A cross-sectional study was designed that included patients currently attending a primary health care centre in Barcelona. The primary care clinic Les Corts is a general clinic focussed in both adult and pediatric medicine. The population registered at the clinic totaled 32,318 at the time of recruitment (May 2007), representing 2% of the population living in the city. The study was approved by the Ethics Committee of the reference hospital for the primary health care centre (Hospital Clinic, Barcelona), complying with all laws and international ethics guidelines outlined in the Declaration of Helsinki.

All the patients who had received a diagnosis, according to database information, of type 2 diabetes mellitus and/or a cardiovascular disease were included in the study. Patients were recruited by performing a search of the clinic's database for diagnoses of type 2 diabetes mellitus and/or a cardiovascular disease according to the ICD-10 international codes. Cardiovascular disease was defined as evidence of a previous episode of coronary heart disease, stroke or peripheral arterial disease. The clinic's database was highly accurate for that purpose. The total population recruited was defined as the total number of patients who had at least one visit to the clinic during the period between May 1, 2006, and May 10, 2007. Inclusion criteria for patients included previous diagnosis of type 2 diabetes mellitus and/or cardiovascular disease, and age of at least 18 years. Patients who were included in a palliative care program were excluded.

Data from all participants were recorded regarding sex, age, smoking habit, level of physical activity reported, body mass index (BMI), flu vaccination in the previous winter period, duration of type 2 diabetes or cardiovascular disease since original diagnosis appearing in the database (in years), morbidity (calculated using the Charlson Index, see supplementary table) and treatment (use of lipid lowering drugs, antihypertensive drugs, antiplatelet drugs and anticoagulation therapy). In addition, the most recent values within the previous year of BP, total cholesterol, LDL-cholesterol and HbA1c concentration (in diabetic patients) were taken. Levels of physical activity reported were defined as a categorical variable with 3 categories: 1 for optimum level, 2 for moderate exercise and 3 for a sedentary person. During current nursery care, each patient answers a physical activity questionnaire and then nurses assign a score (1, 2 or 3) depending on their answers. Sedentary patients were defined as the percentage of score 3 among the total. BMI was defined as the individual's body weight (in kg) divided by the square of his or her height (in m). Duration of T2DM or cardiovascular disease was registered as from the year when the condition really began (rather than the data entry of the condition into the database) until the year when the cross analysis was done. The Charlson Index is a continuous measure of comorbidity. It contains 19 categories of comorbidity, which are primarily defined using ICD-9-CM diagnoses codes. Each category has an associated weight, taken from the original Charlson paper,22 which is based on the adjusted risk of one-year mortality. The overall comorbidity score reflects the cumulative increased likelihood of one-year mortality; the higher the score, the more severe the burden of comorbidity.

Lost patients were defined as patients having no open clinical history in the clinic's database when the cross-analysis was done. Manual inspection of the clinical histories corresponding to all lost patients within the previous year was then performed to try to elucidate the cause of the loss (whether due to death, move or other reasons). Dead patients were defined as such only when the investigator saw the death certificate in the database. The cause of death was sought as well in the death certificate. Patients who changed addresses were defined as so if in the clinical history there was a clear statement saying the change of address. All other reasons were classified as “other reasons”.

Diabetes and cardiovascular disease durations were recorded through manual inspection of each individual's clinical history. In the present study, the Charlson Index was calculated creating a model that gave each ICD-9-CM diagnosis from the Charlson Index its corresponding weight (see supplementary table). Then, for each patient, all weights were added up to get the total score of the Index per patient. Means of the Charlson Index score for each group of patients (DM, CVDM, and CVnoDM) were compared. Losses within the year of the cross-sectional analysis (with its cause specified, including mortality data) were assessed through manual inspection of the clinical history of each lost patient. The rest of variables included were taken automatically from the clinic's database using the last value registered in the database within the previous year.

The primary outcome had two endpoints. First, the aim was to compare the cardiovascular risk factors control (BP, cholesterol serum levels, smoking status, levels of physical activity, and obesity), nursery care interventions (flu vaccinations in the previous winter period), glycaemic control in diabetic patients, and pharmacological therapy between DM, CVDM and CVnoDM patients. Secondly, we aimed to compare the clinical burden of these patients which was assessed through the Charlson Index. The secondary outcome was the comparison of the proportion of dead patients (with their cause specified where possible) within the year of the cross analysis between the different groups.

