Drinking patterns and CHD risk in women and men
Tolstrup J, Jensen MK, Tjønneland A et al
BMJ May 2006 doi:10.1136/bmj.38831.503113.7C
What should we advise our patients about the healthiest patterns of drinking alcohol? According to this study from Denmark, it may depend on whether they are male or female. As the article describes, there is a well-established relationship between alcohol intake and CHD risk. The higher risk is associated with high consumption and teetotalism – the source of the J-shaped curve.
Moderate consumption appears to cut risk by a combination of factors, including beneficial effects on HDL and fibrinogen levels, and on platelet aggregation. What this study adds is the relevance of the pattern of drinking through the week, which is much less adequately researched.
The authors invited 160,725 men and women aged 50–65 years to take part in a questionnaire survey to record a detailed profile of their diet and lifestyle. Despite a response rate of only 35 per cent, this created a large study dataset of 28,448 women and 25,052 men. The inclusion of many different lifestyle variables in the questionnaires meant that the statistical analysis could take account of numerous possible confounding factors.
The results indicated that while the relationship between alcohol consumption and CHD risk exists for both sexes, the pattern of drinking may be more important in men. A moderate intake spread evenly across the week appears more beneficial than the same intake consumed at weekends.
For women, this effect was not evident and the weekly amount consumed was the more important predictor of risk. However, the volume of data for the women studied, in whom CHD is rarer, was less adequate than that for the men.
A further problem with the study was the fairly low response rate, creating the possibility of bias. It would also have been interesting to see whether the same results would be obtained for cardiovascular disease, rather than just CHD.
While the authors did not condone binge drinking in women, the implication was that within the recommended weekly limit, the benefits of alcohol might not be offset by distributing it unevenly, as it would be in men, at least in this age group. But in younger women, where habits are becoming established and CHD is much rarer, the chances of benefiting in any way from taking a weekly allowance all on one evening are small, and the potential risks significant. As the authors emphasise, CHD is only one health issue influenced by alcohol, and an unusual one, in that for other issues, alcohol generally increases risk.
- Dr Tim Holt is a GP and clinical lecturer at Warwick Medical School
Long-term medication after MI
Gislason GH, Rasmussen JN, Abildstrøm SZ et al
Eur Heart J 2006; 27: 1153-8
This study, also from Denmark, investigated patterns of long-term compliance with beta-blockers, ACE inhibitors and statins in 55,315 survivors of acute MI. Patients were identified from national administrative registers and followed up at one, three and five years to see if they were taking these drugs.
An important finding, particularly for beta-blockers, was that the probability of having the drug initiated and continued long-term was greatly increased if it was initiated soon after the event, and this relied predominantly on hospital physicians.
In the UK, in theory, the GMS contract provides a safety net for identifying patients who need such treatments but are discharged without them, so this might be different here.
The conclusions should support the use of recall systems to prompt clinicians treating such patients, clinicians who may not have been involved in the early phase of care. This project involved those having a ‘first ever’ MI. There might be wider implications for other acute illness, where the timing of the initiation of medication is clearly important. As well as patient education to improve concordance, there may be issues to do with continuity of care across secondary and primary care. TH
Smoking status and the risk of glucose intolerance
Houston TK, Person SD, Pletcher MJ et al
BMJ, April 2006 doi:10.1136/bmj.38779.584028.55
Using data from the CARDIA study, a cohort of young US adults recruited in 1985–6, the authors report an association between smoking status (active and passive) and the risk of developing glucose intolerance.
The purpose of the study was to investigate the development of risk factors for cardiovascular disease, hence the relatively young age group (median 25 years). Exposure to tobacco carries a higher risk of developing glucose intolerance, with evidence of a dose-response effect. Ex-smokers who did not smoke passively and lifelong non-smokers were least likely to be affected; passive smokers were more likely and active smokers more likely still.
Despite attempts to control for confounding factors, the authors could not rule out the possible influence of unhealthy eating habits in smokers as a source of this association, rather than a direct causal effect. Nevertheless, the results are interesting, because they may help us to recognise individuals at risk among our patients even if the association is not causal. This might assist in the development of screening programmes for type-2 diabetes. Finally, the results suggest that the development of glucose intolerance is a further health risk to add to the list of detrimental effects of passive smoking. TH
The arterial consequences of recreational drug use
Coughlin PA, Mavor AID
Eur J Vasc Endovasc Surg 2006 (ePub ahead of print) doi: 10.1016
This review of published literature, by vascular surgeons from Leeds, searched under cocaine, cannabis and heroin. They focused on complications of vascular access, the problem most frequently seen by vascular surgeons.
