Ann Robinson’s research reviews—7 August 2020

Ann Robinson reviews the latest research from the top medical journals

Association of hormone replacement therapy and breast cancer

This long term follow-up of two placebo controlled, randomised studies involving over 27 000 women, found that women who had undergone a hysterectomy and took oestrogen alone (conjugated equine estrogen (CEE)) had a slightly lower risk of breast cancer than those who took placebo (annualised incidence 0.3% v 0.37%). This equated to only two fewer deaths from breast cancer, and two fewer after breast cancer for every 10 000 person-years of prior CEE use. However, on a population basis that would scale up to a lot of women. The same possible advantage didn’t hold for women who had not undergone hysterectomy who had to take a progestogen (medroxyprogesterone acetate) in addition to CEE. They had a significantly higher risk of breast cancer compared with placebo (0.45% v 0.36%) though, importantly, no increased risk of dying from breast cancer.  There are lots of gaps in the data: for instance, the findings aren’t applicable to other doses and types of HRT, and information about breast cancer recurrence wasn’t available. In a broadly reassuring landscape of HRT and breast cancer data, every woman must surely be encouraged and helped to decide for herself.

JAMA doi:10.1001/jama.2020.9482

No bones about it: hip fractures are falling

Good news; the incidence of age adjusted hip fractures is decreasing in the US, confirming other studies among men and women from many countries. This huge cohort study from the Framingham Heart Study (over 105 000 person-years followed up over 40 years) found that people born more recently have fewer hip fractures for a given age. Falls in smoking and heavy drinking as well as improved detection and treatment of osteoporosis were all associated with the observed fall of 4.4% per year from 1970 to 2010. Other risk factors such as underweight, obesity, and early menopause didn’t change much over the 40 year period. There was a lack of data on bone mineral density data and medications across the study period, especially in the earlier years. Despite the caveats, public health measures aimed at smoking cessation and limiting heavy drinking seem to be paying dividends in terms of bone health.

JAMA Intern Med doi:10.1001/jamainternmed.2020.2975

Poor herd immunity in Geneva despite high covid-19 rates

Herd immunity seems to be disappointingly low in Geneva, Switzerland, despite high prevalence of covid-19, according to this population based study that tested randomly selected participants and household members over the age of 5 years for SARS-CoV-2 IgG antibodies on a weekly basis and used Bayesian modelling to estimate seroprevalence in the Geneva population. There were 5000 reported clinical cases over 2.5 months in a population of 500 000 (1%) and an estimated 11.6 infections in the community for every reported confirmed case. The seroprevalence of antibodies was estimated as rising from 4.8% in the first week to 10.8% by the fifth week. Participants aged 20-49 years were most likely to be seropositive, and those under 9 or over 65 years were significantly less likely to be seropositive. Is that because the young and old were less likely to be out and about than those of working age, less likely to seroconvert, or a mix of both? Assuming that having antibodies confers immunity; these findings of relatively low seroprevalence don’t provide much reassurance for authorities hoping to ease restrictions on the basis of widespread immunity.

Lancet doi:10.1016/S0140-6736(20)31304-0

Hydroxychloroquine with or without azithromycin; doesn’t help against covid-19

Are hydroxychloroquine and azithromycin effective and safe against covid-19? This Brazilian multicentre, randomised, controlled trial of hospitalised, non-ventilated covid-19 patients with mild to moderate disease found no significant difference in clinical status (using a seven point scale ranging from good health to death) after 15 days between those treated with standard care alone, those given hydroxychloroquine, and those given hydroxychloroquine and azithromycin. Hydroxychloroquine caused more QT prolongation and abnormal liver enzymes than the control group, and there were more reports of adverse effects in the hydroxychloroquine and azithromycin group (39.3%) or hydroxychloroquine alone (33.7%) than in the group given neither drug (22.6%). However, this study has some important limitations: it wasn’t blinded, the design cannot rule out substantial benefit or harm of the drugs, there were deviations from the protocol, some of the patients had already been prescribed the drugs in the past, and some patients were enrolled on the trial up to 14 days after the onset of symptoms, by which time the drugs may have had less impact.

N Engl J Med doi:10.1056/NEJMoa2019014

Obesity and covid-19—a complex picture

Obesity is bad for you, especially if you get covid-19. But this retrospective cohort study of 2466 patients from New York City found that the situation is complex. Over a median of seven days in hospital, 25% of the patients died, 22% were intubated, and 2% remained in hospital. In patients under 65 years old, being obese increased the risk of intubation or death, but in those over 65, it didn’t. There are lots of reasons why that that might be so: age trumps obesity as a risk factor, and older people who are obese may be less frail than their skinnier peers. Surprisingly, body mass index wasn’t associated with admission levels of biomarkers of inflammation, cardiac injury, or fibrinolysis. The study is limited as data on body mass index was missing in 28% of patients.

Ann Intern Med doi:10.7326/M20-3214

Ann Robinson is an NHS GP and health writer and broadcaster