Metabolic diseases

This case describes refeeding syndrome associated with volitional oral nutrition in a patient with chronic alcoholic abuse admitted for detoxification. Refeeding syndrome is an under-recognised and undertreated condition1-5 of severe, acute electrolyte, fluid-balance and metabolic abnormalities in chronically malnourished patients undergoing renutrition. Refeeding syndrome was first described in Japanese prisoners during World War II.6 Since then, it has been described in patients being refed after hunger strikes, starvation after being lost, chronic alcoholism, anorexia nervosa, malignancy, kwashiorkor and marasmus and in obese patients who have had duodenal switch operations.4,5 Reports conflict over whether it is more common following parenteral7 or enteral tube4 nutrition, but descriptions following volitional oral refeeding are less frequent. In people in a chronically starved state, insulin secretion is reduced in parallel with low carbohydrate intake. Fat catabolism predominates, and free fatty acids and ketone bodies replace glucose as the major energy source. If starvation is severe, body stores of phosphate, potassium and magnesium may be depleted, although serum levels are often maintained.3-5 With refeeding, there is a shift back to carbohydrate metabolism and an increase in insulin levels. Insulin stimulates the movement of phosphate, potassium and magnesium into the cells, leading to a fall in their serum concentrations.2 In addition, tissue anabolism increases cellular demand for phosphate, glucose, potassium and water.1 Hyperphosphaturia may occur with alcoholism,7,8 and thiamine, required for the intracellular transport of glucose, may be depleted.9 The principal biochemical hallmark of refeeding syndrome, as seen in this case, is severe, acute hypophosphataemia that usually occurs within 3–4 days of refeeding.2,3,10 This is often associated with hypokalaemia, hypomagnesaemia, sodium and fluid retention, thiamine deficiency and hyperglycaemia. Phosphate is the body’s major intracellular anion.4 Daily oral phosphate intake is about 1000–1400 mg, the major sources being meat, poultry, eggs, cereals and dairy products.4 Wine contains little phosphate.11 Phosphate is found in phospholipids, nucleic acids, adenosine triphosphate and 2,3-diphosphoglycerate in red blood cells. It is important for intracellular buffering, enzymatic phosphorylation, glucose metabolism, nervous system conduction and leucocyte function. Hypophosphataemia-induced depletion of 2,3-diphosphoglycerate in erythrocytes results in a left shift of the haemoglobin/oxygen dissociation curve, increasing haemoglobin affinity for oxygen and predisposing to local tissue hypoxia.1,2,4 However, the clinical manifestations of refeeding syndrome are varied and non-specific (Box 3). Potentially life-threatening sequelae include acute cardiac failure, respiratory failure, Wernicke’s encephalopathy, sepsis and acute renal failure. Sudden cardiac death has been reported in two chronically malnourished patients experiencing acute hypophosphataemia after initiation of total parenteral nutrition.12 Although non-specific, we believe the constellation of symptoms and signs observed in our patient is typical of refeeding syndrome. Most importantly, acute, severe hypophosphataemia not present on admission was noted 4 days after oral refeeding with a ward diet. The presence of a serum phosphate level of 0.15 mmol/L in our patient represents extreme hypophosphataemia. (The lowest published level we are aware of in a patient who survived is 0.07 mmol/L.11) Given the “low normal” value on admission and the patient’s risk factors for refeeding syndrome, the serum phosphate level should have been monitored more closely during the first few days of admission. We did consider several differential diagnoses. Although hypophosphataemia is commonly seen in sepsis,14 clinical and haematological evidence suggested that the patient’s respiratory infection had largely resolved by Day 4. Acute respiratory alkalosis may also cause hypophosphataemia,15 but was unlikely in this case, in view of the normal serum phosphate level on admission. Severe hypokalaemia was already being corrected by intravenous replacement from the day of admission. The development of paraesthesiae, myalgias, groin candidiasis, diarrhoea and sinus tachycardia (which may have indicated incipient cardiac failure4,9) was consistent with the diagnosis of refeeding syndrome.4 Unfortunately, the creatinine kinase level was not measured to exclude rhabdomyolysis. Cerebellar signs may have been secondary to alcoholic degeneration or mild Wernicke’s encephalopathy. Management of refeeding syndrome includes slowing of caloric intake, correcting electrolyte and metabolic abnormalities, monitoring fluid balance and treating complications. Thiamine and B-complex vitamins should be prescribed prophylactically before refeeding.4 Interestingly, the early administration of intramuscular thiamine for chronic alcoholism may have protected our patient against Wernicke’s encephalopathy secondary to refeeding syndrome. Ideally, patients at risk of developing refeeding syndrome should be identified and a prophylactic low-caloric low-carbohydrate dietary regimen implemented.3 Initially, 85 kJ per kilogram of body weight per day, with a generous protein allowance (1.2–1.5 g protein per kilogram of body weight per day), has been suggested.4,13 Caloric intake can then be gradually increased over the following 1–2 weeks, ensuring that clinical and biochemical parameters are closely monitored.4,12,13 It is important to note that most current recommendations are based on parenteral or enteral tube nutrition. Protocols are not well developed for volitional oral refeeding. However, “slow” refeeding in these patients could be achieved by providing a similar low daily caloric intake with reduced food portions. Ideally, an experienced dietitian should be consulted.4 In our case, a dietitian was not available on site, and, given the prompt correction of electrolyte abnormalities and absence of acute cardiac failure, no change to diet was made. Levels of serum electrolytes, urea and creatinine should be monitored at least daily in the acute phase. Prophylactic phosphate and potassium supplementation is often required at the time of refeeding in high-risk patients. Phosphate replacement is recommended if serum levels are below 0.3–0.5 mmol/L3,4,10 or if the patient is symptomatic. As oral replacement at these levels is often inadequate, intravenous replacement is advised.7,10 Complications of overzealous intravenous phosphate replacement may include hyperphosphataemia, hypocalcaemia, tetany, hypotension, hyperkalaemia, hypernatraemia, renal failure and metastatic calcification.2,10 Although successful intravenous regimens based on patient weight and serum phosphate levels in intensive care settings have been described,15 these are often complicated and impractical for ward patients. Terlevich et al10 described the use of 50 mmol intravenous phosphate over 24 hours in 30 ward patients with refeeding syndrome and normal renal function. Twenty-eight patients safely achieved a serum phosphate level above 0.5 mmol/L within 72 hours. We used 42 mmol intravenous phosphate over 36 hours to normalise serum levels in our patient. While less aggressive than the protocol described by Terlevich et al, it was deemed sufficient given that the patient was largely asymptomatic and that serum phosphate levels were improving. Intravenous phosphate was dispensed on site in 14 mmol aliquots, and prescribing this amount over 12 hours simplified the dosing regimen. Clinical diagnosis of refeeding syndrome requires a high index of suspicion.1 Its hallmark of acute, severe hypophosphataemia in chronically malnourished patients after refeeding may occur even in patients who are largely asymptomatic and orally fed. Prevention of morbidity and, in some cases, death requires careful management of diet, vitamin intake and electrolyte and fluid balance. Lessons from practice Refeeding syndrome is a potentially lethal condition in chronically malnourished patients undergoing renutrition. The syndrome is under-recognised and undertreated. Electrolyte levels should initially be monitored daily in at-risk patients, as acute, profound hypophosphataemia may develop even in asymptomatic patients. Regimens for volitional oral refeeding are not well developed, but a prophylactic low-caloric (85 kJ per kilogram of body weight per day), low-carbohydrate diet has been advised. Prophylactic thiamine, phosphate and potassium supplementation is often required for at-risk patients. Patients with serum phosphate levels below 0.3–0.5 mmol/L or symptoms of hypophosphataemia require intravenous phosphate replacement. 1 Serum electrolyte levels over the first 8 days after admission Day Electrolyte Reference range 1 4 8 Potassium (mmol/L) 3.6–5.1 2.4 3.5 3.9 Calcium (mmol/L) (corrected for serum albumin) 2.25–2.58 2.27 2.60 2.60 Magnesium (mmol/L) 0.74–1.03 0.71 0.69 0.70 Phosphate (mmol/L) 0.80–1.50 0.84 0.15 1.56 2 Serum phosphate levels over the first 8 days after admission* * Dotted line indicates the direction of change only (no data were available for Days 2 and 3). 3 Clinical features of refeeding syndrome4 Clinical feature Possible mechanisms Cardiovascular Acute cardiac failure Fluid retention (secondary to carbohydrate intake1,2,4 and hyperinsulinaemia9), arrhythmias, cardiomyopathy1,2,4 Arrhythmias, sudden cardiac death3,12 Electrolyte disturbance1,2 Respiratory Respiratory failure Diaphragmatic myopathy2,8 Neurological Seizures, paraesthesiae Electrolyte and/or metabolic disturbance,1,4 cellular hypoxia secondary to reduced 2,3-DPG and ATP Wernicke’s encephalopathy Thiamine deficiency1,2,13 Gastrointestinal Diarrhoea or constipation Electrolyte and/or metabolic disturbance,4 intestinal atrophy following malnutrition13 Haematological Sepsis Leukocyte dysfunction, hyperglycaemia, acid–base disturbance1,4 Haemorrhage Thrombocytopaenia,2,9 platelet dysfunction9 Haemolytic anaemia Depletion of erythrocyte ATP, resulting in increased cell membrane rigidity1 Metabolic Hyperglycaemia4 Glucose ingestion4 Acid–base disturbance1,4 Impaired phosphate renal buffering2,12 Renal Acute tubular necrosis Rhabdomyolysis4 Musculoskeletal Myopathy Depletion of muscle ATP,1,9 electrolyte disturbance1 Rhabdomyolysis Impaired production of phospholipid cell membranes causes sarcolemma dysfunction1,2 ATP = adenosine triphosphate. DPG = diphosphoglycerate.

