Tired all the time: Cracking the Fatigue Code in Modern General Practice"

My welcoming smile faltered as Alison sat down and opened her appointment with “I’ve come in because, well, I’m just shattered ALL the time”. If there was ever a presenting complaint that sums up the complexity of modern general practice, it is the classic “tired all the time” patient. With only our history-taking ability and a basic examination, general practice has to filter out those patients in whom this vague symptom might be due to a serious underlying diagnosis. We have to keep in mind that the vast majority of patients won’t have a scary illness and are simply – like all of us – juggling work, home, family, finances, and New Year’s resolutions. I often wish “knackered due to life” was a read code….

Many of us approach this presentation with a mental tick list – we gather the information, and then go through relevant red flags. We ask about mood and sleep, then move on to social factors. Hopefully, by the end of the appointment, I’ll know whether I need to ship Alison off to the shiny building ASAP or offer a listening ear with some signposting to our social prescriber and self-help resources. 

Fortunately, a paper in the BJGP delivers us some useful research on this cohort of patients. This population-based cohort study was done in English primary care between 2007 and 2017. It was a big study, with over 300,000 patients in the “fatigue” group, and 420,000 in the “non-fatigue” control group. For each patient, aged 30-99yrs, they then used the Clinical Practice Research Datalink to look at hospital episodes and national cancer registration data, to see whether there was an increased incidence of certain diseases in one group compared to the other. The control group had to have no coding for fatigue or related presentations within 2 years of the data collection starting, and the fatigue group had at least one episode of fatigue recorded with no evidence of other diagnoses that could be causative. 

The authors comment that, while fatigue is associated with a wide range of conditions including cancer, the chance of having a particular one of these when presenting is unknown. The aim of this paper was to look at the short-term disease risk when a patient presents with fatigue. The current approach in the UK is to follow diagnostic guidelines based on case-control studies examining isolated prodromal features of disease. This is the first time fatigue has been looked at as the primary presentation, and the risk of causative underlying disease ranked based on risk. The list of diseases came from a 2019 Lancet paper that produced the first chronological map of physical & mental health conditions from over 4 million patients in the English NHS. The BJGP authors took the top 237 diseases from this Lancet paper, and used these as their list of conditions.

Women made up 67% of the fatigue group, higher than in the non-fatigue cohort. The peak age for women presenting with fatigue was 30-49yrs of age. Among men, patients who presented with fatigue were generally older, with a peak between 50-69yrs. 

The results showed variation between men and women. In men, 127 of the 237 diseases (54%) were more common in the fatigue group. In women, it was 151 of the 237 diseases (64%). The diagnoses most strongly associated with a fatigue presentation were: depression, respiratory tract infections, insomnia/sleep disturbances, hypo/hyperthyroidism (women only) and cancer (men only).  

Of particular importance was that in men who presented with fatigue and were over the age of 80yrs, cancer was the 3^rd most common disease with an absolute excess risk of 3.65%. Men presenting with fatigue had double the risk of having cancer compared with the non-fatigue group. In women at all ages, cancer risk was relatively infrequent, and only the 13^th most common disease in the 80+yrs age group. 

The paper really is a fascinating read if looking at the original literature floats your boat. Suffice to say, the takeaway from this is a shift in approach. The current medical model tells us to work through a list of all possible diseases for which fatigue is a symptom. I suspect if we did this in reality, we’d all be pretty tired too…

But the shift from the scattergun approach of current guidance that suggests “hmm, dengue fever can cause fatigue – better check for that”, means that instead we can combine our history-taking and diagnostic skills with a ranked probability, and focus on the conditions most likely to have caused the presentation. It feels like research has finally caught up with the experience of GPs – and that has to be a good thing for 2025.