The Greek physician, Herophilus (335–280 BC), compared the pulsation of blood vessels to musical rhythms. He is thought to be the first physician to count the pulse rate. He used a water clock or clepsydra which he took to patients’ houses. Avicenna, or Abu Ali Ibn Sina, was a physician in Persia in the 11th century. ‘Abu Ali placed his hand on the patient's pulse, and mentioned the names of the different districts and continued until he reached the name of a quarter at the mention of which, as he uttered it, the patient's pulse gave a strange flutter. Then Abu Ali repeated the names of different streets of that district and different houses till he reached the name of a house at the mention of which the patient's pulse gave the same flutter…Thereupon he said: “This man is in love with such-and-such a girl, in such-and-such a house…the girl's face is the patient's cure”.’ The English physician, William Harvey (1578–1657), discovered that the arterial pulse was caused by the contraction of the left ventricle. He also identified the right atrium as the source of the heartbeat.1 Nowadays, in addition to feeling and counting the pulse (with analogue or digital timers) or listening to the heart, we have ECG recording and monitoring, both superficial and implantable. Exclude an irregular pulse before using an automatic blood pressure device. Electronic monitoring may be misleading. A colleague rushed to a patient's bedside to treat the life-threatening ventricular tachycardia shown on their cardiac monitor, only to find the patient apparently well. The monitor was displaying electrical interference. This problem is rare nowadays. This is the first of two articles about cardiac arrhythmias in people with diabetes. ‘Valarie Tucker, a 58-year-old procurement manager in Grandview, Missouri, had noticed her heart periodically racing on and off for years. But she shrugged off her symptoms as just nerves and anxiety…[I] came home feeling just fine…after taking just five steps up my stairs, it was clear something was off. I felt exhausted, and suddenly, my heart began racing. It literally felt like it was pumping outside of my chest.’ Valarie, who had diabetes and hypertension, was found to have atrial fibrillation.2 Atrial fibrillation (AF) is the commonest sustained cardiac arrhythmia, or may be paroxysmal. The rhythmic atrial impulses that usually trigger heart beats are replaced by quivering fibrillation causing an irregular, usually fast, heartbeat. AF is often asymptomatic, but may cause palpitations, breathlessness, fatigue, or light-headedness. People have described AF as a fluttering in their chest ‘like butterflies’, or like a bird or ferret jumping around.3 One study estimated that the 1.2 million American people with AF in 2010 would increase to 2.6 million in 2030.4 One would expect people with diabetes, with their propensity for coronary heart disease, to be more at risk of AF than the rest of the population. Study results depend on the groups compared, and which risk factors and/or co-conditions, are accounted for in analysis. An extensive review for the American College of Cardiology of the multiple studies concluded that diabetes is an independent risk factor for AF.5 The Framingham Heart Study found, on multivariate analysis, that diabetes was a risk factor for AF: odds ratio 1.4 for men and 1.6 for women.6 A meta-analysis in 1,686,097 people included 108,703 people with AF. People with diabetes were more likely to have AF: relative risk 1.39 (1.10–1.75). The risk was reduced but not abolished by adjusting for multiple risk factors.7 A study followed 421,855 patients with type 2 diabetes from the Swedish National Diabetes Registry and 2,131,223 age-, sex-, and county-matched controls for 13 years. People with pre-existing AF were excluded. Among people with diabetes, 8.9% had AF compared with 7.0% of people without diabetes. The incidence rate for AF was 15.99/1000 person-years among people with diabetes, compared with 11.89 for people without diabetes. After correcting for risk factors, people with diabetes had a hazard ratio for developing AF of 1.28 (1.26–1.30).8 A Danish study included 253,374 (5%) people with diabetes compared with 4,827,713 (95%) without diabetes. Age-stratified incidence rates of AF for people with diabetes were 0.13 (0.09–0.20) per 1000 person years aged 18–30 years, rising to 20.1 (19.4–20.8) aged 70–100 years. However, compared with people without diabetes the incidence rate ratios in diabetes were 2.34 (1.52–3.60) aged 