Adrian T Fung MB BS · Janet Rimmer MB BS, MD, FRACP

Diane A Longstreet,* Deanne L Heath,† Robert Vink‡ * Dietitian, † Research Scientist, Townsville Aboriginal and Islander Health Services, 57–59 Gorden Street, Garbutt, QLD 4814; ‡ Head, Department of Pathology, University of Adelaide, SA. dlongstreetATtaihs.net.au To the Editor: Diabetes in Indigenous Australians occurs at a younger age and at almost four times the rate in non-Indigenous Australians. The age-adjusted prevalence of diabetes among Indigenous people is 16% in remote areas and 9% in non-remote areas, with the actual prevalence estimated to be between 20% and 25%, and possibly higher than 30% in some remote areas.1 The cause for this disparity in diabetes incidence is multifactorial, and recent evidence suggests that nutrition — particularly magnesium intake — may play a role. Although central obesity remains a major risk factor, magnesium deficit has been posited to be an underlying common mechanism for the insulin resistance found in type 2 diabetes, as well as in metabolic syndrome, hypertension, and impaired glucose tolerance.2 The clinical correlations between low magnesium and diabetes have been well documented,3 with serum magnesium deficits being reported in 25%–39% of diabetic outpatients in the United States and Switzerland, and up to 73% of diabetic outpatients in Mexico. With magnesium deficits being observed in diabetes, studies examining the effects of magnesium-rich foods on diabetes risk become relevant. The Nurses’ Health Study and the Health Professionals’ Follow-up Study, which included 85 060 women (18 years follow-up) and 42 872 men (12 years follow-up), demonstrated that, after adjusting for confounding variables, a magnesium-rich diet reduced the relative risk of developing diabetes by 34% in women and 33% in men.4 A similar inverse correlation between magnesium intake and diabetes risk was shown in the Iowa Women’s Health Study with a cohort of 35 988 older women,5 and in the Honolulu Heart Program and the Women’s Health Study with cohorts of 8006 men and 39 345 women, respectively.6,7 Despite this growing body of evidence supporting the involvement of magnesium in diabetes, consideration of magnesium status has not been integrated into Australian medical care for diabetes, and more specifically, for Indigenous Australians. It is known that the traditional diet of hunter-gathers such as Indigenous Australians was much more nutrient- and magnesium-rich than the current estimated Australian intake.8 Nonetheless, there remains a lack of information about current magnesium status, including dietary intake, in Indigenous Australians. It is possible that dietary magnesium intake may be too low to maintain normal serum magnesium homoeostasis, and that this might contribute to the development of type 2 diabetes. Further research into this issue may provide this information.

Diane A Longstreet · Deanne L Heath · Robert Vink

Canteens are not the main source of food for Australian school kids, but their symbolism is big There is widespread awareness of the obesity epidemic in Australian children,1 and the focus has now, quite appropriately, turned to action. In the United Kingdom, celebrity chef Jamie Oliver is trying to transform a 100-year-old school lunch service from “soggy and fried” to “crisp and fresh”. In Australia, the question is whether school canteens should be a high priority for action, because of their accessibility and visibility, or a low priority, on the grounds that canteen foods contribute little to children’s energy intake. Negative ripples from canteensOver the period of a year, children aged 5–15 years obtain only about 16% of their total energy intake from food eaten at school, and probably less than 3% comes from canteens.2 But while the energy contribution is small, the symbolism is big. Canteen users consume significantly greater amounts of foods likely to promote unhealthy weight gain, such as fast foods, confectionery and packaged snacks.2 The types of foods and beverages that predominate in school canteens not only undermine the health and nutrition curriculum, but also create the impression that foods and drinks that are high in fat, sugar and salt belong on the plate as “everyday foods”, rather than on the side as “occasional foods”. Other common practices in schools that undermine healthy eating messages include rewarding children with sweets, having soft-drink and confectionery vending machines, holding sporting events with fast-food vouchers as prizes, and using chocolate drives for fundraising. All these practices create negative ripple effects on Australian family eating practices and beliefs.3 Children are developing the food preferences that they will carry with them into adulthood, so strengthening family and school environments for enjoying healthier food choices is critical. In a 2004 survey of 18 Victorian primary schools (unpublished data), we found that, of the 17 with a food service, all sold meat pies, but only five sold fruit on a regular basis. As a rule, canteen managers provided foods that sold well and had a long shelf life. They usually had no mandate or support to do otherwise. A reliance on profits from canteens, vending machines and “junk food fundraising” also makes it hard for schools, particularly high schools, to model healthy eating. In common with a survey of 500 New Zealand schools,4 we found that schools readily recognise the rather poor job they do of providing a healthy food environment. Most schools do not see food provision as part of their core business and lack the inclination or resources to take on this “added” responsibility. Private enterprise fills this vacuum, with the result that the health of profits increasingly dominates the health of pupils. An extreme example is the “cola war” in the United States, in which the weapons of choice have been contracts with schools to sell minimum volumes of Coca Cola or Pepsi.5 Creating a new epidemic?Perhaps we should consider the task ahead of us as the creation of a new epidemic of healthy eating rather than reducing an obesity epidemic. Using the principles in Gladwell’s recent bestseller The tipping point,6 the school canteen and students themselves could be the catalyst for healthier eating among children and adolescents — turning negative ripples into positive waves. Can a “tipping point” be created from a handful of champion schools that decide to embrace the whole-of-school policies and strategies needed to get their canteens right (healthy, enjoyable, profitable and supported), hoping that others will follow their lead? This is almost certainly too much to expect to happen in 9000 schools across Australia, which tend to function semi-autonomously on these matters. Lessons from successful public health programs, such as sun protection and injury prevention, show that tipping the balance in targeted behaviours from unhealthy to healthy requires a backbone of strong central policy, ongoing social marketing, and supported and coordinated implementation of programs. Some Australian modelsVarious government-supported models influence how school canteens operate in Australia. Probably the least effective include the Victorian model of simply disseminating canteen guidelines7 and the Australian Government model from the pre-election spending spree, wherein each school could apply for $1500 to reinvent the “healthy canteen” wheel. Neither has policy, social marketing or implementation support. A third model, which has some merit, is exemplified by the Western Australian (StarCAP8) and Tasmanian (Cool CAP9) school canteen accreditation programs. Both have well developed criteria and processes for schools to work through to achieve program accreditation. StarCAP is backed by the WA government, but is managed on a shoestring budget, without policy and social marketing support, and thus has a low accreditation rate (7% of schools) and declining reach.10 Cool CAP is newer, with a higher accreditation rate (42% of schools accredited or working towards it), and so far has been successful in securing legislative and monetary support. Ultimately, however, the impact of these types of programs will probably be modest as long as the impetus to change remains with each individual school. Because they are well supported and centrally driven, the most promising models come from New South Wales and South Australia. The NSW Healthy School Canteen Strategy (“Fresh Tastes @ School”)11 grew out of the NSW Government Childhood Obesity Summit in 2002. It is now mandatory for state schools to provide food and beverage choices consistent with the Australian guide to healthy eating.12 NSW Health has also boosted support for the NSW Canteen Association so that it, in turn, can support schools to operate economically viable, nutrition-oriented school canteens. Early positive waves include support from parents, canteen managers, some food companies and, increasingly, local health and education services. A similar model released in 2004 in South Australia brings SA government backing to a set of healthy eating guidelines.13 The guidelines encourage links between the canteen, the community and teaching about nutrition food skills. Both the NSW and SA government models would now benefit from social marketing explaining the rationale, processes and support for the program. It would make sense, for example, to link these strategies with the national “Go for 2&5” (2 serves of fruit and 5 serves of vegetables) campaign.14 If we are serious about the childhood obesity epidemic, school canteens are a good place to start, because they carry a symbolism that ripples into the Australian diet far beyond their contribution to energy intake. At the moment, the ripples are a negative and undermining force. However, full implementation of the NSW or SA models for school canteens throughout the country could just tip the balance towards an outbreak of healthier eating.

