- Ann Robinson, NHS GP and health writer and broadcaster
Careful what you wish for
Do we underestimate the harm done by overdiagnosis as a result of screening? We worry so much about missing pathology and far less about the consequences of overdiagnosis. This interesting retrospective cohort study of nearly 55 000 US women aged over 70 years compared the cumulative incidence of breast cancer in those who continued being screened with those who didn’t. Adjusted cumulative breast cancer incidence was greater among 70-74 years old who were screened compared with those who weren’t (6.1 v 4.2 cases/100 women), but an estimated 31% of breast cancer among screened women was potentially overdiagnosed. For women over 85 years, the cumulative incidence was 2.8 v 1.3/100 women with up to 54% considered to be overdiagnosis. Overall, there was no reduction in breast cancer-specific mortality associated with screening. A flaw of the study was that it couldn’t measure other benefits or harms of screening or other factors that might differ between women who go for screening and those who don’t.
Ann Intern Med doi:10.7326/M23-0133
“Just” a spoonful of sugar
The ubiquity of sugary drinks shows no signs of abating. It’s clear that their nutritionally empty calories contribute to obesity, but is there evidence of other harms? This large study of just under 100 000 women followed up for a median 20.9 years found that having one or more sugary drinks a day was associated with a substantially higher incidence of liver cancer and death from chronic liver disease compared with those who drank a maximum of three sugary drinks per month (18 v 10.3 and 17.7 v 7.1 cases/100 000 person-years respectively). The possible mechanism isn’t clear because the association held even after adjustments for body mass index; it could be that the rapid rise in blood sugar after a drink promotes insulin resistance, which is a risk factor for liver disease. In contrast, the women who drank at least one artificially sweetened drink a day didn’t have a significantly raised risk of liver cancer or chronic liver disease mortality. The observational nature of the study leaves many questions unanswered; the question that really bothers me is how much evidence do we need before action is taken on the scourge of sugary drinks?
JAMA doi:10.1001/jama.2023.12618
One size doesn’t fit all
If I can’t find a large blood pressure cuff in the surgery, I try to persuade myself that the regular size will be fine. But it’s not. This randomised crossover trial of 195 people with a range of mid-upper arm circumferences found that using a regular blood pressure cuff resulted in an average 3.6 mm Hg lower systolic reading among individuals who needed a small cuff. The more common scenario is that we squeeze a bigger arm into a cuff that’s too small. Using a regular cuff in people needing a large or extra-large size led to higher systolic readings of 4.8 and 19.5 mm Hg respectively. So our patients with larger arms may be overdiagnosed and treated, and those with skinny arms may be undertreated; which is the opposite of what I’d have guessed. Further research is needed, but it seems that customised cuffs are warranted. Automated devices are potentially far more accurate than the old sphygmomanometers—but only if we attach the right sized cuff.
JAMA Intern Med doi:10.1001/jamainternmed.2023.3264
Mix and match your exercise regimen
A cohort study of over half a million US adults over a median 10 years assessed the best combination of moderate aerobic physical activity (MPA), vigorous aerobic physical activity (VPA), and muscle strengthening activity (MSA) to achieve the lowest risk of dying—whether from all causes, cancer, or cardiovascular disease. Unsurprisingly, balanced amounts of MPA (150-225 minutes/week), VPA (≤75 mins/week), and MSA (≥2 sessions/week) combined were associated with a lower risk of mortality compared with an inactive reference group. Adjusted mortality rates represented an approximately 50% lower mortality for all-cause and cancer mortality and an approximately threefold lower mortality for cardiovascular disease mortality among the most active individuals. Which is a darned sight more than you can say for most interventions. There’s a lot of scope for unmeasured variables and recall bias to creep into this kind of study, and fitness fanatics will be frustrated at the lack of detail. The observed risk reduction in this study can probably only be achieved by doing more intense aerobic exercise than current guidelines recommend, although for most of us any activity would be better than none.
JAMA Intern Med doi:10.1001/jamainternmed.2023.3093
Good enough?
Can people living with HIV, who have low level viraemia despite taking antiretroviral therapy (ART), have unprotected sex with their uninfected partner without fear of transmission? It’s a salient question, especially in resource poor settings, where access to blood tests to repeatedly test viral load can be limited. The goal of ART is to achieve undetectable viral loads, with the mantra that undetectable equals untransmittable designed to remove stigma and provide a positive public health message. It’s accepted that very low level viral loads (≤200 copies/mL) carry no risk of sexual transmission, but many people live with higher levels of up to 1000 copies/mL. This systematic review of eight studies and expert input showed that the risk of sexual transmission by individuals with low level viraemia (200-999 copies/mL) was also next to nothing. Low level viraemia isn’t perfect, but it may be good enough.
Lancet doi:10.1016/S0140-6736(23)00877-2