifiers in Danish data
.
Clin Epidemiol
2023
;
15
:
569
581
17.
WHO Collaborating Centre for Drug Statistics Methodology
. Guidelines for ATC classification and DDD assignment.
2022
. Accessed 18 January 2024. Available from https://www.whocc.no/atc_ddd_index_and_guidelines/guidelines/
18.
Schjerning Olsen
A-M
,
Gislason
GH
,
McGettigan
P
, et al
.
Association of NSAID use with risk of bleeding and cardiovascular events in patients receiving antithrombotic therapy after myocardial infarction
.
JAMA
2015
;
313
:
805
814
19.
Gislason
GH
,
Rasmussen
JN
,
Abildstrøm
SZ
, et al
.
Long-term compliance with beta-blockers, angiotensin-converting enzyme inhibitors, and statins after acute myocardial infarction
.
Eur Heart J
2006
;
27
:
1153
1158
20.
WHO Collaborating Centre for Drug Statistics Methodology
. ATC classification index with DDDs.
2022
. Accessed 18 January 2024. Available from https://atcddd.fhi.no/atc_ddd_index_and_guidelines/guidelines/
21.
Schneeweiss
S
,
Avorn
J.
A review of uses of health care utilization databases for epidemiologic research on therapeutics
.
J Clin Epidemiol
2005
;
58
:
323
337
22.
Cleveland
WS
,
Devlin
SJ.
Locally weighted regression: an approach to regression analysis by local fitting
.
J Am Stat Assoc
1988
;
83
:
596
610
23.
Statistics Denmark
. Documentation of statistics: the population. Accessed 21 October 2022. Available from https://www.dst.dk/en/Statistik/dokumentation/documentationofstatistics/the-population
24.
Pottegård
A
,
Schmidt
SAJ
,
Wallach-Kildemoes
H
,
Sørensen
HT
,
Hallas
J
,
Schmidt
M.
Data resource profile: the Danish National Prescription Registry
.
Int J Epidemiol
2017
;
46
:
798
798f
25.
Pottegård
A
,
Christensen
Rd
,
Houji
A
, et al
.
Primary non-adherence in general practice: a Danish register study
.
Eur J Clin Pharmacol
2014
;
70
:
757
763
26
Kornholt
J
,
Feizi
ST
,
Hansen
AS
, et al
.
Medication changes implemented during medication reviews and factors related to deprescribing: post-hoc analyses of a randomized clinical trial in geriatric outpatients with polypharmacy
.
Br J Clin Pharmacol
2023
;
89
:
3291
3301
27.
Johansson
KS
,
Kornholt
J
,
Bülow
C
, et al
.
Physician-led medication reviews in polypharmacy patients treated with at least 12 medications in a type 2 diabetes outpatient clinic: a randomised trial
.
Diabet Med
2023
;
40
:
e15052
28.
Christensen
MB
,
Jimenez-Solem
E
,
Ernst
MT
,
Schmidt
M
,
Pottegård
A
,
Grove
EL.
Low-dose aspirin for primary and secondary prevention of cardiovascular events in Denmark 1998-2018
.
Sci Rep
2021
;
11
:
13603
29.
Kearney
PM
,
Blackwell
L
,
Collins
R
, et al.;
Cholesterol Treatment Trialists’ (CTT) Collaborators
.
Efficacy of cholesterol-lowering therapy in 18,686 people with diabetes in 14 randomised trials of statins: a meta-analysis
.
Lancet
2008
;
371
:
117
125
30.
Navarese
EP
,
Robinson
JG
,
Kowalewski
M
, et al
.
Association between baseline LDL-C level and total and cardiovascular mortality after LDL-C lowering: a systematic review and meta-analysis
.
JAMA
2018
;
319
:
1566
1579
31.
Brunström
M
,
Carlberg
B.
Effect of antihypertensive treatment at different blood pressure levels in patients with diabetes mellitus: systematic review and meta-analyses
.
BMJ
2016
;
352
:
i717
32.
Weich
S
,
Pearce
HL
,
Croft
P
, et al
.
Effect of anxiolytic and hypnotic drug prescriptions on mortality hazards: retrospective cohort study
.
BMJ
2014
;
348
:
g1996
33.
Straube
S
,
Tramèr
MR
,
Moore
RA
,
Derry
S
,
McQuay
HJ.
Mortality with upper gastrointestinal bleeding and perforation: effects of time and NSAID use
.
BMC Gastroenterol
2009
;
9
:
41
34.
Pan
Y
,
Davis
PB
,
Kaebler
DC
,
Blankfield
RP
,
Xu
R.
Cardiovascular risk of gabapentin and pregabalin in patients with diabetic neuropathy
.
Cardiovasc Diabetol
2022
;
21
:
170
35.
Mo
C
,
Sun
G
,
Lu
M-L
, et al
.
Proton pump inhibitors in prevention of low-dose aspirin-associated upper gastrointestinal injuries
.
World J Gastroenterol
2015
;
21
:
5382
5392
36.
Feng
J
,
Zheng
N
,
Fan
X
, et al
.
Association of laxatives use with incident dementia and modifying effect of genetic susceptibility: a population-based cohort study with propensity score matching
.
BMC Geriatr
2023
;
23
:
122
37.
Rochon
PA
,
Gurwitz
JH.
Optimising drug treatment for elderly people: the prescribing cascade
.
BMJ
1997
;
315
:
1096
1099
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/journals/pages/license.