18–39 years falling to 0.99 (0.97–1.01) aged 75–100 years.9 Higher HbA1c or raised urinary albumin increased the risk of AF: HbA1c ≤ 6.9% (≤52mmol/mol) and normoalbuminuria, adjusted hazard ratio 1.16 (1.14–1.19); HbA1c ≥ 9.7% (≥ 83mmol/mol) and normoalbuminuria 1.35 (1.20–1.52); HbA1c ≤ 6.9% and albuminuria 1.55 (1.50–1.60); HbA1c ≥ 9.7% and albuminuria 2.04 (1.82–2.28).8 The increasing risk of AF with increasing HbA1c has been seen in other studies. A case-control study found that the risk of developing AF increased by 3% per additional year of diabetes duration. People with diabetes with a short duration of diabetes had less risk of AF than those with longer duration: <5 year duration OR 1.07 (0.75–1.51) and >10 year duration 1.64 (1.22–2.20).10 A Taiwanese study of 645,710 patients with type 2 diabetes included 85,198 people on metformin and compared them with the remainder. None of the patients were using other anti-diabetic medication. Over 13 years, 9983 patients developed AF – an incidence rate of 1.5%. ‘After adjusting for co-morbidities and medications, metformin independently protected the diabetic patients from new-onset AF with a hazard ratio of 0.81 (95% confidence interval 0.76–0.86, p<0.0001). Metformin significantly decreased the extent of pacing-induced myolysis and the production of reactive oxygen species.’11 Thiazolidinediones (TZDs) may also reduce the risk of developing AF. A meta-analysis including 130,854 people with diabetes showed that people on TZDs had an odds ratio for developing AF of 0.73 (0.62–0.87) compared with those not using TZDs. This reduced risk was significant only in the observational studies and not in the randomised controlled trials. The risk reduction was largely in people taking pioglitazone rather than rosiglitazone.12 Prompt diagnosis of AF is essential to treat symptoms, and to prevent potentially serious or even fatal complications. Seek the cause of the AF, e.g. valvular heart disease, excess alcohol. Long-term anticoagulant therapy is required in most patients to prevent embolic stroke. Medication is used to restore or maintain sinus rhythm, and to control the heart rate in people with sustained AF. Cardioversion may be used to restore sinus rhythm. If drug treatment is unsuccessful or inappropriate, catheter electric ablative treatment is used. Some patients may need cardiac pacing. One study of 228 people with paroxysmal AF found 65 with diabetes or impaired fasting glucose. After catheter ablation AF was more likely to recur in people with an abnormal glucose metabolism (18.5% vs 8.0%) than in those without.13 Direct current cardioversion was used to treat persistent AF and was successful in 66.6% of people with type 2 diabetes compared with 84.3% of people without diabetes. After successful cardioversion, only 37.5% of people with diabetes remained in sinus rhythm at about 10 weeks, compared with 61.8% of people without diabetes.14 In a cohort study of 9749 people with atrial fibrillation, 29.5% had diabetes. Compared with those without diabetes, people with diabetes were younger, and more likely to have coronary heart disease, heart failure, hypertension, chronic kidney disease, stroke, and poorer quality of life. They were more likely to be anticoagulated but had a higher mortality – adjusted hazard ratio aged <70 years 1.63 (1.04–2.56), ≥70 years 1.25 (1.09–1.44); both cardiovascular and non-cardiovascular mortality were increased, as was hospitalisation, but not bleeding.15 Suggested mechanisms for the increased risk of AF among people with diabetes include myocardial fibrosis, glucose fluctuations, oxidative stress, and inflammation. This can cause atrial structural remodelling and dilatation, and also electrical, electromechanical, and autonomic remodelling.5 AF is relatively common and may be asymptomatic. Diabetes is a risk factor for AF. It is more likely in those with longer duration of diabetes, higher HbA1c, and albuminuria. If untreated, AF may be distressing, but can also have life-changing consequences. It may be fatal. Treatment includes rate-controlling medication, and sometimes cardiac interventions such as cardioversion or ablative therapy. Long-term anticoagulation is essential to prevent embolic strokes. Check the pulse rate and rhythm for AF or other dysrhythmias at each diabetes review for prompt treatment, and to ensure it is appropriate to use automated blood pressure recording. People with diabetes in AF have a worse prognosis than people without diabetes.