Association between age and the initiation of antihypertensive, lipid lowering and antiplateletet medications in elderly individuals newly treated with antidiabetic drugs

SIR—The elderly are the age group most affected by diabetes mellitus [1]. Given that the general population is ageing, the number of older individuals with diabetes is likely to increase. Moreover, diabetes is often accompanied by several co-morbidities [2], not to mention cardiovascular disease (CVD), the main complication of diabetes [3]. Antihypertensive [4], lipid-lowering [5] and antiplatelet [6] drugs are known to decrease cardiovascular complications in diabetes. Even though there is little evidence from clinical trials in elderly patients with diabetes, most older patients with this condition require pharmacological treatments in order to reach the clinical targets recommended in guidelines [7–9]. The Canadian Diabetes Association recommends the same targets for older patients than for younger ones (130/ 80 mmHg; LDL-cholesterol 2.6 mmol/l), although more conservative objectives are suggested for individuals with multiple co-morbidities [7]. Similar recommendations are found in guidelines from other associations [8, 9]. In many instances, there is evidence that quality of treatment diminishes with increasing age [10–14]. In a study of 6,626 prevalent users of antidiabetic drugs, the older the individuals were, the less likely they were to use a statin or an angiotensin converting enzyme (ACE) inhibitors [10]. However, since patients could have been using cardiovascular protective treatments before they started oral antidiabetic drugs, the question remains as to whether or not age affects initiation of cardioprotective treatments among individuals with diabetes. The objective of this study was to assess the association between age and the initiation of antihypertensive, lipid-lowering and antiplatelet drugs during the year following commencement of an oral antidiabetic drug among individuals aged 66 years and over. We performed a population-based cohort study using the Quebec Diabetes Surveillance Database (QDSD). The database results from the linking of five public health administrative databases. For individuals who had been diagnosed with diabetes, it includes information on patient demographics, physician and hospital services and data from the public drug plan. From the QDSD, individuals aged ≥66 years were selected if they had initiated an oral antidiabetic drug between 1 January 1998 and 31 December 2002. The date of the first claim formed the index date. We excluded all individuals who (1) had a claim for insulin or any oral antidiabetic drug in the year before the index date, (2) had not been beneficiaries of the provincial prescription drug insurance plan for the entire year before the index date and (3) did not have at least one year of follow-up. From the remaining patients, three sub-cohorts were identified: the antihypertensive drug-free, the antiplatelet drug-free and the lipid-lowering drug-free sub-cohorts. The sub-cohorts were made up of patients with respectively no claim for any antihypertensive, antiplatelet and lipid-lowering drug during the year before the index date. The outcome variable for the antihypertensive drug-free sub-cohort was defined as a prescription for an antihypertensive drug in the year following the index date: likewise an antiplatelet drug for the antiplatelet drug-free sub-cohort, and similarly, a lipid-lowering drug in the lipid-lowering drug-free sub-cohort. The proportion of individuals initiating therapy according to four age groups (66–70, 71–75, 76–80 and ≥81 years) in each sub-cohort was calculated. A trend analysis was done using the Mantel–Hanszel chi-square test (α = 0.05). We computed adjusted odds ratios (OR) with 95% confidence intervals (CI) of initiating an antihypertensive, antiplatelet or lipid-lowering drug treatment according to the age groups. Analyses were performed using SAS, version 9.1 (SAS Institute, Inc., Cary, NC, USA). Access to the data used in this research was authorized by the Commission d’accès à l’information du Québec. During the study period, a total of 48,505 individuals who initiated an oral antidiabetic drug treatment fitted the inclusion criteria. Their characteristics are presented in Table 1. Among the individuals included in the antihypertensive drug-free sub-cohort (n = 15,290 (32%)), the risk of initiating an antihypertensive drug treatment in the year following the index date decreased with age (P for trend <0.001). Similarly, the risk of initiating an antiplatelet or a lipid-lowering drug treatment decreased as age increased in the antiplatelet drug-free (P = 0.119) and lipid-lowering drug- free (P < 0.001) sub-cohorts respectively, although the trend was not