A Colin Bell BSc(Hons), MSc, PhD · Boyd A Swinburn MB ChB, MD, FRACP

Clinical record A 22-year-old woman who was 12 weeks pregnant presented to a country hospital with uncontrollable nausea and vomiting. This was initially controlled with antiemetics, but over the next 4 weeks, she continued to vomit five to six times a day. She presented again at 16 weeks’ gestation when the vomiting became more frequent, and also complained of poor visual acuity and diplopia of several days duration. At the time, there was no gait disturbance, no change in mentation, nor any associated headache. At this stage she was admitted and treated with intravenous fluids including glucose. One week into her admission, her vision deteriorated and she became confused and unsteady on her feet; she was referred to the Royal Melbourne Hospital. On initial examination in Melbourne, she was normotensive (120/80 mmHg) but drowsy. There was no neck stiffness or rash. She now had a prominent gait ataxia and a corresponding mild lower-limb peripheral neuropathy with absent deep tendon reflexes. Additionally, she had bilateral horizontal nystagmus and poor visual acuity (right, 6/36; left 6/18). There was an associated mild right gaze paresis with intact visual fields on confrontation. There was no pain with eye movements. The results of fundoscopy are shown (Figure 1). Her past history was unremarkable. She did not take regular medications, and this was her first pregnancy. Ultrasonography of the fetus and placenta at 17 weeks’ gestation excluded multiple and molar pregnancy. She did not drink or smoke. Haematological and biochemical test results were largely normal, but her serum levels of creatinine (0.04 mmol/L; normal range [NR], 0.05–0.10 mmol/L) and urea (2.2 mmol/L; NR, 2.5–6.1 mmol/L) were slightly low, and her INR (international normalised ratio) was 2.0 (NR, 0.8–1.2). There were also some elevated levels in her liver function test results: γ-glutamyl transferase, 60 U/L (NR, < 50 U/L); alanine aminotransferase, 486 U/L (NR, < 55 U/L); and bilirubin, 32 U/L (NR, 0–19 U/L). Cerebrospinal fluid (CSF) examination and magnetic resonance imaging (MRI) were also undertaken to exclude intracranial causes. The CSF findings were normal, but the MRI revealed bilateral thalamic lesions (Figure 2). Given the new clinical features and her MRI result, a diagnosis of Wernicke’s encephalopathy was made. Thiamine deficiency was suspected to be secondary to hyperemesis gravidarum. Interestingly, her serum thiamine level was only marginally low at 64 nmol/L (normal range, 66–200 nmol/L). Therapy with pyridoxine, folic acid and intravenous thiamine replacement was commenced. She made a rapid clinical recovery. Visual acuity returned to normal within 3 days and her ataxia, gaze limitation and neuropathy improved. Repeat MRI imaging 7 days later showed moderate resolution of the hyperintensities. Our patient went on to successfully deliver a healthy baby girl at 37 weeks’ gestation. 1 Fundal photographs showing gross papillitis of both the right and left optic discs, respectively. A small retinal haemorrhage is shown in the left eye. 2 Magnetic resonance imaging (T1-weighted, T2-weighted axial, and diffusion-weighted image, respectively), with arrows showing bilateral symmetrical thalamic hyperintensities with restricted diffusion. Bilateral thalamic ischaemia is most often seen in thiamine deficiency. This patient’s rapid response to intravenous thiamine replacement confirms the diagnosis of Wernicke’s encephalopathy secondary to thiamine deficiency. Thiamine, or vitamin B1, plays an important role in carbohydrate metabolism through decarboxylation of alpha-keto acids. Moreover, it also functions as a co-enzyme to the apoenzyme transkelotase in the pentose monophosphate pathway for glucose.1 Low thiamine levels can be associated with poor intake or absorption, increased demand, or poor utilisation. It is well understood that thiamine requirements are increased in pregnancy, and this is thought to be the result of sequestration of the vitamin by the fetus and placenta.2 With hyperemesis gravidarum, the deficiency is further compounded by the impaired absorption related to the intractable vomiting.3 In some situations, prolonged use of intravenous glucose without thiamine can be a potent precipitant of Wernicke’s encephalopathy.4 This phenomenon is well documented,5 and in our patient, might have contributed to the evolving deficiency. Abnormal liver function test results are also seen in Wernicke’s encephalopathy, and it is possible that hepatic derangement may have also played a pathogenic role in the evolution of our patient’s disease. 6 Our case is unusual in that visual deterioration was the first symptom of an evolving thiamine deficiency. There was a delayed presentation of the classical triad of Wernicke’s encephalopathy — ataxia, mental confusion and ocular gaze problems. Moreover, in our case, MRI scanning was able to detect sensitive neurological changes, raising the suspicion of thiamine deficiency. This information was important in prompt treatment of the condition when clinical signs began to emerge. Indeed there are reports of the usefulness of MRI imaging in diagnosing cases of Wernicke’s encephalopathy.7 Our patient’s vitamin B1 level was only mildly reduced, but this may have been the result of partial correction with food intake while an inpatient at the country hospital. Unfortunately, intracellular thiamine levels could not be measured before replacement. Intracellular thiamine level is usually a more accurate marker of thiamine deficiency, measured as erythrocyte thiamine diphosphate by high pressure liquid chromatography, and is often abnormal in such cases of Wernicke’s encephalopathy, even when the serum thiamine level is normal. If left untreated, thiamine deficiency can lead to severe neurological, cardiovascular and gastrointestinal sequelae. If not treated, the result is often death. Delayed correction of low thiamine levels may lead to persistent nystagmus, ataxia and poor concentration with multiple tasks.8 For the fetus of an affected mother, slow in-utero development is possible. Ophthalmic symptoms may include subacute, progressive visual loss, with central field defects and associated poor colour vision. Papillitis can be seen in cases of nutritional optic neuropathy. In severe cases, blindness may occur as a result of complete optic atrophy.9 As thiamine deficiency has severe consequences, we believe that vitamin B1 replacement should be considered in all pregnant women with hyperemesis gravidarum. We further suggest that in such cases MRI imaging may play an important role in early diagnosis. Lessons from practice Visual deterioration may sometimes be the first presenting sign of thiamine deficiency. The visual symptoms of Wernicke’s encephalopathy are reversible with prompt administration of thiamine. Early correction of thiamine deficiency is essential in all patients with hyperemesis gravidarum. Intravenous glucose should be given together with thiamine for patients being treated for hyperemesis gravidarum. The best laboratory measure of thiamine deficiency is erythrocyte thiamine diphosphate by high pressure liquid chromatography. Magnetic resonance imaging may help in early diagnosis of patients with Wernicke’s encephalopathy where optic neuropathy is the first clinical sign.

Mervyn D Ferdinands MB BS(Hons) · Janaka Seneviratne FRACP · Owen White MD, PhD, FRACP

Timothy C Crowe,* David Cameron-Smith† * Lecturer, † Senior Lecturer, School of Exercise and Nutrition Sciences, Deakin University, Burwood, VIC 3125. tcroweATdeakin.edu.au To the Editor: Low-carbohydrate diets have re-emerged into the public spotlight and are enjoying widespread popularity. However, current evidence indicates that low-carbohydrate diets have no significant advantage over more traditional energy-restricted diets for long-term weight loss and maintenance.1-3 While these diets have shown short-term efficacy in modifying some lipid parameters and measures of insulin sensitivity, questions remain about the risk of adverse effects with long-term carbohydrate restriction.4 The scientific literature has not addressed the questions of how the general public perceive these diets, and what dieting approaches they adopt. Dieting perceptions and practices within the community may be far removed from the strictly controlled situation of published research. A national telephone survey of 1200 adults aged 18 years and over was conducted by the private market research company Newspoll from 6 to 8 August 2004. The survey asked about knowledge of and attitudes to carbohydrates and dieting. Telephone numbers were randomly selected, with a quota for capital city and non-capital city areas. Selection of an individual in each household was based on the last birthday. Response rate to the survey was 11%. Sex, age, marital status and working status demographics were representative of the Australian adult population. The main findings are summarised in the Box. Most of those surveyed correctly identified foods such as pasta and bread as “carbohydrate foods”. Only a third of people identified soft drinks and lollies as carbohydrate foods, and 20% incorrectly identified cheese and eggs as carbohydrate foods. Almost 17% of people had either tried, or intended to try, a low-carbohydrate diet, with women more likely to have tried this diet. Half of those surveyed believed that carbohydrate foods should make up a quarter or less of the daily diet (current health recommendations are that about half the diet should comprise carbohydrates). Almost 70% of those surveyed believed they needed to cut back on carbohydrates to lose weight. Based on this survey, low-carbohydrate dieting practices are as widespread in Australia as in the United States.5 Interestingly, the US study noted a greater propensity to use carbohydrate-reduced diets among those who were obese, had diabetes, hypertension or high cholesterol. Our data demonstrated widespread misunderstanding of what constitutes a high-carbohydrate food, which may leave many individuals at risk of choosing a diet that selectively excludes wholegrain foods, fruits and some dairy products. Health professionals should be aware that low-carbohydrate diets remain popular, that the people who are following these diets may represent a more “at risk” population, and that the food choices made by those following this dietary pattern may have adverse long-term health effects. Respondents’ knowledge of and attitudes to carbohydrate foods and low-carbohydrate dieting Question Total (n = 1200) Men (n = 600) Women (n = 600) Which of the following foods, if any, do you regard as carbohydrate foods? Bread 89% 85% 93% Pasta 90% 88% 93% Rice 80% 76% 83% Breakfast cereal 79% 77% 80% Lollies 35% 36% 35% Soft drink 34% 31% 36% Cheese 20% 22% 19% Eggs 18% 20% 16% Have you tried or do you intend to try the Atkins diet, or some other low-carbohydrate diet? 17% 11% 22% Based on official recommended guidelines for a healthy diet, about how much of a person’s diet should be made up of foods such as bread, breakfast cereal, pasta and rice? Less than a quarter 7% 7% 7% About a quarter 43% 34% 51% About half 29% 31% 26% About three quarters 6% 8% 4%

