The Di@bet.es study is a cross-sectional population survey carried out in Spain from 2009 to 2010 and including adult patients older than 18 years. The sample size calculation was based on a 15% prevalence of DM and a sampling error of 1%. One hundred clusters were randomly selected from the total of health centers or their equivalents in all parts of Spain (the design effect was 1.5, as it was considered that the clusters would be heterogeneous with respect to the study variables). The final sample size allowed for a participation rate of 55% (data from similar studies) and included an additional 30% to compensate for losses in a future study on incidence. Of the 10 227 adult candidate participants, 5728 (56%) came to the study visit. Of these, 9.9% were excluded by protocol (institutionalization, severe disease, pregnancy, or recent delivery). Thus, 5544 patients were ultimately included in the study.2

The study was approved by the Ethics and Clinical Research Committee of Hospital Carlos Haya, and by other pertinent regional ethics and clinical research committees of Spain. All participants signed an informed consent before beginning the study.

Participants were invited by letter or phone contact to attend a study visit in their respective health centers. Information was collected by an interviewer using a structured questionnaire with items related to sociodemographic data, clinical variables (current diagnoses and history of hypertension, cardiovascular disease, stroke, and peripheral artery disease), and lifestyle (smoking, alcohol consumption, and level of daily physical activity). In addition, participants were queried about the therapeutic drugs they were taking, which were recorded according to their trade name or active principle. A single investigator (physician) reviewed all the answers. We analyzed the following groups of drugs because they are related to the treatment of DM and other cardiovascular risk factors: lipid-lowering drugs (statins, fibrates, anion exchange resins, and omega-3 fatty acids), antihypertensive agents (angiotensin converting enzyme inhibitors [ACEI], angiotensin receptor blockers, direct renin inhibitors, beta blockers, alpha blockers, calcium channel blockers, and diuretics), oral antidiabetic (OAD) agents (biguanides, sulfonylureas, thiazolidinediones, mitiglinide, dipeptidyl peptidase IV inhibitors, and others), insulin, uricosurics (allopurinol), thyroid hormone (sodium levothyroxine), psychoactive drugs (hypnotics, anxiolytics, antidepressants, antipsychotics, and others), and nonsteroidal anti-inflammatory drugs (NSAIDs) (salicylates, propionic acid, acetic acid and enolic acid derivatives, cyclooxygenase 2 inhibitors, and others). For each therapeutic drug group, another category was created for use when the participant could not remember the name of the drug, but knew why it had been prescribed (“does not recall the medication”).

A previously trained nurse conducted a physical examination. Weight and height were measured following standardized methods. The body mass index (BMI) was calculated as weight (kg)/height2 (m). Two blood pressure measurements separated by 1 to 2min were obtained with the patient in a sitting position, using a blood pressure monitor (Hem-703C, Omron; Barcelona, Spain) according to the manufacturer's instructions; the mean of the 2 measurements was used in the analyses.

Fasting blood samples were drawn for all participants. Patients with fasting capillary blood glucose<140mg/dL and those who were not receiving DM treatment additionally underwent oral glucose tolerance testing. Samples were immediately centrifuged and stored at −18°C (maximum 15 days) until transfer to a centralized biobank, where they were stored at −80°C until analysis.

Obesity was defined as a BMI≥30.

Hypertension was established when a patient was receiving antihypertensive medication or when systolic pressure was ≥140mmHg and/or diastolic pressure was ≥90mmHg.

Dyslipidemia was established when a patient was under treatment with lipid-lowering drugs or when plasma low-density lipoprotein cholesterol level was ≥100mg/dL or triglyceride level was ≥150mg/dL or high-density lipoprotein cholesterol level was ≤40mg/dL for men and ≤50mg/dL for women.

The diagnosis of DM and carbohydrate metabolism abnormalities was based on the results of plasma glucose measurement, according to the 1999 World Health Organization criteria.16 Carbohydrate metabolism changes included the following categories: impaired fasting glucose (IFG), impaired glucose tolerance (IGT), IFG+IGT, known DM, and unknown DM.

Continuous variables are expressed as the mean (standard deviation) and categorical variables as percentages. Comparison of means with variables having 3 or more categories was performed using analysis of variance, and adjusted for potential confounding variables (age, sex, BMI, hypertension, and carbohydrate metabolism abnormalities). Associations between categorical variables and calculation of odds ratios (OR) with 95% confidence intervals (95%CI) were carried out using a logistic regression model, in which the dependent variables were use or no use of each therapeutic drug group and the covariates were: age, sex, BMI, hypertension, carbohydrate metabolism abnormalities, cardiovascular disease, stroke, and peripheral artery disease. P values <.05 were considered statistically significant.

