Antidepressants as risk factor for ischaemic heart disease: case-control study in primary care
BMJ 2001; 323 doi: https://doi.org/10.1136/bmj.323.7314.666 (Published 22 September 2001) Cite this as: BMJ 2001;323:666
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
The finding of a link between increasing doses of dothiepin (1) and
development of ischaemic heart disease is an important one, which if
genuinely causative should influence our prescribing for depression. It
deserves careful scrutiny.
The authors have found an effect more than five years after
prescription of dothiepin which is at least as strong as that for more
recent exposure. This challenges biological plausibility as their listed
mechanisms all operate only for the duration of prescription and are
essentially reversible. There are other explanations for this finding.
Tricyclic antidepressants may directly damage the myocardium in a less
reversible manner, inducing cardiac enzyme release and antimyosin antibody
response (2), an effect which demands further investigation in humans.
Perhaps a more likely explanation is that any relationship with drug
therapy is confounded by depression itself as a risk factor for ischaemic
heart disease, and the authors acknowledge this. If so, why is there no
link between SSRIs and ischaemic heart disease in this study? In the same
issue of the BMJ, the link between serotonin reuptake properties of
antidepressants and increased risk of gastrointestinal bleeding is
confirmed (3), so SSRIs may in fact protect against cardiovascular disease
via their antiplatelet effects.
Finally we should consider the most basic test of causation applied
to any association in an observational study, that of temporality. The
outcome in this study is diagnosis of ischaemic heart disease rather than
onset of the condition itself. Occult cardiovascular disease usually
precedes the onset of symptoms by several years, and it is possible that
ischaemic heart disease may have predated dothiepin prescription in at
least some cases in this study. The ‘vascular depression’ hypothesis
ranks occult vascular disease as a potent risk factor for late-life
depression (4,5)which could obviously coexist with occult ischaemic heart
disease. The consequent prescription of dothiepin could fully explain
this association, particularly as the study population is predominantly
elderly.
The association of tricyclic antidepressant therapy in normal doses
and ischaemic heart disease is very worthy of further investigation via
other methodologies including cohort studies and biological studies of
mechanisms, but it would be premature on the basis of this paper alone to
alter prescribing. The emphasis should remain on dothiepin’s acute and
profound cardiotoxic effects in overdose and in those with existing
cardiac disease, which are reasons enough to be cautious about its use.
1. Hippisley-Cox J, Pringle M, Hammersley V, Crown N, Wynn A, Meal A,
Coupland C. Antidepressants as risk factor for ischaemic heart disease:
case-control study in primary care. BMJ 2001; 323: 666-669
2. Marti V, Ballester M, Udina C, Carrio I, Alvarez E, Obrador D, et al.
Evaluation of myocardial cell damage by In-111-monoclonal antimyosin
antibodies in patients under chronic tricyclic antidepressant treatment.
Circulation 1995;91:1619-1623.
3. Walraven van C, Mamdani MM, Wells PS, Williams JI. Inhibition of
serotonin uptake by antidepressants and upper gastrointestinal bleeding in
elderly patients: retrospective cohort study. BMJ 2001; 323:655-657
4. Alexopoulos GS, Meyers BS, Young RC, et al. “Vascular depression”
hypothesis. Arch Gen Psychiatry 1997; 54:915–22
5. Thomas AJ. Ferrier IN. Kalaria RN. Perry RH. Brown A. O'Brien JT. A
neuropathological study of vascular factors in late-life depression.
Journal of Neurology, Neurosurgery & Psychiatry. Vol 70(1) (pp 83-87),
2001
Kevin Lewis, Consultant Psychiatrist; Joe Reilly (Corresponding
Author), Consultant Psychiatrist, Tees and North East Yorkshire NHS Trust,
Parkside Community Mental Health Centre, Park Road North, Middlesbrough
TS1 3LF
email reilljg@lineone.net;
Alison Coak,
Assistant Psychologist; Sue Hunter, Principal Pharmacist, Tees and North
East Yorkshire NHS Trust, St Luke's Hospital Marton Road, Middlesbrough.
