Alex Nowbar reviews the latest research from the top medical journals

Thou shalt have a fishy

Cardiovascular disease is less than desirable, but it does get tiresome hearing about one or other dietary thing that could reduce risk of it. Mohan and colleagues report that a fish intake of at least 175 g (roughly two servings) a week is associated with a lower risk of major cardiovascular disease in people with prior cardiovascular disease, but not in the general population. This is based on large numbers of patients across the globe in a mixture of observational and randomised studies, but this analysis is not informative for changing practice. The authors write: “The consumption of fish (especially oily fish) should be evaluated in randomized trials of clinical outcomes among people with vascular disease.” Well yes, but I’m not sure such a trial would be practical, although there have been randomised trials of fish oil supplements (which have been a mixed bag). We’ll probably never know if a fish diet in general is protective because those more likely to eat fish are also more likely to do all sorts of other things—the classic nutrition research dilemma.

JAMA Intern Med doi:10.1001/jamainternmed.2021.0036

CNS drugs in dementia

Maust and colleagues’ prevalence study of over a million older adults with dementia in the community in the US found that 13.9% of them had prescriptions for three or more central nervous system (CNS)-active drugs for at least 30 days. These could include antidepressants, opioids, benzodiazepines, and antipsychotics. This won’t come as much of a surprise, but it is worrying because these drugs can increase the risk of impaired cognition, falls, and death. More importantly, we don’t know how best to manage these people’s complex needs. The authors comment that they could not assess the appropriateness of the prescriptions. Even if these data had been available, it still would have been debatable whether reducing pain is worth an increased risk of falls, for example. A rate of 13.9% polypharmacy could well be achieving the best outcomes possible in this population—these patients have pain and mental disorders among many other conditions—but I find it hard to believe that three or more CNS drugs can be in many patients’ best interests.

JAMA doi:10.1001/jama.2021.1195

STEMI without risk factors

In a large Swedish registry, 14.9% of ST elevation myocardial infarctions (STEMIs) occurred in people without standard modifiable cardiovascular risk factors. The “big four” of these risk factors (known as SmuRFs) are smoking, hypertension, hypercholesterolaemia, and diabetes. People without SmuRFs were less likely to receive statins, ACE inhibitors or angiotensin receptor blockers, or β blockers at discharge. This seems suboptimal. Worse still, 30 day mortality was higher in those without SMuRFs. This underlines the need for evidence based post-STEMI medication initiation even in people perceived to be low risk. People without SMuRFs who have STEMI must have a substrate for developing cardiovascular disease even if we don’t yet know how to measure this.

Lancet doi:10.1016/S0140-6736(21)00272-5

Gestational diabetes screening

Hillier and colleagues compared two gestational diabetes screening tests in a large randomised trial with the same subsequent guideline-directed management. The one-step approach (a fasting 2 hour oral glucose tolerance test) resulted in more diagnoses of gestational diabetes (16.5% v 8.5%) than the two-step approach (non-fasting 1 hour glucose tolerance followed by a 3 hour fasting test if positive), but there was no difference between the groups in subsequent perinatal or maternal complications from gestational diabetes. This suggests that there may be little advantage to the broader diagnostic approach compared with the two-step approach. This means that interventions can be targeted at the people identified in the two-step approach, and screening for the additional cases isn’t beneficial. This is an example where the additional diagnoses detected by the one-step approach can be considered overdiagnosis—where there was no advantage to knowing or treating the condition for maternal or perinatal outcomes. Perhaps these people are more likely to develop diabetes subsequently, but that doesn’t mean we need to screen for it in pregnancy. There is already enough to worry about that has more immediate implications.

N Engl J Med doi:10.1056/NEJMoa2026028

New drug for early Alzheimer’s disease

Eli-Lily’s new drug donanemab was tested in the 257-patient double-blind randomised TRAILBLAZER-ALZ trial. It was positive for the primary outcome of change in the Integrated Alzheimer’s Disease Rating Scale at 76 weeks, but the aim to reduce disease progression by at least half was not achieved. Patients had to have early symptomatic Alzheimer’s disease with evidence of tau and amyloid deposition on positron emission tomography to be eligible to receive the intravenous injections of donanemab or placebo every four weeks. There was no benefit of donanemab over placebo seen on the secondary outcomes, which further calls into question the significance of the positive primary endpoint. There were no safety concerns.

N Engl J Med doi:10.1056/NEJMoa2100708

Alex Nowbar is a clinical research fellow at Imperial College London.