Introduction

A recent epidemiological study reported an excess risk of depression requiring pharmacological treatment after treatment with calcium channel blockers and angiotensin converting enzyme inhibitors but not after treatment with digoxin, anti-arrhythmics, nitrates, diuretics, or  blockers.¹ There have also been case reports suggesting depression^2-5 as well as psychosis⁶ after treatment with calcium channel blockers. As depression may promote suicide we investigated possible ecological associations between suicide rates and the rates of use of eight cardiovascular drug groups in 152 Swedish municipalities. In addition, we investigated the risk of suicide in users and non-users of calcium channel blockers who had purchased prescription drugs mainly used to treat hypertension.

	Methods

This report concerns two different studies: firstly, an ecological study with data from Swedish municipalities on rates of suicide and rates of use of eight groups of cardiovascular drug groups; secondly, the hypothesis generated by this study tested in a cohort study on historical data from users of different antihypertensive drugs.

The ecological study
Sweden is administratively divided into 284 municipalities. Suicide rates for men and women standardised for age for these municipalities during the 5 year period 1989-93 were obtained from the epidemiological centre of the Swedish Board of Health and Welfare. Data on incidence of cause specific mortality were missing for eight municipalities. Data on suicide mortality (ICD-9 (international classification of diseases, 9th revision), codes E950-E959 and E980-E989) were available for 152 municipalities in which the expected number of people committing suicide during the 5 year period was more than five men and five women, as assumed from the overall suicide rates in Sweden.

The cohort study
Data on individual prescription drug use in a municipality located in mid-eastern Sweden (population about 20 000 in 1989) have been compiled and studied since the 1970s. All prescription drugs purchased by residents from pharmacies within the municipality are registered according to the ATC system.⁷ For the purpose of the present study all inhabitants of the municipality were identified who during 1988 or 1989 had purchased cardiovascular medications with ATC codes (in 1988 and 1989) C02 (centrally acting antiadrenergic agents, ganglion blocking antiadrenergic agents, peripherally acting antiadrenergic agents, calcium channel blockers, angiotensin converting enzyme inhibitors), C03 (diuretics), and C07 ( blockers). ATC codes for calcium channel blockers and angiotensin converting enzyme inhibitors were later changed to C08 and C09, respectively. The subjects were classified as users or non-users of calcium channel blockers. Mortality data for the cohort until the end of 1994, including the cause of death, were derived from the Swedish mortality register.⁸ Deaths from suicide were defined by the ICD-9 codes E950-E959 and E980-E989. The codes E980-E989 include cases with uncertain intention for suicide. Data on purchased medications and cause of death were linked by the Swedish personal identification number.

	Results

The ecological study
The number of suicides in the 152 municipalities during the 5 year period ranged from 5 to 652. The total number of suicides in the municipalities during this period was 5648. The total population was 7.3 million, and the mean (range) municipality population was 48 042 (13 722 - 679 364) in 1991. Age adjusted suicide rates varied from 0.76 to 3.69 deaths per 10 000 inhabitants per year. The mean (SD) suicide rate for the 152 municipalities was 2.06 (0.49) suicides per 10 000 inhabitants per year. The rates of use were 79.7 defined daily doses per 1000 inhabitants per day for diuretics, 36.8 for  blockers, 22.5 for calcium channel blockers, 20.4 for nitrates, 16.0 for angiotensin converting enzyme inhibitors, 15.3 for cardiac glycosides, 9.4 for low dose aspirin, and 2.9 for lipid lowering agents.