Timothy C Crowe · David Cameron-Smith

Marc M Cohen Professor of Complementary Medicine, School of Health Sciences, RMIT University, PO Box 71, Bundoora, VIC 3083; and President, Australasian Integrative Medicine Association. marc.cohenATrmit.edu.au To the Editor: I read Constable’s article1 on age-related macular degeneration with interest, but was surprised at its somewhat guarded advice on nutritional supplementation and the fact that it gave only passing reference to uncontrolled studies of the carotenoids lutein and zeaxanthin and failed to mention a number of controlled studies that have recently shed light on the potential for these nutrients to influence the progression of age-related macular degeneration (ARMD). Lutein and its isomer, zeaxanthin, are deposited in the macula, where they make up the macular pigments and act as a blue-light filter to protect the underlying tissues from phototoxic damage, as well as providing antioxidant activity. Lutein was not available in supplement form at the time of conducting the Age-Related Eye Disease Study (AREDS).2 However, the more recent Lutein Antioxidant Supplementation Trial (LAST), a double-masked, placebo-controlled, randomised trial of lutein and antioxidant supplementation in people with ARMD, demonstrated that taking 10 mg lutein daily, with or without additional nutrients, improved visual function.3 Interestingly, lutein supplementation was also found to improve vision in a small, randomised, placebo-controlled study of people with cataracts. It was suggested that these improvements were due to improved macular function and increased macular pigment density.4 Furthermore, a recent study showed that lutein supplementation results in increased macular pigment density in both normal and ARMD patients.5 Lutein occurs naturally in foods such as eggs, spinach, romaine (cos) lettuce, broccoli, zucchini, corn, peas and Brussels sprouts. Although lutein is readily absorbed from foods and dietary supplements, surveys indicate that average lutein intake may be below levels that are associated with disease prevention.6 Toxicology studies have established that lutein is generally safe, with potential for use as a supplement in foods and beverages.6 While advice on smoking cessation and increasing fruit and vegetable intake is useful for a wide range of conditions, including ARMD, Constable’s statement that “antioxidant supplements should be recommended if a fresh diet is impractical and if retinal signs of progression are present”1 appears overly cautious. In light of recent findings on the potential benefits of antioxidants such as lutein, and the low cost and minimal risks associated with supplementation compared with the potentially devastating consequences of blindness from ARMD, it may be prudent to make more general recommendations on nutritional supplements, rather than waiting until signs of retinal progression are evident.

Marc M Cohen

Ian J Constable Director, Lions Eye Institute, Centre for Ophthalmology and Visual Science, University of Western Australia, 2 Verdun Street, Nedlands, WA 6009. ijcATcyllene.uwa.edu.au In reply: While the published literature on dietary supplementation with the antioxidants lutein and zeaxanthin is highly encouraging, it does not yet pass the requisite standards for public endorsement provided by large-scale, independent, evidence-based medical trials. The controlled (“LAST”) trial of lutein1 cited by Cohen consisted of just 91 patients divided into three subgroups including the placebo group, who were followed for only 1 year. The measurement of visual improvement on a Snellen chart would not be accepted as “gold standard” evidence by major granting agencies and regulatory affairs bodies, who demand the higher discrimination of a logarithmic visual acuity chart. Moreover, the data were not derived from multicentre trials and independently assessed. It is instructive to compare the methodology of the LAST trial with that of the ARED Study,2 which involved 11 centres, 3640 patients and an average follow-up of 6.3 years. For these reasons, although I mentioned lutein in an encouraging fashion, I did not endorse it to the same extent as vitamin C, vitamin E and zinc supplements. The other reasons for giving limited recommendation of antioxidant supplements at this stage relate to the fact that it is not yet clear to what extent supplements would be beneficial over and above a diet targeted to provide these antioxidants in plentiful supply. It may yet be shown that a nutritious diet — with emphasis on brightly coloured and leafy vegetables, fresh fruits, nuts and fish, coupled with reduced processed vegetable oils (except olive oil) — can, alone, provide substantial protection. While the supplements are generally deemed to be safe, they are not without occasional serious side-effects and have not been followed long term. Richer, one of the authors of the lutein study, 1 acknowledges commercial relationships with the supplement suppliers, and states that the study requires greater numbers and long-term follow-up to be confirmed. Cohen cited a second article3 that refers to a 2-year study of cataract, in which a mere 17 patients were allocated to three subgroups including the placebo group. It is not usual practice to quote conclusions from such a small, and therefore potentially unreliable, trial design. Cohen is right to point to lutein and possibly zeaxanthin supplements as an encouraging possibility for preventing blindness from macular degeneration, and I hope the evident enthusiasm and rapid marketing of lutein proves justified in the long run by forthcoming major trials.

Ian J Constable

Weight loss occurs in the short term, but not enough is known to recommend long term use After health professionals have promoted a low fat, high carbohydrate model of eating for more than 20 years, the prevalence of overweight and obesity in Australia (as elsewhere) has climbed.1 Very few people are able to attain and maintain a truly low fat eating plan, but that has not stopped the low fat orthodoxy being blamed for the obesity epidemic.2 Yet, it is not sufficient to focus on a single aspect of diet — low fat diets are not intrinsically “healthy”, especially if they contain high levels of simple sugars, low levels of complex carbohydrates and are nutrient poor. In contrast to the low fat, high carbohydrate diet, a popular approach to weight loss is the Atkins diet,3 a “controlled carbohydrate” dietary regimen. One of the many reasons for its popularity is that, as society has become increasingly concerned about body image and weight, the Atkins regimen promises quick weight loss without hunger, allows a wide range of foods and has simple “rules”. All this is supported by consumer “how-to” books, celebrity endorsement, and food product innovation and marketing. The Atkins diet — “kerbing the carbs”Atkins’ theory rests on a belief that a high intake of refined carbohydrate, especially simple sugars, causes overstimulation of insulin and results in uncontrolled hunger and eating, while the excess insulin also favours fat storage. Thus, the Atkins diet relies primarily on controlling carbohydrate intake and progresses through four phases. The strict induction phase, intended to produce ketosis, allows only 20 g of carbohydrate a day for a minimum of 2 weeks. Fruit, bread, grains, starchy vegetables or dairy products other than cheese, cream or butter are eliminated. Sugar and alcohol are not allowed, and caffeine is discouraged. Mineral and vitamin supplementation, dietary fibre, and eight glasses of water a day are recommended. During this restrictive phase, weight loss is rapid. While this initial weight loss may be due in part to water loss as body glycogen stores are depleted, low carbohydrate diets also result in a reduced caloric intake.4-7 Factors contributing to the lower caloric intake may be the satiating effect of a high protein diet, a lower absolute fat intake due to restricted food choice, and possibly appetite suppression due to ketosis.8-10 However, the exact mechanisms of the weight loss are as yet unknown.10 The second and third phases of the diet allow a gradual liberalisation of food intake by an incremental increase in total carbohydrate: fruits, nuts, more vegetables and some cereal foods are added. The final, maintenance phase is intended to be permanent, and aims to keep daily dietary carbohydrate intake to a known (relatively low) amount. Advice for patients wanting to follow the Atkins diet While low carbohydrate diets appear to work for weight loss in the short term (6 months), not enough is known to recommend them in the long term. All weight-reduction diets are difficult to follow over a long period of time and have limited long term success. Follow the complete Atkins plan (not only parts of it), including regular physical activity, vitamin and mineral supplementation, a daily fibre supplement, eight glasses of water a day, and minimally processed foods. Maintain a high daily intake of fruit and vegetables (at least two serves of fruit and five serves of vegetables from the “allowed” foods) and avoid saturated fats. A dietitian can help with your dietary intake plan if you are having difficulties. Does the Atkins diet work?If weight loss is the goal, the answer appears to be a qualified “yes”. For obese people, it works a little better than a low fat diet over 6 months. Recently, four randomised controlled trials in obese men and women (two lasting 6 months, two lasting 12 months) compared a low carbohydrate diet to a conventional low fat weight-loss diet.5-7,11 Although the studies differed in design and had different subjects, in each study the weight loss at 6 months was 4–6 kg greater for the low carbohydrate group than for the low fat group. However, the weight loss difference between groups at 12 months was no longer statistically significant.6,11 The dropout rates in all of the trials were high (21%–43%), with a general non-significant tendency for better retention in the low carbohydrate group. So, in the long term, low carbohydrate diets do not necessarily offer better weight control than lower fat, higher carbohydrate diets. Is the Atkins diet safe?During weight loss, a low carbohydrate regimen appears to have no adverse effects on cardiovascular risk factors such as serum lipid levels (total and low-density lipoprotein cholesterol) or blood pressure, or on fasting glucose and fasting insulin levels.6,7 In fact, randomised controlled trials comparing a low carbohydrate diet with a low fat diet up to 12 months consistently indicate a beneficial effect on serum triglyceride and high-density lipoprotein cholesterol concentrations. However, the low carbohydrate regimen is associated with a greater incidence of constipation, headache, halitosis, muscle cramps, diarrhoea, general weakness and rash.5 Strictly limiting carbohydrates could also reduce intake of plant-based foods rich in phytochemicals, bioflavinoids, carotenoids and other micronutrients now regarded as important in a healthy diet.12 The regimen developed by Atkins3 encourages fruit and vegetable intake, and minimally processed food, so a low carbohydrate diet should not necessarily imply an intake low in fibre and low in plant-based food. Low carbohydrate diets may also be beneficial by removing simple sugars and sugary foods, including fructose sweeteners, which could be responsible for excess energy intake.8 Overall, however, the safety of low carbohydrate diets beyond 12 months is largely unknown, and there is speculation that the regimen may have adverse health implications for cardiovascular disease, renal function (through an observed cross-sectional association of high dietary protein intake with proteinuria) and bone health (through relatively low calcium intake and the association of high protein intake with hypercalcinuria).13 Information is also lacking on the long term effect of a low carbohydrate regimen for the young, the elderly, people of normal weight (or for people who are not losing weight), and those with chronic conditions, such as diabetes or cardiovascular disease. As with many dietary regimens, the nutritional quality of low carbohydrate diets varies according to how the dietary rules are applied. The Atkins diet calls for a drastic dietary reduction of foods with a significant starch and sugar content — in doing so, the intake of many energy-dense but micronutrient-poor foods is reduced. There is the potential for these to be replaced with foods that are moderate in energy intake, and rich in fibre and micronutrients. However, in any regimen to reduce or control weight, particular attention should be given to ensuring that the reduced food intake is of high nutritional quality. A sensible way to follow an Atkins diet is to include plenty of the allowed fruits and vegetables, and to prefer food sources of unsaturated fat over those with saturated fat.