Competing interests: No competing interests
Editor- The paper from the Division of General Practice at the
University of Nottingham disturbed me. In my facial pain patients I have
found dolesulpin (dothiepin) very effective in the treatment of atypical
facial pain. This condition is defined as pain for which no other cause is
apparent, which does not respond significantly to standard non-opiate
analgesic agents and which is usually accompanied by early morning waking.
The dose needed ranges from 25mg to 225mg nightly. The correct dose for
the individual patient is that which corrects the sleep disturbance. Six
months treatment is usually effective with the occasional patient needing
twelve.
After reading the paper I spent the next few hours checking my
records for all patients prescribed dolesulpin by their GP's, at my
suggestion, in the past year. I wrote suggesting that their GP's comsider
switching them to amitriptyline, at the same dose, though in the past I
have found this agent less effective than dolesulpin and more difficult
to establish the individual effective dose. I have not found lofepramine
effective in this group.
Dolesulpin was introduced more than thirty years ago. Its rapid
correction of early morning waking and relief of other symptoms in three
weeks, with tolerable side effects, have led to its continued use. I have
read Peter Cansfields' and W P Plummers' responses on your web site and
fear I may have acted too quickly in response to this paper. On
medicolegal grounds alone we need an authoritative opinion.
Competing interests: No competing interests
Dear Editor,
Hippisley-Cox et al. present the results of a case controlled study
in which a statistically significant association is found between past use
of dothiepin (dosulepin) and ischaemic heart disease. The associations
between other tricyclic antidepressants and other classes of anti-
depressant with ischaemic heart disease are not significant.
It is not correct to conclude from this analysis, as the authors do,
that 'the increased risk [of ischaemic heart disease] is due to
dosulepin.'
There is, in fact, an increased risk with all the antidpressants
reported and the confidence intervals for all of them overlap
considerably. All confidence intervals include the odds ratio of 1.63 for
taking 'any antidepressant ever.' It is possible for groups to be
statistically distinct when their confidence intervals overlap, but this
distinctness has not been demonstrated. The authors have not presented any
analyses which suggest that the odds ratio for any individual
antidepressant differs from this overall value.
In order to do so, they would need to perform a test of heterogeneity
between the individual antidepressants, or classes of antidepressants.
This is the equivalent in logistic regression to an analysis of variance
in ordinary regression.
Alternatively, if they believed a priori that dothiepin was different
they could carry out various tests of whether its odds ratio was different
to that of the other antidepressants or groups of antidepressants. This
means making a direct comparison between the groups, not simply comparing
p - values for comparison of the odds ratios with the 'null' value of one.
(Comparisons between dothiepin and other antidepressants would not be
valid, if the investigators simply found during analysis that dothiepin
had the most extreme value for the odds ratio and then tested it).
This study is also unable to distinguish the effects of
antidepressant treatment from any possible direct effects of depression.
The prescription of antidepressents seems to be almost completely
confounded with the illness of depression itself. Associations are
reported between higher doses of dothiepin and longer treatment with this
drug and ischaemic heart disease, but there is likely to be a strong
association between the severity of depression and the dose and length of
any antidepressnat treatment.
It is interesting to note that only a third of patients were treated
with a dose of dothiepin which would have had any antidepressant activity
(>100 mg per day) and only a third were given more than four
prescriptions, when it is recommended that antidepressant treatment should
continue for at least six months after the remission of depressive
symptoms. If antidepressants do have health risks, it is worrying that
they are being used in treatment regimes which are likely to be
ineffective.
Another, more esoteric, statistical point is that when matched case
controlled studies are analysed, it is good practice to include the
factors which were used for matching in the logistic regression equation
when undertaking analysis. The authors do not specifically state that they
have included age sex and general practice as factors in either the raw
analysis or the adjusted analysis of their data. In general, to do so will
increase the effect size of any risk factor under investigation, provided
that the factors used for matching are genuinely confounding variables (If
they are not, there is no point in matching!).
Competing interests: No competing interests
Julia Hippisley-Cox et al claim they have found good evidence for an
association between dosulepin (Dothiepin) and subsequent ischaemic heart
disease. However this interpretation of their results appears unwarranted.