False aneurysms most commonly develop in the femoral artery because of the close relationship to the vein, and should be considered as differential diagnoses of a swelling in the groin. They are associated with pain and may be pulsatile, indurated owing to infection, or haemorrhaged. Diagnosis is made using Doppler scan. Management is associated with DVT, graft sepsis and even loss of limb. Intra-arterial injection may be associated with limb ischaemia and tissue necrosis because of toxic effects of the drug on the arterial wall. Injection is associated with intense pain and flushing at the site, and patchy mottling of the limb.
Cocaine is a procoagulant and activates platelets, causing MIs even in normal coronary arteries. It also stimulates the sympathetic nervous system, causing vasospasm and hypertension, and increasing ventricular irritability and the risk of VF. It increases LDL uptake in vessel walls and activates the development of atherosclerosis.
Arrhythmias are also promoted by local anaesthetic properties prolonging QRS and QT intervals. Aortic dissection is more common and severe elevation in BP is especially associated with crack cocaine. Increased use of cocaine, amphetamine and ecstasy is reported to be the greatest cause of stroke in young adults.
Cannabis can also affect the arteries and may be a co-factor in the development of peripheral ischaemia, as seen in Buerger’s disease in young smokers.
- Dr Kathryn Griffith is a GPSI in cardiology in York
Adiposity in adolescents: where does obesity start?
Wardle J, Henning Broderson N, Cole TJ et al
BMJ 2006; 332: 1130-5
This article presents a study begun in 1999. A cohort of almost 6,000 students was followed from age 11–12 (school year seven), for five years. They were randomly selected from 36 London schools, giving a mix of ethnic and social backgrounds, as part of the Health and Behaviour in Teenagers Study (HABITS).
Weight, height and waist circumference were measured annually. Overweight and obesity were defined according to International Obesity Task Force criteria, because adult BMI calculation underestimates adiposity in adolescents, so BMI was corrected to that at 18.
At the beginning of the study, almost 25 per cent of the students were overweight (BMI 25kg/m2 or more) and 4.6 per cent obese (BMI 30kg/m2 or more). This was higher in girls at 29 per cent, and students from lower socio-economic groups, and highest in black girls (38 per cent), while boys who were black had the least proportion of overweight of all boys.
During the study, more students became obese but numbers with healthy weight were unchanged, at 75.9 per cent in year seven and 76.3 per cent in year 11. It therefore appears that persistent obesity is established before age 11–12. The study could be criticised for the small number of students completing measurements for five years. Although 84 per cent of eligible students enrolled, only 36 per cent completed. It may be interesting to consider that the drop-outs may be the proportionally more overweight, who underestimate the problem.
The numbers of Asian students, particularly Asian girls, was low. The Asian girls had the smallest waists and boys, the largest; this is not a good prognosis for future metabolic syndrome.
Children fatter than average at 11 are likely to be fatter at 15 and the paper discourages the notion of benign puppy fat. It is disappointing, although not surprising, to see the numbers of adolescents who are overweight, because they will have all the long-term problems associated with obesity, as well as the additional problems of body image and self-esteem.
I think perhaps Jamie Oliver missed the boat with his school dinners campaign in secondary school; we need to concentrate on nursery and junior school.KG
The importance of silence
Bernardi L, Porta C, Sleight P
Heart 2006; 92: 445-52
This study from Italy compared the autonomic responses to different types of music of 12 advanced conservatoire musicians and 12 age-matched non-musicians. Apparently, musicians are likely to be more attentive, using their dominant hemispheres, while non-musicians use the non-dominant hemisphere.
Subjects were monitored at rest and while listening to six types of music, in random order. The music included ‘classical slow, 70 beats per minute with rhythmic structure’ and ‘techno, 136 beats per minute with a strong non-syncopated structure’.
Music with a faster tempo induced an arousal response, with increases in breathing rate, BP and heart rate. Fast classical (Vivaldi) was as likely to do this as techno. The effect lasted for the whole four-minute test periods.
Sitar ragas, with a tempo of 55 beats per minute, were associated with the slowest heart and breathing rates and BP, although only the heart rate was statistically significant. There was no difference with subjects’ musical preferences, but the responses were amplified in the musicians who had the most marked responses to faster tempos.
The lowest heart and breathing rates and BP were during two-minute periods of silence interspersed between the music. The effect of change of tempo might be important in determining the relaxing effects of music, where composers always include pauses and changes of tempo, even in techno. It is proposed that these alternating fast and slower rhythms may induce relaxation and potentially be of benefit in cardiovascular disease.
The paper concludes that even short periods of music can induce an arousal response that is most marked with fast tempo, whatever the style. It seems that dad’s Vivaldi and son’s techno produce the same response, but it is spells of silence that are associated with true harmony. KG
Ethnic differences in drug related adverse events
McDowell SE, Coleman JJ, Ferner RE
BMJ 2006; 332: 1177-81
This paper employs a systematic review of the medical literature with subsequent meta-analysis of the retrieved data. The design is gold standard, but the validity of the conclusions is dependent on the quality of the data analysed.
