Ann Robinson reviews the latest research from the top medical journals

JAMA

US diet—improving, but could do better

This US study of nearly 44 000 people who monitored their food intake from 1999-2016, showed a trend towards eating fewer low quality carbohydrates (mostly added sugar) and more high quality carbohydrates (such as whole grains), plant protein, and polyunsaturated fat. But overall, 42% of energy intake was still derived from low quality carbohydrates and the intake of saturated fat remained above 10% of energy. It’s a shift in the right direction and shows that population wide dietary change can happen. The usual caveats apply; self-reported dietary data are subject to error and variation, and the population sampled may have changed demographically over time. Most importantly, there’s no way of knowing whether these dietary changes had any impact on health outcomes.

NEJM

Asthma in African-American children and adults—what works?

African-American asthmatic patients in the US have higher morbidity than white patients and are under-represented in trials. Two US prospective, randomized, double-blind trials, one in children and the other in adolescents and adults, studied poorly controlled asthmatics with black ancestry, which the studies defined as at least one grandparent who identified as black. It found that 46% of the children improved with quintupling the inhaled corticosteroid dose (to 250 μg fluticasone twice a day) or doubling the fluticasone and adding a long acting beta agonist, salmeterol. Adolescents and adults, by contrast, responded better to added salmeterol than to an increase in fluticasone. A potential worry is that the highest dose of inhaled steroid resulted in adrenal axis suppression in children under 8 (as measured by reduced urinary cortisol) although the study wasn’t powered to assess the impact on growth or other harms. Interestingly, the phenotype and biomarkers didn’t help to predict response to a particular therapy, but larger trials would be needed to explore this further. 

Ustekinumab for ulcerative colitis—a new option

This study found that ustekinumab is more effective than placebo in induction and maintenance therapy in moderate-severe ulcerative colitis, even in patients who are still having treatment with other drugs (conventional or biologics) or have had in the past.  For instance, clinical remission at week 8 among patients who received intravenous ustekinumab was around 15% compared with 5% given placebo, with similar rates of adverse events. The randomised withdrawal design of the phase 3 study meant that only those patients who had a response to induction therapy with ustekinumab (clinical remission at week 44) underwent a second randomization in the maintenance trial, which may have skewed the results. As you’d expect, patients who had done poorly on other biologics were less likely to respond to ustekinumab.

Lancet

Drugs for severe hypertension in pregnancy

This open label randomised controlled trial, the first to directly compare the three oral drugs most commonly used to treat severe hypertension in pregnancy (blood pressure >160/110 mm Hg), found that oral methyldopa, along with nifedipine and labetalol, are effective and safe. Of the three, methyldopa was the least effective (76% of women achieved blood pressure control within six hours on methyldopa compared with 84% on nifedipine). More neonates born to mothers on nifedipine were admitted to intensive care units, mainly because of low birthweight although, reassuringly, rates of stillbirth, neonatal mortality, and morbidity didn’t vary between the groups.

JAMA Internal Medicine

Use of insulin in older people

This cohort study of 21 531 adults found that patients in poorer health were most likely to continue to use insulin at the age of 75 and that subsequent discontinuation of insulin over the next four years was more common in healthier patients even after accounting for level of glycaemic control. This is despite the fact that older people in poor health who continue their insulin are at increased risk of hypoglycaemia with limited advantage to their future health. Guidelines say that doctors should consider reducing treatment intensity as health status gets worse, but in practice, we may not be grasping the nettle. The results need to be treated with caution. It is observational so causality can’t be proven, although the cohort was representative of the population at large. The fact that this study specifically studied older people is welcome. Clear guidance on the safe use of insulin in older people as they become increasingly frail and unwell is needed.

Annals of Internal Medicine

Follow up after adenomas—what’s best?

Screening by colonoscopy or faecal immunochemical testing (FIT) for colorectal cancer at the age of 50, is known to reduce the risk of subsequent colorectal cancer death by allowing removal of precancerous adenomas and early detection. But what is the safest option for follow up of patients once they’ve had adenomas removed? This modelling study suggests that high-intensity surveillance by colonoscopy after three years for high risk adenomas (HRA) and five years for low risk adenomas (LRA) provides “modest but clinically relevant benefits” over low-intensity surveillance (colonoscopy after 10 years for LRA and 5 years for HRA) or no further colonoscopy. The cost is acceptable within the US healthcare system ($30 000 QALY difference between low and high intensity surveillance.) Benefit is less at older ages. Colonoscopy isn’t available or acceptable to all patients worldwide and some may prefer a small increased lifetime risk of colorectal cancer in exchange for not having to undergo intrusive testing. Longer term follow up (over 20 years) and less intrusive alternatives to colonoscopy are sorely needed.

Ann Robinson is an NHS GP and health writer and broadcaster