The prevalence of use of the therapeutic drug groups evaluated is shown in Table 2. The findings show that the use of any type of medication increases with age (adjusted by sex, BMI, obesity, hypertension, and carbohydrate metabolism abnormalities). Furthermore, the percentage of patients who take 6 or more drugs daily rises with increasing age. Men use lipid-lowering agents and allopurinol more frequently than women, whereas women more often take levothyroxine, psychoactive drugs, and NSAIDs.

The risk of being under treatment with any lipid-lowering agent, antihypertensive drug, or NSAID was significantly higher in patients with known DM than in those with normal carbohydrate metabolism. In addition, the possibility of receiving treatment with antihypertensive drugs was higher in patients with other carbohydrate metabolism abnormalities than in healthy individuals. However, although the use of thyroid hormone, uricosurics, and psychoactive drugs was higher in patients with abnormal carbohydrate metabolism compared with healthy individuals, the differences were not statistically significant (Table 3).

The use of lipid-lowering and antihypertensive agents did not differ between the various categories of carbohydrate metabolism abnormalities. This was not the case with NSAIDs: patients with known DM used salicylates (Fig. 1A) more commonly than other types of NSAIDs. Furthermore, the probability of being under treatment with salicylates was significantly greater in this group of individuals (OR=5.41; 95%CI, 2.89-8.97; P<.0001; adjusted by age, sex, and obesity). There were no differences in salicylate use versus other types of NSAIDs in individuals with IFG, IGT, or IFG+IGT (OR=1.50; 95%CI, 0.87-2.58; P=.08), or in patients with unknown DM (OR=1.70; 95%CI, 0.80-3.60; P=.12).

Figure 1.

Use of salicylates and other nonsteroidal anti-inflammatory drugs according to age group and carbohydrate metabolism category. A: Use of salicylates (%) and other nonsteroidal anti-inflammatory drugs according to age group and carbohydrate metabolism category. B: Use of salicylates (%) and other nonsteroidal anti-inflammatory drugs according to age group (years). IFG, impaired fasting glucose; IGT, impaired glucose tolerance; KDM, known diabetes mellitus; NSAID, nonsteroidal anti-inflammatory drugs; UDM, unknown diabetes mellitus. *No carbohydrate metabolism abnormalities.

(0.17MB).

Among the total population, 66% were taking at least one of the therapeutic drugs studied, with a median of 2 drugs per day. In general, the number of drugs used daily increased with age (P<.0001), and was higher in women (P<.0001) (Fig. 2A), obese individuals (P=.005) (Fig. 2B), and patients with known DM (P<.0001) (Fig. 2D). In contrast, hypertension diagnosis (Fig. 2C), and education level (Fig. 2E) were not associated with use of a larger number of drugs.

Figure 2.

Number of drugs used daily according to age groups, sex, body mass index, hypertension, carbohydrate metabolism categories, and education level. A: Number of drugs used daily according to age group and sex. B: Number of drugs used daily according to age group and body mass index. C: Number of drugs used daily according to age group and hypertension diagnosis. D: Number of drugs used daily according to age group and carbohydrate metabolism category. E: Number of drugs used daily according to age group and education level. BMI, body mass index; HT, hypertension; IFG, impaired fasting glucose; IGT, impaired glucose tolerance; KDM, known diabetes mellitus; UDM, unknown diabetes mellitus. *No carbohydrate metabolism abnormalities.

(0.36MB).

Concomitant use of the therapeutic drug groups studied is summarized in Table 5.

With regard to the limitations of this study, we highlight its cross-sectional design, the fact that only 8 drug groups were analyzed, and the lack of information on dose, treatment duration, treatment adherence, and correct prescription; in addition, the information given by the participants was not confirmed with medical or pharmaceutical registries. Nonetheless, the results obtained in the evaluation of allopurinol and levothyroxine use, which showed a clear, known distribution by sex, are in keeping with the data from other studies,7 which could indicate that bias was not produced during the data collection and analysis.

The study also has some strengths. It is a population study with a representative sample of the Spanish population large enough to detect prevalence differences of around 10% (a value similar or superior to the prevalence of use of the majority of the drugs evaluated). Systematic analysis of the medication, together with the information collected about other diseases and lifestyle, can enable assessment of the determinants of therapeutic drug use in our population.