Malcolm D Riley PhD · John Coveney PhD

Recent data from Australia, the United States and Europe show increased self-reported energy intake associated with obesity, in contrast to earlier suggestions that the obesity epidemic has occurred despite minimal or no increase in per capita energy intake from food. The effect of increased energy intake is compounded by sedentary lifestyles. Both physical activity and nutrition must be addressed to reduce the prevalence of obesity and improve the health of Australians. During the past decade, several authors from the United States and the United Kingdom have suggested that the obesity epidemic has occurred despite minimal or no increase in per capita energy intake and/or energy from the food supply.1,2 This has been described as the “American paradox”,2 and cited as evidence that the obesity epidemic is due to decreased physical activity and not to changes in eating patterns.3 There are dangers inherent in this viewpoint. According to the International Obesity Taskforce (IOTF), the food industry seeks to focus on inactivity and promote sports to divert attention from the role of foods and drinks.4,5 The IOTF asserts that the causes of the obesity epidemic are twofold: an abundance of energy-dense foods and drinks, leading to a pervasive “passive over-consumption” of energy; and an environment that limits opportunities for physical activity, leading to an almost universal sedentary state.4 Energy intakeTrends in energy consumption are difficult to establish because of a variety of measurement issues. People under-report their energy intake, and higher BMI is associated with a greater degree of under-reporting.6 This introduces a systematic bias, resulting in the paradoxical observation that obese individuals appear to eat less than lean people. Techniques have been developed to reduce this bias by excluding the most blatant under-reporters. The Australian Food and Nutrition Monitoring Unit undertook a bridging study to compare the food and nutrient intake data from the 1983 and 1995 Australian national nutrition surveys and the 1985 Australian Council for Health, Physical Education and Recreation (ACHPER) study.7 For these studies, energy intake was estimated from 24-hour intake and food frequency data, collected by dietitians at in-home interviews. After adjusting for differences in the food composition databases and demographic changes in the Australian population, the study found that mean energy consumption of Australian adults living in capital cities increased significantly by around 3%–4% (about 350 kJ/day) between 1983 and 1995.7 Between 1985 and 1995, mean energy intake increased greatly and significantly by 11% for girls and 15% for boys aged 10–15 years. The major source of increased energy intake was carbohydrate. Absolute fat intake did not increase among children, and declined slightly among adults.7 Fat as a percentage of energy decreased because of the increase in energy intake. An energy imbalance of 3%–4% in adults (about 350 kJ/day, equivalent to a slice of bread, or 30 minutes of sitting instead of brisk walking) would produce weight gain of about 1 kg per year until equilibrium is again reached, when the higher energy expenditure at a higher body mass equals energy intake. The Australian intake data are supported by recent data from the US. A significant increase in total energy intake between 1971 and 2000, particularly from carbohydrate, has been reported in the US, based on intake data from the National Health and Nutrition Examination Survey (NHANES).8 Data on food supply in the US and Europe also indicate that energy supply has increased.9,10 In Europe, data from the MONICA study indicate that per capita energy supply correlates with prevalence of obesity.10 In that study, trends in total energy supply per capita explained 41% of the between-population trends in BMI. Energy supply combined with prevalence of ex-smokers (who are more likely to be overweight) explained 69% of the between-country differences in change of prevalence of overweight. An economic analysis of factors underlying the trends in obesity in the US found that the per capita increase in the numbers of restaurants accounted for 61% of the increase in BMI and 65% of the increase in the percentage of the population who are obese.11 Energy expenditureAlthough it is not possible to measure energy expenditure at a population level, it has been suggested that population Total Daily Energy Expenditure (see Box) has declined.1,2,12 However, all of these reviews rely on the argument, based on the UK data,1 that energy intake has not increased and therefore energy expenditure must have decreased. Objective analysis of Total Daily Energy Expenditure and Activity Energy Expenditure in free-living individuals is possible using doubly-labelled water techniques. A large review of objective studies of energy expenditure found that Resting Energy Expenditure, Activity Energy Expenditure and Total Daily Energy Expenditure are all substantially and progressively higher with obesity. For BMI categories ranging from < 25 to > 35 kg/m2, Total Daily Energy Expenditure ranges from 9.5 to 13.5 MJ/day for women and from 12.9 to 17.5 MJ/day for men.13 This contradicts the popular view that obesity is due to “low metabolism” and is maintained despite a low level of food intake. After adjustment for body size, energy expenditure of obese individuals is similar to that of lean individuals. Population body mass has increased, so population per capita Total Daily Energy Expenditure must have increased. Clearly, energy intake must also have increased to maintain energy balance. The recent studies quoted above, based on both food supply and self-reported intake data in the US, Australia and Europe, confirm this. Physical activityAt an individual level, physical activity is clearly important for weight control. Physical activity is an effective adjunct to dietary management for weight loss and maintenance.14 However, at a population level, measurement of physical activity relies on self-report and suffers from methodological difficulties similar to those for determining nutrition. Prospective observational population studies of the effect on body weight of physical activity measured at baseline are few and the results are inconsistent.15-17 Physical inactivity appears to be both a cause and a consequence of obesity. Many studies have shown that television viewing is associated with obesity, although this is independent of physical activity level.18 However, it is very clear that a sedentary lifestyle is almost universal in developed societies, and that this is related to physical, technological and economic environmental conditions.19 In Australia, population surveys suggest that the proportion of the population reporting undertaking 30 minutes of moderate physical activity daily has declined from 62% in 1997 to 57% in 2000.20 However, there is now consensus that 45–60 minutes of moderate physical activity may be required for some people to prevent weight gain in the current environment of abundant energy-dense food.15,17 Relative contributions to obesityTwo longitudinal studies using objective measures of energy expenditure have attempted to determine the relative contribution of energy intake and expenditure to subsequent risk of obesity. In the first study, involving infants, measured energy intake determined weight gain between 3 months and 12 months, but measured energy expenditure did not.21 The other study showed that, among Pima Indian adults, baseline total energy intake (calculated from energy expenditure measured by using doubly-labelled water) is a predictor of weight gain over 4 months or more.22 However, baseline Activity Energy Expenditure and physical activity were not related to weight gain. As in many other studies, the physical activity level was lower in obese subjects, and the authors concluded that this was secondary to obesity. ConclusionAn increase in energy supply and consumption has made a major contribution to the obesity epidemic. It is probable that population physical activity level has also decreased, as both a cause and a consequence of the obesity epidemic. There are many other benefits of physical activity and healthy eating besides weight management, and both physical activity and nutrition must be addressed together to improve the health of all Australians. This requires multistrategy interventions across a range of sectors, including food producers, manufacturers, wholesalers, retailers, restaurateurs, caterers, transporters, advertisers, urban planners, employers, sporting associations, the fitness industry, community groups, the media, and policymakers at all levels of government, in addition to interventions in schools. A national, coordinated, systematic approach to monitoring overweight and obesity, dietary intake and physical activity is essential to both inform and evaluate interventions. Key evidence-based interventions to address childhood obesity through improved nutrition and increased physical activity are outlined in the National Obesity Taskforce report Healthy weight 2008 — Australia’s future.23 Energy expenditure terminology Total Daily Energy Expenditure is the amount of energy expended by an individual in one day. This is made up of Resting Energy Expenditure and Activity Energy Expenditure. Resting Energy Expenditure is the energy expended by the body in the resting state to maintain itself and digest food. This is made up of Basal Metabolic Rate (energy expended in the rested and fasted state) plus the Thermic Effect of Food (the energy used in digesting food — estimated at ~10% of Total Energy Expenditure). Activity Energy Expenditure is the energy expended by the body in movement, and is the only aspect of energy expenditure that is under conscious control. Activity Energy Expenditure makes up 20%–40% of Total Energy Expenditure, depending on activity level. Physical Activity Level is expressed as the ratio of Total Daily Energy Expenditure divided by Basal Metabolic Rate.

Christina O Stubbs BSc, GradDipDiet, APD · Amanda J Lee BSc, GradDipDiet, PhD

The Victorian Supreme Court has decided that artificial nutrition and hydration provided through a percutaneous gastrostomy tube to a woman in a persistent vegetative state may be withdrawn. The judge ruled, in line with a substantial body of international medical, ethical and legal opinion, that any form of artificial nutrition and hydration is a medical procedure, not part of palliative care, and that it is a procedure to sustain life, not to manage the dying process. Thus, the law does not impose a rigid obligation to administer artificial nutrition or hydration to people who are dying, without due regard to their clinical condition. The definition of key terms such as “medical treatment”, “palliative care”, and “reasonable provision of food and water” in this case will serve as guidance for end-of-life decisions in other states and territories. The case also reiterates the right of patients, and, when incompetent, their validly appointed agents or guardians, to refuse medical treatment. Where an incompetent patient has not executed a binding advance directive and no agent or guardian has been appointed, physicians, in consultation with the family, may decide to withdraw medical treatment, including artificial nutrition or hydration, on the basis that continuation of treatment is inappropriate and not in the patient’s best interests. However, Victoria and other jurisdictions would benefit from clarification of this area of the law.