As they state, depression is a known independent risk factor for
ischaemic heart disease. Their best estimate for the increased risk due to
depression alone is their adjusted odds ratio of 1.41. Adjusted odds
ratios for most of their different groupings of antidepressants are of
this order and all include the value 1.41 well within their confidence
intervals.
Moreover where antidepressants are more likely to have been
prescribed for reasons other than depression (‘amitritptyline ever’ and
‘other antidepressant’ categories) the best estimates of adjusted odds
ratios fall much closer to 1.0.
Historically amitriptyline and dosulepin (Dothiepin) have greater
cumulative use and therefore representation than some of the other groups
and statistical significance appears to be related to this.
As the authors state, they were unable to control for level of
depression and it is not unreasonable to postulate that a ‘dose response’
relationship between depression and antidepressant exposure is related to
depth and severity of depression.
Whilst there are short-term cardiovascular effects with most
antidepressants that may precipitate myocardial infarction, the argument
for cardiotoxicity many years after exposure seems less plausible.
However, depressive illness is often recurrent and may be the underlying
risk factor operating over longer time periods.
Competing interests: No competing interests
Depression not Dosulpin increases IHD
The Editor
Hippisley-Cox and colleagues reported1 previous use of the
antidepressant dosulepin (dothiepin) is associated with increased ischemic
heart disease (IHD) and suggested the relationship may be causal. However,
we believe there are good reasons for a different interpretation.
The
unadjusted odds ratios in table 2 run from 1.28 to 1.73 for all types of
antidepressants as well as for those receiving no antidepressant
treatment. Taken together, an alternative interpretation is that it is
depression itself, rather than antidepressant use, that is the main factor
driving the subsequent increase in IHD. The apparent specificity for
dosulpin might well be, as the authors suggest, due to its more frequent
use in more severe depression. There have been a large number of
prospective studies demonstrating depression to be an independent risk
factor for the subsequent development of IHD and the relative risk in
these studies is higher than that found in the report by Hippisley-Cox,
e.g. 4.5 by Pratt et al 19962 and 2.1 by Ford et al 19963. These studies
also found a dose-response relationship between the severity of depression
and the risk of IHD. This would be consistent with the view that it is
depression itself that is a risk factor for IHD and that dosulpin (and
other antidepressants) may not have any direct effects.
Furthermore, there
are a number of potential biological mechanisms by which depression could
either lead to or exacerbate IHD. For example, depression has robust
associations with abnormalities in platelet activity and raised levels of
proinflammatory cytokines4 which would increase the atherogenic process.
Such mechanisms have a greater biological plausibility for explaining an
increase in IHD than those mentioned in the report to try and explain the
association with dosulpin, which are more relevant to increasing
arrhythmias and sudden cardiac death.
Dr A.J.Thomas
Lecturer in Old Age Psychiatry
a.j.Thomas@ncl.ac.uk
Prof J.T.O'Brien
Professor of Old Age Psychiatry
j.t.o'brien@ncl.ac.uk
Department of Psychiatry,
University of Newcastle upon Tyne,
Wolfson Research Centre,
Newcastle General Hospital
NE4 6BE
We are not aware of any competing interests.
1. Hippisley-Cox J, Pringle M, Hammersley V, Crown N, Wynn A, Meal A,
et al. Antidepressants as risk factor for ischemic heart disease: case-
control study in primary care. BMJ 2001;323:666-9.
2. Pratt LA, Ford DE, Crum RM, Armenian HK, Gallo JJ, Eaton WW.
Depression, psychotropic medication, and risk of myocardial infarction.
Prospective data from the Baltimore ECA follow-up. Circulation
1996;94(12):3123-9.
3. Ford DE, Mead LA, Chang PP, Cooper-Patrick L, Wang NY, Klag MJ.
Depression is a risk factor for coronary artery disease in men: the
precursors study. Archives of Internal Medicine 1998;158(13):1422-6.
4. Musselman DL, Evans DL, Nemeroff CB. The relationship of depression to
cardiovascular disease: epidemiology, biology, and treatment. Archives of
General Psychiatry 1998;55(7):580-92.
Competing interests: No competing interests