There has been little systematic enquiry into ethnic-specific variations in drug-related adverse events. The authors conclude that different ethnic groups have different risks for important adverse events for cardiovascular drugs.
There was a threefold increased risk of angio-oedema due to ACE inhibitors among black patients and a similar increase in ACE reported cough among East Asian patients. There was a
1.5-fold increase in the risk of intracranial haemorrhage with thrombolytic therapy for black patients.
The authors hypothesise that ethnicity may be a surrogate marker for a genetic make-up or other cultural factors associated with increased cardiovascular risk in some patient groups.
There is potential for results to be affected by publication bias, because it is possible that researchers may more (or less) readily report differences in adverse events among these groups. There is no systematic approach to classifying ethnicity or adverse events, so the results may only be regarded as approximate.
The main value of this study is to generate a hypothesis that there may be important variations in adverse event rates among different ethnic groups. The authors recognise this and call for more prospective research examining the issue and better reporting of ethnic-specific adverse event rates in clinical trials.
In the meantime, clinicians need to factor in ethnicity as a potential risk factor for adverse events when prescribing commonly used drugs, such as ACE inhibitors.
- Dr Rubin Minhas is a GPSI in cardiology in Gillingham, Kent, and Medway PCT CHD lead
Lipid-lowering drug treatment in diabetes mellitus
Costa J, Borges M, David C, Carneiro AV
BMJ 2006; 332: 1115–24
This was a systematic review and meta-analysis of randomised, placebo-controlled, double-blind trials, with a follow-up of at least three years, that evaluated lipid-lowering drug treatment in patients with and without diabetes mellitus. The design is gold standard and the trials included are high quality. Two of the authors independently assessed the studies using standardised protocols, with any differences in grading resolved by consensus. Twelve studies were included.
Lipid-lowering drug treatment was found to be at least as effective in diabetic patients as in non-diabetic patients.
In primary prevention, the relative risk reduction for major coronary events was 21 per cent in diabetic patients and 23 per cent in non-diabetic patients. In secondary prevention, the relative risk reductions were similar, at 21 per cent and 23 per cent respectively. However, the absolute risk difference was three times higher in secondary prevention.
When results were adjusted for baseline risk in primary and secondary groups, diabetic patients appeared to benefit more in secondary prevention. The authors conclude that there is strong evidence that lipid-lowering drugs (especially statins) significantly reduce cardiovascular risk in diabetic and non-diabetic patients and suggest that diabetic patients receive more benefit in primary and secondary prevention. RM
Guidelines on the prevention of cardiovascular disease
Heart 2005 Dec; 91 Suppl 5: v1-52
This 52-page document was produced by a group of professional societies. The introduction discusses the role of systematic search and evidence-based medicine, but does not state whether these were undertaken and if a formal methodology was employed.
This detailed guideline attempts to consider all atherosclerotic cardiovascular disease through addressing multifactorial risk factor assessment. Unfortunately, there is greater emphasis on drug interventions than on many lifestyle-related issues that form the foundation of much cardiovascular disease.
There are some algorithms for managing impaired glucose tolerance, which GPs may find useful. The BP-lowering algorithms from the British Hypertension Society (BHS) are recounted, although those from the joint BHS/NICE consultation earlier this year now supersede them. However, within the ABCD algorithm, it is reassuring to see that the ACE inhibitors precede the angiotensin receptor antagonists.
It is disheartening to find errors in BP and lipid targets throughout the document, because they often contradict each other. Any guideline needs to be at least internally valid and basic proofreading errors of this type reinforce the suspicion that the methodological quality of the guideline is poor.
Guidelines make recommendations, but it is unclear how these have been developed from the evidence that the authors selected, although it is clear that patients were not involved.
The sweeping recommendation that all people between the ages of 40 and 80 should be screened for cardiovascular disease is made without providing supporting evidence. This is a cause of concern, because it targets just less than half the UK population.
It is unclear whether the authors were aware of this but practitioners and commissioners need such details before they act. Some of the far-reaching recommendations and conclusions are without evidence. For example, lower lipid targets are advocated for primary prevention, despite the law of diminishing returns and the absence of trials. Clinicians should expect some good reasons why they should be adding to their workload and expending scarce resources by chasing dubious policies.
Applying a recognised guideline-scoring tool, this did not score very well, mainly because of lack of disclosure of conflicts of interest and a rather opaque methodology. Given the potential for a positive bias towards drug treatment recommendations, it would have been reassuring to highlight the degree of editorial independence through stating the conflicts of interest. RM