Software STATA.11 was used for the statistical analyses of this study. Results are expressed as means±SD, confidence interval (CI) for continuous variables or as % for categorical variables. Chi-squared tests were performed for categorical variables and ANOVA tests for comparisons between continuous variables. Adjustments for age and sex have been done as necessary. The significance level was set up at p<0.05 (two-sided). Proportion of deaths was calculated as the number of dead patients divided by the total of lost patients in each group (CVnoDM, CVDM and DM) within the year of the cross analysis.

855 patients had established cardiovascular disease and had not been diagnosed with T2DM (CVnoDM), 301 suffered from a cardiovascular disease and also had T2DM (CVDM) and 1012 patients had type 2 diabetes (DM) but had not an established cardiovascular disease. Baseline characteristics and cardiovascular risk factors of these patients are shown in Table 1. The control group was defined as the group of patients with a history of cardiovascular disease but no previous diagnosis of T2DM.

In all three groups, the majority of patients were men, however the proportion of men to women was lower in the DM group (54% as compared to 65% and 71% in CVnoDM and CVDM, respectively). The DM group was significantly younger than the other two groups (70.7 mean age in years as compared to 74.4 mean years in CVnoDM and 74.7 mean years in CVDM). Overall, 11% of patients were currently smoking. The proportion of patients leading a sedentary lifestyle differed among the groups, with a maximum in CVDM (39%), 34% in CVnoDM, and a minimum in DM (27%, p<0.0001). Patients with T2DM had a higher BMI as compared with those with cardiovascular disease but without T2DM (p<0.01). Levels of systolic blood pressure (SBP) and diastolic blood pressure (DBP) were similar in the three groups and followed ADA recommendations (<135/85mmHg in average). Lipid profile (total cholesterol serum levels and LDL cholesterol serum levels) was significantly lower in CVDM as compared with CVnoDM and with DM. Glycaemic control was similar between the two groups with T2DM (CVDM and DM), with HbA1c mean levels less than 7%. There was a significant lower proportion of flu vaccinations in the diabetic patients (12%) as compared with the other groups (65% and 71%, p=0.004).

Data about pharmacological drugs used by the patients are presented in Table 2. In summary, the most important result in Table 2 revealed that patients with established cardiovascular disease were more frequently on antiplatelet or anticoagulation therapy (67% in CVnoDM and 77% in CVDM) as compared with type 2 diabetes but with no history of cardiovascular disease (47%, p<0.01). There was a higher use of ACE inhibitors or ARA II drugs found in patients with diabetes and established cardiovascular disease (64% versus 43%, p=0.003) as compared with patients with cardiovascular disease but without diabetes. No significant differences were observed in the use of lipid lowering drugs among the three groups, with an average use of 61% in total. Patients with both diabetes and cardiovascular disease had a higher use of antihypertensive drugs as compared with non-diabetic patients with cardiovascular disease (control group) (75% versus 61%, p=0.03). There was more than twice the use of beta blockers in the control group as compared with diabetics (28% versus 13%, p=0.009). No differences in the treatment of T2DM were found between CVDM and DM groups (data not shown).

General morbidity (assessed using the Charlson Index, see supplementary table) reached its maximum when both cardiovascular disease and diabetes were present (mean of 3.24, 95%CI 3.03–3.44), then diminished to 2.09 (95%CI 2.00–2.18) if only diabetes was present, whereas the group with less associated morbidity was the one with only cardiovascular disease (mean index 1.83, 95%CI 1.72–1.93, p<0.0001). Results did not change after adjustments for age and sex (results not shown). The Charlson Index was created initially to provide physicians with a rough idea of the general comorbidity facing patients and to give an approximation of ten-year survival predictions (the higher the score, the lower the survival rate).22 According to this statement, in our study, patients with T2DM would have a worse 10-year survival rate as compared to patients with cardiovascular disease but without T2DM, with the poorest 10-year survival rate found in those with both cardiovascular disease and T2DM.

There was on average a 10% of losses in the three groups. The proportion of deaths within the last year in each group did not differ significantly among them (2.2% in the control group, 3.2% in the diabetic, cardiovascular disease group and 1.1% in the diabetic group). Results did not change after adjustments for age and sex (data not shown). 70% of deaths were men in all three groups, with an average age of 79 years in the diabetic group, 83 years in the control group and 84 years in the diabetic group with established cardiovascular disease.