Michael A Ashby MD, FRACP · Danuta Mendelson MA, LLM, PhD

Objectives: To review the evidence for the effectiveness of complementary and self-help treatments for anxiety disorders.Data sources: Systematic literature search using PubMed, PsycLit, and the Cochrane Library.Data synthesis: 108 treatments were identified and grouped under the categories of medicines and homoeopathic remedies, physical treatments, lifestyle, and dietary changes. We give a description of the 34 treatments (for which evidence was found in the literature searched), the rationale behind the treatments, a review of studies on effectiveness, and the level of evidence for the effectiveness studies.Conclusions: The treatments with the best evidence of effectiveness are kava (for generalised anxiety), exercise (for generalised anxiety), relaxation training (for generalised anxiety, panic disorder, dental phobia and test anxiety) and bibliotherapy (for specific phobias). There is more limited evidence to support the effectiveness of acupuncture, music, autogenic training and meditation for generalised anxiety; for inositol in the treatment of panic disorder and obsessive-compulsive disorder; and for alcohol avoidance by people with alcohol-use disorders to reduce a range of anxiety disorders.

Anthony F Jorm PhD, DSc · Helen Christensen PhD · Kathleen M Griffiths PhD · Ruth A Parslow PhD · Bryan Rodgers PhD · Kelly A Blewitt BAppPsych

“Patient is embarking on one-month McDonald’s binge”, writes pragmatic GP Daryl Isaacs in his medical record for a fighting-fit Morgan Spurlock. The young American filmmaker has decided that he is going to eat nothing but McDonald’s food, for three meals of every day for a whole month. He will try everything on the menu, and if asked to “super size”, he will accept the challenge. It seems this intriguing n-of-1 trial, documented on film in the tragi-comic Super size me, was the brainchild of a creative man. Or maybe it was a “gut response” to the judgment handed down in failed US legal action against fast food giant McDonald’s. The legal team for the plaintiffs (two very overweight teenagers) had failed to establish that McDonald’s should be liable for their clients’ obesity, in part because we all know (don’t we?) that fast food may be harmful and that it isn’t meant to be eaten for three meals a day. At the start of Spurlock’s McTrial, his GP and other professional advisers — a gastroenterologist, a cardiologist, a dietitian and an exercise physiologist — were expecting to see relatively minor changes as a result of his “Mac attack”, among them a rise in serum lipid levels and an increase in weight. None of the team seemed particularly worried. But fairly early on they start advising him to stop, or at least modify, what he is doing. A little later they are trying to convince him, in no uncertain terms, to stop the experiment altogether. His dietitian worried about weight gain and poor nutrition; the doctors were concerned about his rising liver enzyme levels; and his mother offered him a portion of her liver should he need a transplant! His girlfriend Alex — who just happens to be a vegan chef — stood by her man despite her fears for his health. Everyone can see where Spurlock is heading, and the trial doesn’t end a moment too soon for this young man’s peace of mind. Throughout, Morgan’s ever-present, ironic humour keeps the viewer entertained and engaged — for example, after his baseline physical, which included a per-rectal examination, he quipped: “I like my doctor to be thorough”, and after his first super size meal he reports symptoms of “McGas”, “McSweats” and McTwitches”, as well as “feeling a little McCrazy”. The film travels well beyond the filmmaker’s personal experiences. Although McDonald’s has received special attention in the film because, as one interviewee said, it “lures in young children”, Spurlock also takes us to schools, professorial offices and boardrooms around America, showing us how fast foods in general and their supporters have conquered not only neighbourhoods but also school canteens and government legislators. Former US Surgeon General David Satcher and Professor Kelly Brownell, Director of the Yale Center for Weight and Eating Disorders, are among the many experts interviewed. Spurlock’s film explains how fast food can conquer us as individuals (and whole nations?), with its cocktails of addictive constituents and why it has a starring role in the current epidemic of obesity and obesity-related diseases. It is no surprise that this documentary is already winning major awards, including for Best Director at this year’s Sundance Film Festival. Super size me exhibits a wonderful balance: as funny as it is informative, as intimate and personal as it is professional, and as affectionate about McDonald’s as it is castigating. If only all fast food was as balanced nutritionally.

Ann T Gregory

Leon A Flicker,* Samuel D Vasikaran,† Jenny Thomas,‡ John G Acres,§ Paul E Norman,¶ Konrad Jamrozik,** Nicola T Lautenschlager,†† Peter J Leedman,‡‡ Osvaldo P Almeida§§ * Professor of Geriatric Medicine, ‡ Research Nurse, School of Medicine and Pharmacology, § Research Fellow, School of Medicine and Pharmacology and School of Psychiatry and Clinical Neurosciences, ¶ Associate Professor of Surgery, †† Senior Lecturer in Psychiatry of Old Age, ‡‡ Professor of Medicine, §§ Professor of Psychiatry of Old Age; University of Western Australia, Royal Perth Hospital, Box X2213, Perth, WA 6000. † Head, Department of Core Clinical Pathology and Biochemistry, Royal Perth Hospital, Perth, WA. ** Professor of Primary Care Epidemiology, Imperial College London, London, UK. leonflicATcyllene.uwa.edu.au To the Editor: Elevated levels of homocysteine (Hcy) have recently been associated with increased risk of vascular events1 and dementia.2 The clearance of Hcy is dependent on three vitamins — folate, B6, and B12. Vitamin B12 deficiency has been described in older people for over 40 years,3 and may have wide-ranging effects through this vitamin’s influence on Hcy. The aims of this study were to examine serum B12 and folate status, and their relationships with plasma Hcy concentrations, in community-dwelling healthy older people living in Perth. Older men and women were recruited from two different sources: 299 men aged 75 years and over were recruited from a large population-based study of screening for abdominal aortic aneurysm,4 where 70% of those invited joined the project; and we recruited 273 community-dwelling women aged 70 years and over through advertisements. Exclusion criteria for both groups included significant cognitive impairment, severe physical illness and current use of B-group vitamin supplements. The Human Research Ethics Committee at the University of Western Australia approved the study, and all participants provided informed consent. Fasting total plasma Hcy, serum B12 and folate concentrations were measured in all participants, and serum creatinine concentration was measured in the men only to calculate glomerular filtration rate (cGFR). For analyses, the variable plasma Hcy was heavily skewed to the right and natural logarithmic transformation was used. Pearson’s product moment correlations were calculated for univariate analyses of continuous variables. Descriptive statistics are presented in Box 1. Fourteen per cent and 1% of the men, and 6% and 1% of the women, were deficient in B12 and folate, respectively. Hcy concentrations above upper reference limits (15 μmol/L for men and 13 μmol/L for women) were found in 24% of both men and women. There were significant (P < 0.001) positive correlations between age and log Hcy concentration for men (r = 0.23; 95% CI, 0.12–0.33) and women (r = 0.25; 95% CI, 0.13–0.36), inverse correlations between B12 and log Hcy concentrations for men (r = – 0.25; 95% CI, – 0.14 to – 0.35) and women (r = – 0.30; 95% CI, – 0.19 to – 0.41), and inverse correlations between folate and Hcy concentrations for men (r = – 0.43; 95% CI, – 0.33 to – 0.52) and women (r = – 0.28; 95% CI, – 0.16 to – 0.39). Plots of log Hcy against B12 and folate concentrations for all participants are presented in Box 2. Under multiple regression, the association of B12 and folate concentrations with log Hcy concentration remained after adjustment for age and cGFR in men only; beta values (SE) were: – 0.00060 (0.00011) for B12 concentration; – 0.0155 (0.0017) for folate concentration; – 0.0029 (0.0008) for cGFR; and 0.012 (0.005) for age. In this sample there were high prevalences of B12 deficiency and hyperhomocysteinaemia. Although the prevalence of folate deficiency was substantially lower, there were still moderate inverse associations between serum folate and Hcy concentrations. Unfortunately, vitamin B12 deficiency of this kind may not be universally corrected with small doses of oral supplements,5 and this has intensified concerns about precipitating neurological complications by population-based folate supplementation.6 There is a need for intervention studies of B-group supplements to evaluate whether the risks associated with hyperhomocysteinaemia can be ameliorated. 1: Demographic characteristics, serum B12 and folate, and plasma homocysteine in 299 older men and 273 older women Men Women Mean (SD) Range Mean (SD) Range Age (years) 78.9 (2.8) 68–86 74.8 (4.4) 70–92 Weight (kg) 78.4 (1.2) 50.6–119.5 69.3 (1.3) 39.0–120.0 Height (cm) 171 (6.5) 150–197 159 (6.7) 132–176 Body mass index (kg/m2) 26.6 (3.5) 16–37 27.4 (5.3) 17–52 Ever smoked 66% 41% Ever drank alcohol 95% 66% Serum folate (nmol/L) 24.3 (7.6) 5.5–45.3 (RI, 7–34) 25.3 (7.6) 3.9–45.2 (RI, 7–34) Serum B12 (pmol/L) 254.5 (116.7) 57–890 (RI, 140–646) 313.5 (158.7) 59–1270 (RI, 140–646) Plasma Hcy (μmol/L) 13.50 (5.3) 6.7–70.5 (RI, 6.0–15.0) 11.46 (6.8) 3.8–96 (RI, 5.0–13.0) Glomerular filtration rate (mL/min) 78.3 (16.3) 35.8–142.4 SD = standard deviation. Hcy = homocysteine. RI = reference interval. 2: Plot of serum B12 and folate concentration against log homocysteine concentration (with regression line for B12) in 299 older men and 273 older women