statistically significant in the antiplatelet drug-free group (Table 2). Baseline characteristics of the 48,505 individuals included in the total population aA person was deemed to have a cardiovascular disease if there were any outpatient or hospital ICD (International Classification of Diseases)-9 codes for ischaemic cardiac disease, heart failure or cerebrovascular disease or a procedural code of revascularization, coronary bybass or peripheral limb revascularization recorded. A person with a prescription claim for a nitrate was also considered to have cardiovascular disease. Baseline characteristics of the 48,505 individuals included in the total population aA person was deemed to have a cardiovascular disease if there were any outpatient or hospital ICD (International Classification of Diseases)-9 codes for ischaemic cardiac disease, heart failure or cerebrovascular disease or a procedural code of revascularization, coronary bybass or peripheral limb revascularization recorded. A person with a prescription claim for a nitrate was also considered to have cardiovascular disease. Association between age and initiation of antihypertensive, antiplatelet and lipid-lowering drugs in elderly individuals in the year following the start of an oral antidiabetic drug treatment aAdjusted odds ratios (AOR): adjusted for sex, area of residence (metropolitan: urban core of at least 100,000 people and its adjacent municipalities; agglomeration: urban core of at least 10,000 people but <100,000 and its adjacent municipalities; rural: municipalities not included in the former two categories), calendar year of oral antidiabetic drug initiation, and for the following variables measured in the year prior to index date: hospitalization, number of outpatient physician consultations, number of different drugs used, and cardiovascular disease. Association between age and initiation of antihypertensive, antiplatelet and lipid-lowering drugs in elderly individuals in the year following the start of an oral antidiabetic drug treatment aAdjusted odds ratios (AOR): adjusted for sex, area of residence (metropolitan: urban core of at least 100,000 people and its adjacent municipalities; agglomeration: urban core of at least 10,000 people but <100,000 and its adjacent municipalities; rural: municipalities not included in the former two categories), calendar year of oral antidiabetic drug initiation, and for the following variables measured in the year prior to index date: hospitalization, number of outpatient physician consultations, number of different drugs used, and cardiovascular disease. Among elderly who started an oral antidiabetic drug, age had an influence on whether or not they were exposed in the following year to a cardioprotective drug that they had not been using before. Other studies have shown that older age is associated with underuse of evidence-based therapies. In a study of patients with atrial fibrillation, warfarin use decreased with each advancing decade of life, independently of other risk factors and bleeding contraindications [13]. In another study, older patients experienced a decreased probability of receiving thrombolysis for acute myocardial infarction [14]. Lastly, Glynn et al. [11] reported that the likelihood of being exposed to insulin or to oral antidiabetic drugs decreased with increasing age after hospital discharge. However, the question of whether the lower use of evidence-based treatments among older patients reflected appropriate clinical judgment or an inappropriate care gap is difficult to answer [14]. We propose seven hypotheses that could explain the lower odds of initiating cardioprotective drug treatments among the oldest individuals in the cohort. First, individuals who need oral antidiabetics at older age and who were not treated before for cardiovascular risk factors may be healthier. However, in a study of 1,375 patients aged 50–74 years with prevalent diabetes, there was no evidence of a decrease in the prevalence of systemic hypertension or dyslipidaemia among the older age groups [15]. Second, the ‘older older’ individuals may have more co-morbidities and contraindications for drug treatments than the ‘younger older’. As the databases do not capture clinical data, it was not possible to take all co-morbidities into account in the analysis. Third, some physicians may be reluctant to prescribe additional drugs to older patients as the risk of interactions and side-effects is likely to increase [16]. Nonetheless, it has been reported that elderly patients with high cardiovascular risks are often denied statins without there being any meaningful basis for the decision [17]. Many older patients with diabetes would therefore benefit from a thorough examination of the risks and benefits of cardioprotective