Unexpected weight loss is one of the non-specific signs of cancer. However, there are many non-cancer reasons for weight loss and so a key question facing GPs is how to prioritise patients presenting to primary care with unexpected weight for further investigation.
Two measures that might help guide prioritisation are the amount of weight loss and over what time period it occurs, but the association of these measures with underlying cancer is currently poorly defined. Research led by Dr Brian Nicholson and Dr Jason Oke (Nuffield Department of Primary Care Health Sciences) in collaboration with colleagues from the Kaiser Permanente Washington Health Research Institute, Seattle, set out to quantify the association between cancer diagnosis rates and measured weight loss in primary care.
In UK primary care records, weight loss is only recorded if the GP is sufficiently concerned, leading to incomplete data. By contrast, in the USA, people attending primary care are routinely weighed and so weight data is more complete.
The study, published in the Journal of Cachexia, Sarcopenia and Muscle, analysed recorded weight over time for 43,302 people over 40 years old who are registered with a Kaiser Permanente Washington primary care provider. Reductions in measured weight were positively associated with a subsequent diagnosis of five types of cancer (pancreatic, myeloma, gastro-oesophageal, colorectal and breast) and cancers at stages 2 and 4. The risk of cancer diagnosis increased linearly with the amount of weight loss, independently of the patient’s age, sex, starting weight, weight measurement interval and other medical conditions.
This study confirms the importance of weight loss as a sign of undiagnosed cancer and recommends that instead of arbitrary cut-offs for the interval of weight loss, clinicians should assess the percentage of weight lost and the patient’s age independent of the measurement interval.