Leon A Flicker · Samuel D Vasikaran · Jenny Thomas · John G Acres · Paul E Norman · Konrad Jamrozik · Nicola T Lautenschlager · Peter J Leedman · Osvaldo P Almeida

Possible protection against cancer and improved immune function make supplementation with selenium attractive, but its toxicity and other unknown effects urge caution Selenium presents a nutritional conundrum because of its dual status as a highly toxic, but essential, trace element.1 The eightfold gap between the estimated average requirement2 and the upper limit of safe intake is relatively narrow, so questions of too much and too little are important. Additional key questions pertain to adequate versus optimal status, and reflect the shift in focus of nutrition from preventing deficiency towards promoting optimal health. As is the case for many micronutrients, the quest for protective effects of selenium (Se) intakes above requirements is rapidly gaining momentum. Ever since the biological role of Se in humans was first delineated less than 30 years ago, evidence for its increasing scope and importance to human health has been rapidly accumulating. Increasingly, Se depletion, as opposed to “deficiency”, is being associated with a range of health outcomes, including viral infection, reproduction, mood, thyroid function, cardiovascular disease, inflammatory conditions, immune function and cancer protection. 3,4 It is possible to advance a range of theoretical arguments that suboptimal Se status may have an effect on health, but evidence of a direct relationship to health outcomes is very limited. Se research is incipient and there is a paucity of data. For example, Se reference values that have been adopted in the United States2 are based on only two studies, one of which was a 1983 Chinese study of poor quality. Research is hampered by substantial difficulties in assessing and interpreting Se intakes and status, which become even more significant in the context of extreme global variations in the Se content of soil, food and human tissue.3,4 We do know that Se has key roles in redox regulation and antioxidant function, and hence in membrane integrity, energy metabolism and protection against DNA damage. 1,3,4 These and other functions are mediated through over 35 selenoproteins, which require adequate Se intake for synthesis and expression. Selenoproteins include several forms of the enzymes glutathione peroxidase (GPx), thioredoxin reductase and iodothyronine 5'-deiodinase. The number of known selenoproteins has almost trebled over the past 7 years,1,3 although the roles of several remain undefined. Plasma Se concentration is the most commonly used indicator of Se status.2 Low Se intakes, plasma Se concentrations and GPx activities have direct, linear associations up to a threshold plasma Se concentration (70–100 μg/L), beyond which GPx activity plateaus. This maximum GPx concentration is thought to represent repletion, and commensurate Se intake forms the basis of recommended dietary requirements.2,5 Concentrations of other selenoproteins are also influenced by Se intake and may have a role as functional indicators of Se status, 2,3,5 but assay methods and reference standards are at an early stage, and comparisons between studies are difficult. There is differential hierarchical expression of the selenoproteins, with relative preservation of the presumably more metabolically important at lower intakes of Se. 1,3 However, we do not clearly understand the health implications of submaximal expression of the selenoproteins. There are enormous geographical variations in the Se content of soil and food, and hence in Se intakes and concentrations in human blood and tissues.3,4,6 Thus, it is essential to use local data for monitoring and interpreting Se status.6 The 2000 US Recommended Dietary Allowance (RDA) is 55 μg/day.2 The 1987 Australian RDAs are 70 μg/day for women and 85 μg/day for men, but these are currently under review.7 Organic selenomethionine is the predominant form of Se in food. The most important dietary sources of Se are meat, poultry, fish and cereals (although brazil nuts are very high in Se and certain fish also have particularly high levels). Accurate assessment of intake is exceptionally difficult because the Se content of food is so variable. 2,3,6,8 Estimates of Se intakes include 106 μg/day in a large representative US sample,2 and a range of 29–70 μg/day in Europe.3 Extremely limited Australian data suggest intakes of around 90 μg/day,6 while more comprehensive New Zealand data show intakes as low as 28 μg/day, and indicate that conventional dietary intake methods are inadequate for estimating Se intake.8 Inorganic Se (selenite and selenate) is only available through supplementation, is generally less bioavailable and produces a different physiological response than organic forms of Se. 1,2 Environmental changes and agricultural practices may be reducing Se concentrations in soil.4 These factors, in conjunction with trade barriers (eg, the cessation of importing high-Se US wheat to the European Union), appear to be associated with a decline in the availability of Se through the food chain and in human Se status, particularly in Europe.3 Changes in food supply and habits, including the importation of Australian wheat, have improved the previously marginal Se status of New Zealanders.9 Twenty-six European studies since 1990 all reported mean plasma Se concentrations below 100 μg/L, the level postulated to be required for GPx saturation and cancer protection. 3,10 Ten of these studies reported plasma Se levels under 70 μg/L,3 postulated by others to be associated with GPx saturation.11 A representative US plasma Se level was 124 μg/L.2 In this issue of the Journal (page 383), Lyons and colleagues present evidence that although mean plasma Se concentrations of South Australians are relatively high by European standards, they may be declining, and over a third of their sample had levels below 100 μg/L.12 Overt human Se deficiency is rare. It is manifested as Keshan disease, an endemic fatal cardiomyopathy, which is virtually unknown outside areas of China, where the levels of Se in soil and dietary Se intake are extremely low. A few studies have reported Se deficiency as a result of long-term total parenteral nutrition.1 Despite myriad claims for potential relationships between a range of diseases and Se status, there are no clear population health outcomes that can be attributed to Se status in countries like New Zealand, where intakes are very low. 1,9 The evidence for an effect of Se on health outcomes is strongest in cancer prevention. Secondary findings from the 10-year US Nutritional Prevention of Cancer Trial demonstrated a protective effect of supplementation with 200 μg/day of organic Se (from yeast) on total cancer incidence and mortality, and on prostrate cancer incidence (relative risk, 0.75, 0.59, and 0.48, respectively).10 The effects were stronger in men and in those with lower baseline plasma Se levels (< 105 μg/L), and were not found for a range of other site-specific cancers. Of concern is that in the top tertile for baseline plasma Se level there may be an association between supplementation and increased risk of breast cancer and melanoma, as well as overall cancer incidence.10 There are numerous limitations to what was a small study, and the relatively high baseline plasma Se levels (114 μg/L) make it difficult to generalise the findings. It appears that protection from cancer may require supplementation beyond correction of depletion and maximal expression of the selenoproteins. Several large trials are under way to clarify the benefits and risks of Se supplementation with respect to tumorigenesis. More speculative and tantalising is the potential association between Se and immune function. Evidence suggests that reduced Se status may be associated with the incidence of clinical infection in adults,13 and supplementation of apparently Se-replete individuals potentially enhances immune function.3 Animal studies indicate that the Se status of the host can genetically alter invading viral pathogens, so that a normally benign strain may become virulent in an Se-deficient host.14 This is of interest given the emergence of new influenza virus strains from China, where there are significant areas of overt Se deficiency, and given the decline in Se status associated with progression of HIV infection.3 In summary, there has been an explosion of interest in the biological role of Se and the potential health implications of Se status. Much of the evidence in humans is descriptive, and the dearth of quality prospective trials means the links with many diseases are still controversial and, in many cases, speculative. There are very limited representative data on Se content of the food supply and Se status in many populous parts of the world, and no nationally representative Australian data. Lyons and colleagues provide the most comprehensive Australian plasma Se data to date,12 but they are neither prospective nor representative, and there may be variations according to states.6 Outcomes of research on Se in the next decade are likely to be important, and we urgently need more Australian data. Meanwhile, it is necessary to remember that selenium is toxic.1-3 Intakes below 400 μg/day are considered safe for almost all individuals.2 As illustrated by the Nutritional Prevention of Cancer Trial,10 outcomes of Se supplementation are variable and may not be without risk. Benefits and an appropriate dose for supplementation remain controversial. Until further evidence is available, supplementation should be recommended with caution, and overconsumption should be avoided.3

Lynne A Daniels PhD, APD

Objective: To assess trends in selenium status in South Australians from 1977 to 2002.Design: Six cross-sectional surveys.Participants: 117 participants in 1977, 30 in 1979, 96 and 103 (separate surveys) in 1987, 200 in 1988, and 288 volunteer blood donors in 2002. A total of 834 healthy Australian adults (mean age, 42 years [range, 17–71 years]; 445 were male).Main outcome measures: Plasma and whole blood selenium concentrations.Results: The 2002 survey yielded a mean plasma selenium concentration of 103 μg/L (SE, 0.65), which reached the estimated nutritional adequacy level of 100 μg/L plasma selenium. Mean whole blood selenium declined 20% from the 1977 and 1979 surveys (mean whole blood selenium concentration, 153 μg/L) to the 1987, 1988 and 2002 surveys (mean whole blood selenium concentration, 122 μg/L). Plasma selenium was higher in men (P = 0.01), and increased with age in both men and women (P = 0.008).Conclusions: In healthy South Australian adults sampled from 1977 to 2002, whole blood and plasma selenium concentrations were above those reported for most other countries and in most previous Australian studies, notwithstanding an apparent decline in selenium status from the late 1970s to the late 1980s.