therapies. Fourth, one might argue that the elderly with frailty or limited life expectancy should not be prescribed protective treatments. Frailty [18] is difficult to assess using administrative data. Also, there were patients whose life expectancy was not long enough to anticipate benefits from long-term therapy and they were not excluded from our study because they survived longer than a year after the index date. Fifth, cost of therapy may be prohibitive. However, since many elderly with diabetes reach their cap on the monthly contribution required under the drug plan, the prescription of cardioprotective drugs would not necessarily add to a patient's financial burden in Quebec. Sixth, there is a lack of clear evidence supporting the clinical benefit of antiplatelet [19] and lipid-lowering therapies in the older patient with diabetes [20]. This could explain partly why the proportion of those who acquired these drugs was low. Nevertheless, considering that the benefits of antihypertensive treatment is well established [21] and that almost all older patients with diabetes would qualify for ACE inhibitors or angiotensin receptor blocker therapy [22], it might be expected that the proportion of individuals newly treated for diabetes who later initiated antihypertensive drugs would be higher than what was observed. Lastly, the seventh hypothesis concerns a latent will to avoid ‘aggressive’ treatment of older individuals as they are ageing. However, as more aggressive clinical interventions such as revascularizations are carried out in older patients with cardiovascular disease [23, 24], it would be unethical not to offer the same level of pharmacological treatment protection, should this pharmacological treatment prove appropriate. This study has some limitations. First, there is yet no clear evidence that all cardioprotective drugs are of benefit in the elderly population with diabetes. Our assumptions are based upon recommendations from clinical guidelines, which are based on data from clinical trials performed in younger individuals. Similarly, since the database does not capture clinical data, it was not possible to assess appropriateness of drug treatments or contraindications. Next, because acetylsalicylic acid can be purchased without prescription, it may have underestimated the proportion of individuals taking antiplatelet drugs. Lastly, analysed data were for the years 1998 to 2003; the situation may have changed since. The recently established evidence of benefits of antihypertensive therapy among the very old [25] may increase future use of these agents. Similarly, an increased use of lipid-lowering drugs may, however, have occurred after 2003 as clinical studies [5, 26] suggested that even the oldest patients benefited from the use of statins, although this might not apply to the very old (>85 years). In conclusion, we observed an age-related gap between the guidelines and clinical practice that may come about from physiologically or medically driven decisions or from unwillingness to treat because of older age. Future research is needed to address these issues more specifically. There is also a need for clinical studies to examine the efficacy of cardioprotective treatments in the older patient with diabetes. Indeed, evaluating the quality of the care process (the use of cardioprotective treatments) is irrelevant if it does not translate into clinical benefits. Antihypertensive, antiplatelet and lipid-lowering drugs are recommended to reduce cardiovascular complications in elderly with diabetes, but whether or not age affects initiation of cardioprotective treatments remains unknown. Our results suggest that exposure to cardioprotective drugs in the year following oral antidiabetic therapy initiation is indeed associated with the patient's age. The age-related gap between the recommendations and clinical practice may come about from physiologically or medically driven decisions or from unwillingness to treat because of older age. The authors thank contributors from the Institut national de santé publique du Québec: Danielle St-Laurent and Louis Rochette for the use of the QDSD, and Valérie Émond and Isabelle Larocque for revising the manuscript. The authors acknowledge the work of Joanne Vidal in editing the manuscript. None. Caroline Sirois is the recipient of a doctoral training award from the Fonds de la Recherche en Santé du Québec (FRSQ). Paul Poirier is a clinical research scholar from FRSQ. Results were presented in part at the 24th International Conference on Pharmacoepidemiology & Therapeutic Risk Management, 17–20 August 2008, Copenhagen, Denmark, and at the 2008 CDA/CSEM Professional Conference and Annual Meetings, 15–18 October 2008, Montreal, Canada.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.