There was almost double the number of cardiovascular deaths in patients with diabetes and established cardiovascular disease as compared with non-diabetics with cardiovascular disease (27.3% versus 14.4%, p=0.02). Cardiovascular deaths were higher in the group with both diabetes and cardiovascular disease (27.3%) as compared with the group with only diabetes (16.7%), but this difference did not reach statistical significance. A similar proportion of deaths due to cancer was found in the three groups (18% on average). These results did not change after controlling for sex and age (data not shown).

Present results show that overall patients with T2DM are younger, and in a larger proportion women and more obese than patients with established cardiovascular disease. Lipid control is more efficiently achieved in patients with both T2DM and cardiovascular disease as compared with patients with only T2DM and patients with cardiovascular disease but without T2DM. This result reflects the fact that physicians make a special effort to reduce lipid profile when both diseases (cardiovascular disease and diabetes) are present. Lipid lowering drugs are used in a similar proportion in all 3 groups. Consequently, the use of statins is now generally widespread in all diabetic patients. Flu vaccinations are given less frequently to patients with T2DM as compared with those with cardiovascular disease. Glycaemic control is similar in diabetic patients regardless of the presence of cardiovascular disease. It must be noted that the three groups shared similar periods of evolution since the time of diagnosis of diabetes or the cardiovascular event (7 years on average) registered in the database. Comorbidity data showed that patients with established cardiovascular disease but without diabetes had a lower comorbidity than patients with T2DM but without cardiovascular disease. Mortality data revealed that there is a similar mortality in patients with diabetes and no history of cardiovascular disease, patients with diabetes and cardiovascular disease, and in non-diabetic patients with established cardiovascular disease. There is a trend towards an increased risk of death if the patient is both diabetic and has established cardiovascular disease. Cardiovascular mortality is almost double in patients with diabetes and cardiovascular disease as compared with non-diabetic patients in secondary cardiovascular prevention. Cancer mortality was similar in the three groups of patients.

The strength of this study lies in its depiction of the real situation of clinical health care practice in patients with T2DM, with and without cardiovascular disease, who attend a primary health care centre in Barcelona. Since this study does not belong to any large epidemiological study in primary care, the present results reflect accurately the clinical management and complications of the patients suffering from T2DM and/or cardiovascular disease in a day-to-day practice. The most important limitation is the study design, a cross sectional study at one year. In order to compare survival rates among these groups of patients, a cohort design with an appropriate follow-up time is needed. Other limitations exist such as that there were not enough cardiovascular events and corresponding data of lost patients may not have been registered correctly in the clinic's database. Indeed, the inclusion of death certificates within the clinic's database was not hundred percent complete, so it remains possible that more deaths occurred but that they were accounted for as “other causes” (misclassification). Therefore interpretation of mortality data must be performed carefully.

Present results are in concordance with the preliminary results of the di@bet.es study,21 a national cross-sectional study about the prevalence of diabetes in Spain and associated cardiovascular risk factors.21 Physicians seem not to treat T2DM as a cardiovascular event equivalent given the lower use of antiplatelet drugs in patients in primary cardiovascular prevention suffering from T2DM. This finding is in accordance with the latest recommendations from the ADA about the use of antiplatelet treatment in primary prevention in people with diabetes.23,24

Regarding clinical practice, it was shown that diabetic patients received lower flu vaccinations. Taking into account that infections in diabetic patients can have a worse prognosis, intervention should be done at primary health care centres aimed at increasing flu vaccinations in patients with T2DM. On the other hand, given the theory that inflammatory status underlies the cause of arteriosclerosis, vascular damage and diabetes,25 flu vaccinations could have an additional benefit in patients with diabetes,26 which future research would need to confirm. In addition, BMI should be reduced in T2DM patients and physical activity promoted in order to ameliorate cardiovascular outcomes.

Future research should focus on performing cohort studies to compare survival rates of CVnoDM, CVDM and DM patients in a primary health care setting since these results show that T2DM confers a higher comorbidity burden than cardiovascular disease without diabetes.

Likewise, preventive programs involving both nursery and medical interventions focussing in an improvement of a healthy lifestyle (diet, exercise, loss of weight, stop smoking, flu vaccinations) in patients with T2DM or in secondary cardiovascular prevention need to be revised and ameliorated in primary care centres.

In summary, special attention to clinical management in patients with T2DM (and without cardiovascular disease) must be made because results show that, even though a similar metabolic control, there is no evidence of a lower clinical burden in patients with diabetes as compared to those in secondary cardiovascular prevention without diabetes in the Mediterranean area.