View larger version (0K): [in this window] [in a new window]   Fig 1.   Correlation between rates of use of calcium channel blockers and rates of suicide in 152 Swedish municipalities

View larger version (11K): [in this window] [in a new window]   Fig 2.   Cumulative rate of suicide over 7 years' follow up in 617 users (continuous line) and 2780 non-users (dotted line) of calcium channel blockers

The cohort study
In all, 3397 patients were identified as purchasers of drugs with ATC codes C02, C03, and C07 in 1988 and 1989. Of them, 617 (18.2%) were classified as users of calcium channel blockers (nifedipine, verapamil, diltiazem, and felodipine) and 2780 (81.8%) as non-users. During the follow up, from the date of purchase until the end of 1994, five users of calcium channel blockers (three men and two women, one with uncertain intent) and four non-users (three men and one woman, none with uncertain intent) committed suicide. The 7 year suicide risks were 9.7 and 2.2 per 1000 persons for calcium channel blocker users and non-users, respectively. The difference in suicide risk was significant (P=0.002.) The average annual absolute risk associated with use of calcium channel blockers was therefore 1.1 suicides per 1000 persons. After adjustment for differences in age and sex the relative risk among users versus non- users was 5.4 (95% confidence interval 1.4 to 20.5). Figure 2 illustrates the cumulative suicide rates during follow up of users and non-users of calcium channel blockers.

	Discussion

In the past several groups of cardiovascular drug have been associated with depressive disorders. As suicide is a serious consequence of depression, the current studies were undertaken to evaluate the possible influence of widely used drugs on risk of suicide. The correlation between use of calcium channel blockers and suicide rates found in the ecological study led us to design a cohort study to test if, among users of antihypertensive drugs, subjects using calcium channel blockers had a higher risk of suicide than subjects not using calcium channel blockers.

Diltiazem,² nifedipine,³ and verapamil⁴ have been associated with depressive disorders in case reports and also in a previous epidemiological study that used data on individual prescriptions of calcium channel blockers and antidepressants.¹ The two current studies imply that calcium channel blockers may also promote suicide. In the ecological study the estimated correlation suggests that about one tenth of the intermunicipality variation in risk of suicide is related to the use of calcium channel blockers. In the cohort study the suicide risk in users of calcium channel blockers adjusted for sex and age was fivefold compared with the risk in non-users treated with other antihypertensive agents.

Clinical trials have a limited ability to detect infrequent or late adverse effects or adverse effects resulting in common symptoms. They also often use a number of inclusion and exclusion criteria, thus reducing their generalisability. Studies with a long follow up and studies encompassing the whole population are therefore needed.

In contrast with most clinical trials the current cohort study included all identified users of the study drugs. Consequently the generalisability of the study is high. In the ecological study low populated municipalities with few expected suicides were excluded, so the influence from extreme rates in combination with small number of events was avoided.

In Sweden, as in most other countries, men have a higher incidence of suicide than women. Men are also more likely to be prescribed a calcium channel blocker (unpublished data on file). In our cohort study the estimated rates of suicide were adjusted for differences in age and sex, and, in the ecological study, the suicide rates were adjusted for differences in age. Additional adjustment of the ecological correlations for the proportion of men in the municipalities did not affect the results (no data given), thus we can eliminate age or sex differences as confounders.

Comorbidity such as cardiovascular diseases might have promoted depressive disorders and suicide. In the ecological study this problem was dealt with by adjusting the correlation for each tested cardiovascular drug group with the rate of use of the seven other drug groups. In the cohort study all subjects were treated with antihypertensive drugs. It follows that although no detailed clinical data were available, users and non-users of calcium channel blockers most probably had similar medical backgrounds.

Populations with a high prevalence of cardiovascular diseases also have a high suicide risk. In the ecological study the rates of use of all but one evaluated cardiovascular drug group also correlated significantly with the suicide rates before adjustment for the rates of use of the other drug groups. After adjustment, however, only the rate of use of calcium channel blockers was significantly and positively correlated with suicide rates (see table 1). Accordingly, the increased suicide risk linked to use of calcium channel blockers would seem independent of cardiovascular comorbidity.

Links with depression
Calcium channel blockers are often prescribed to treat angina pectoris. If angina pectoris causes depression, this might form a link to suicidal behaviour. Nitrates are also prescribed to treat angina pectoris, however, and the rates of use of nitrates did not correlate with suicide rates when we adjusted for the rates of use of other cardiovascular drug groups. Neither did the use of angiotensin converting enzyme inhibitors, often prescribed to patients with diabetes, correlate with suicide rates. Thus, it seems unlikely that the presence of angina pectoris or diabetes help to explain the linking of calcium channel blockers to depression and increased suicide risk.