Intended for healthcare professionals


Hunger strikes

BMJ 1997; 315 doi: (Published 04 October 1997) Cite this as: BMJ 1997;315:829

Understanding the underlying physiology will help doctors provide proper advice

  1. Michael Peel, Senior medical examinera
  1. a Medical Foundation for the Care of Victims of Torture, London NW5 3EJ

    Hunger strikes in different parts of the world are regularly in the news. Doctors with an interest in human rights may be asked to give independent medical advice to an asylum seeker intending to start a hunger strike. Several recent articles have addressed the ethics of treating hunger strikers,1 2 but there is less information available on the physiological issues. It is essential to understand both these issues to be able to advise the individual appropriately.

    There have been several studies of fasting for a few days, but in the past 15 years only three studies have described voluntary total fasting for prolonged periods. The first was of a monk who tried to fast for 40 days for religious reasons but was forced to stop on day 36 because of unacceptable symptoms.3 The second was of four adults who were planning to fast indefinitely. One became very unwell on day 38, and the others ceased fasting on day 40.4 The third was a retrospective study of 33 South African political prisoners on hunger strike for up to 28 days.5

    Hunger strikes have been around since Roman times, and the suffragettes brought the tactic to public awareness in Britain earlier this century. Gandhi fasted at least 14 times but never for more than 21 days. After the second world war Ancel Keys published an extensive review of people subjected to prolonged starvation and a study replicating these conditions in the laboratory.6 The hunger strikers of the Maze Prison in Belfast in the early 1980s died after 45–61 days, but no results have been published.

    For the first few days of starvation the body uses its stores of glycogen in liver and muscle.7 This is accompanied by glucagon induced naturesis, with substantial weight loss. The next phase lasts up to day 10-14, during which time glycogen stores are exhausted and certain amino acids take over as the substrate for gluconeogenesis. This is associated with a loss of muscle, including heart muscle. In the final phase protein is protected, so that it forms only about 10% of energy source. Most energy comes from ketones produced by the breakdown of fatty acids. When fat stores are used up there is catastrophic protein catabolism, but generally other complications arise first.

    Conclusions from studies recommend independent medical monitoring after a weight loss of 10% in lean healthy individuals.7 If the pre-hunger strike weight is unknown, a maximum of 10 days' hunger strike, or a body mass index of less than 16.5 kg/m2, should be the trigger. Major problems arise at a weight loss of about 18%.

    The main disabling symptom is feeling faint and dizzy. Hunger strikers learn to stand up very slowly and may become almost bed bound. This may affect their ability to state their case. Bradycardia and drop in blood pressure are well recognised as effects of even relatively short fasting and were seen in all the individuals studied closely.4 5 Orthostatic hypotension was present by about day 20 in all cases in which it was recorded and in at least one case was almost disabling. Weakness and lightheadedness was common. The cause of this is not clear, but could be partly due to electrolyte imbalance. Although thyroxine concentrations are maintained in fasting, tri-iodothyronine is converted rapidly to an inactive metabolite, thus reducing effective thyroid function. This is an important physiological protective function but will lead to weakness and a sensation of feeling cold. Abdominal pain was described by around three quarters of those studied, even in the early stages.5

    Dehydration is a risk in voluntary total fasting, as individuals may lose their feelings of thirst and hunger.1 This is in complete contrast to prolonged severe undernutrition, where people may drink to relieve feelings of hunger.6 Average fluid intake needs to be maintained at around 1.5 l/day. Ideally water should be supplemented with up to 1.5 g sodium chloride (half a teaspoon of salt) per day. More than this may precipitate hypokalaemia, and monitoring of potassium concentrations may be helpful where possible.

    Of the five individuals monitored closely, one developed symptomatic hypokalaemia, which eventually needed intravenous rectification. This individual went on to develop acute Werneke's encephalopathy. The risk is increased by ingested glucose; some hunger strikers eat small amounts of chocolate that are brought in by friends trying to help.

    The study of detainees showed 77% of hunger strikers to be clinically depressed at the time of admission to hospital, measured by an independent psychiatrist, although they also demonstrated features similar to those of the post-traumatic stress syndrome.5 Emotional liability is a later feature of voluntary total fasting.

    Once a hunger strike of more than three weeks is over, re-alimentation is potentially dangerous. Werneke's encephalopathy has been recorded in patients taking inappropriate food after fasting. Ingesting carbohydrate after fasting will also cause a reverse of the initial naturesis, causing measurable weight gain and potentially acute oedema. In South Africa, diluted proprietary lactose free balanced feed was used until a light diet was tolerated.5 Elsewhere, boiled vegetables have been the nutrients taken in the initial period. The patient needs to consume small amounts of food which are high in neither processed sugars nor protein. Hospital monitoring needs to be continued for several days after eating has restarted.

    Cardiac problems are potential hazards of refeeding. Hypokalaemia is a risk, and a sudden increase in fluid volume can precipitate cardiac failure as the physiological load is increased. The bradycardia and hypotension of starvation resolve and often overshoot.6 The exact cause is unclear. These fatalities may be related to loss of cardiac muscle in parallel to skeletal muscle loss. The deaths were also associated with prolonged QT intervals on electrocardiographic monitoring. Laboratory studies showed that the QT interval was one of the variables that took the longest to recover on refeeding.6

    Hunger strikers are not aware of the complex physiological processes that they are disrupting or the risks on restarting eating. Doctors working with hunger strikers must be aware of the processes and potential problems so that they can advise them fully.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    View Abstract