Graham H Lyons BAgricSc, MPH · James C R Stangoulis BAgricSc, PhD · Lyndon T Palmer BSc · Robin D Graham PhD, DAgricSc · Geoffrey J Judson BRurSc, PhD · Janine A Jones BCom/Eco, MStatSocAust

Re: “Overweight and obesity in Australia: the 1999–2000 Australian Diabetes, Obesity and Lifestyle Study (AusDiab)”, by Adrian J Cameron, Timothy A Welborn, Paul Z Zimmet et al in the 5 May 2003 issue of the Journal (Med J Aust 2003; 178: 427-432). In Box 3 and Box 4 (pages 429 and 430), the row headings for the weekly income categories were reversed — the reference category should have been the lowest (rather than the highest) income group. These sections of Boxes 3 and 4, as they should have appeared, are shown below. This reversal affected our interpretation of the results of the study. In the Results section (page 431), we stated that “Increasing income increased the risk of obesity in women. Although no such association was significant for men, both the BMI and waist circumference data suggested that middle-income men tended to be more obese than the lowest income group.” This should have read: “Increasing income decreased the risk of obesity in women. Although no such association was significant for men, both the BMI and waist circumference data suggested that middle-income men tended to be more obese than the highest income group.” In the Discussion section (page 432), we stated that “In men, minor trends for middle-income groups to be more obese and the least affluent to be thin were observed, although these were not significant. Women, by contrast, showed a strong positive graded association between income and obesity.” This should have read: “In men, minor trends for middle-income groups to be more obese and the most affluent to be thin were observed, although these were not significant. Women, by contrast, showed a strong negative graded association between income and obesity.” The association between obesity and income was not a major outcome of the study, with the results focusing on the prevalence of obesity and the association of obesity with physical activity and television viewing time. The html and pdf versions of the article published in the eMJA were corrected on 20 March 2004. 3: Association between obesity (measured using body mass index [BMI]* [n = 4996] and waist circumference* [n = 4984]) and potential risk factors among Australian men Body mass index Waist circumference n Adjusted odds ratio† (95% CI) n Adjusted odds ratio† (95% CI) Smoking status Non/ex-smoker 4048 1.00 4041 1.00 Smoker 865 0.71 (0.48–1.04) 860 0.63 (0.51–0.78)‡ Physical activity§ Lowest quintile 946 1.00 946 1.00 Highest quintile 1021 0.70 (0.46–1.06) 1022 0.56 (0.42–0.75)‡ Television viewing¶ Lowest quintile 784 1.00 783 1.00 Highest quintile 1094 1.86 (1.30–2.67)‡ 1093 1.97 (1.48–2.63)‡ Education University/Further education 2089 1.00 2086 1.00 Completed high school 901 1.14 (0.92–1.42) 900 0.93 (0.69–1.27) Some high school completed 1693 2.19 (1.6–3.01)‡ 1684 1.65 (1.17–2.33)‡ Primary school/never attended school 309 2.40 (1.59–3.61)‡ 310 2.31 (1.69–3.15) ‡ Country of birth Australia/New Zealand 3727 1.00 3713 1.00 United Kingdom/Northern Ireland 612 0.92 (0.65–1.29) 611 0.89 (0.68–1.16) Rest of world 653 0.85 (0.67–1.08) 656 0.60 (0.40–0.92)‡ Weekly income (A$) 0–199 502 1.00 502 1.00 200–399 1541 1.18 (0.80–1.74) 1533 1.30 (0.93–1.81) 400–599 1162 1.12 (0.78–1.60) 1157 1.29 (0.91–1.82) 600–799 1133 1.07 (0.77–1.50) 1134 1.23 (0.84–1.80) 800–1499 571 1.06 (0.70–1.62) 573 0.97 (0.69–1.36) 1500+ 37 0 .68 (0.22–2.07) 36 0.51 (0.17–1.54) Occupation** Skill level 1 1243 1.00 1244 1.00 Skill level 2 540 0.65 (0.46–0.93)‡ 534 0.74 (0.49–1.14) Skill level 3 658 0.48 (0.32–0.72)‡ 659 0.43 (0.26–0.70)‡ Skill level 4 604 0.95 (0.63–1.43) 602 0.84 (0.58–1.21) Skill level 5 304 0.93 (0.52–1.63) 304 0.92 (0.46–1.84) Others 1629 0.56 (0.35–0.89)‡ 1623 1.07 (0.66–1.72) * Obesity defined as BMI ≥ 30 kg/m2, or waist circumference ≥ 102 cm. † Model adjusted for age and all other risk factors in the table. ‡ Significantly different from reference (P < 0.05). § Quintiles of physical activity: < 30 minutes = lowest, ≥ 550 minutes = highest. ¶ Quintiles of television viewing: < 300 minutes = lowest, ≥ 1260 minutes = highest. ** Skill levels: 1 — managers, administrators and professionals; 2 — associate professionals; 3 — tradespersons, advanced clerical and service workers; 4 — intermediate clerical, sales and service workers, intermediate production and transport workers; 5 — elementary clerical, sales and service workers and labourers; and others — students, retirees, pensioners and unemployed. 4: Association between obesity (measured using body mass index [BMI]* [n = 6071] and waist circumference* [n = 6075]) and each of the potential risk factors among Australian women Body mass index Waist circumference n Adjusted odds ratio† (95% CI) n Adjusted odds ratio† (95% CI) Smoking status Non/ex-smoker 5123 1.00 5128 1.00 Smoker 857 0.70 (0.51–0.97)‡ 858 1.03 (0.76–1.39) Physical activity§ Lowest quintile 1254 1.00 1255 1.00 Highest quintile 1221 0.47 (0.31–0.72)‡ 1221 0.53 (0.34–0.80)‡ Television viewing¶ Lowest quintile 992 1.00 994 1.00 Highest quintile 1292 1.82 (1.19–2.76)‡ 1295 2.27 (1.55–3.32)‡ Education University/Further education 1985 1.00 1985 1.00 Completed high school 1185 1.04 (0.77–1.40) 1185 1.31 (1.01–1.70)‡ Some high school completed 2481 1.48 (1.19–1.83)‡ 2487 1.47 (1.19–1.82)‡ Primary school/never attended school 419 2.12 (1.18–3.80)‡ 417 2.68 (1.64–4.36)‡ Country of birth Australia/New Zealand 4672 1.00 4677 1.00 United Kingdom/Northern Ireland 644 0.95 (0.68–1.34) 646 1.01 (0.69–1.49) Rest of world 754 0.80 (0.66–0.97)‡ 751 0.72 (0.57–0.92)‡ Weekly income (A$) 0–199 1034 1.00 1034 1.00 200–399 2046 0.85 (0.67–1.08) 2046 0.93 (0.74–1.19) 400–599 1270 0.87 (0.66–1.15) 1274 0.79 (0.62–1.02) 600–799 1064 0.57 (0.40–0.82)‡ 1062 0.62 (0.46–0.83)‡ 800–1499 499 0.67 (0.48–0.93)‡ 501 0.59 (0.37–0.94)‡ 1500+ 20 0.63 (0.19–2.11) 20 0.46 (0.13–1.65) Occupation** Skill level 1 990 1.00 989 1.00 Skill level 2 390 0.82 (0.41–1.65) 389 0.80 (0.48–1.34) Skill level 3 354 0.88 (0.55–1.40) 355 0.93 (0.58–1.48) Skill level 4 886 1.11 (0.72–1.71) 884 0.74 (0.53–1.05) Skill level 5 549 0.84 (0.45–1.57) 551 0.90 (0.60–1.36) Others 2869 0.94 (0.54–1.63) 2874 1.08 (0.71–1.64) * Obesity defined as BMI ≥ 30 kg/m2, or waist circumference ≥ 88 cm. † Model adjusted for age and all other risk factors in the table. ‡ Significantly different from reference (P < 0.05). § Quintiles of physical activity: < 20 minutes = lowest, ≥ 390 minutes = highest. ¶ Quintiles of television viewing: < 240 minutes = lowest, ≥ 1200 minutes = highest. ** Skill levels: 1 — managers, administrators and professionals; 2 — associate professionals; 3 — tradespersons, advanced clerical and service workers; 4 — intermediate clerical, sales and service workers, intermediate production and transport workers; 5 — elementary clerical, sales and service workers and labourers; and others — students, retirees, pensioners and unemployed.

Adrian J Cameron MPH · Paul Z Zimmet MD, FRACP FAFPHM · David W Dunstan PhD · Marita Dalton GradDipEpidemiol · Jonathan E Shaw MD, MRCP · Timothy A Welborn MB BS, PhD · Neville Owen PhD · Jo Salmon PhD · Damien Jolley MSc

Lyndon E Llewellyn,* Cedric E Robillot,† Andrew P Negri‡ * Principal Research Scientist, Bioactive Molecule Discovery; † Senior Research Specialist, Bioinnovation; ‡ Senior Research Scientist, Bioinnovation, Australian Institute of Marine Science, PMB 3, Townsville, QLD 4810. L. LlewellynATaims.gov.au To the Editor: Seafood is not only consumed as food, but also as dietary supplements and complementary medicines. Examples are capsules of freeze-dried oysters and mussels, or freeze-dried extract of shellfish meat, sold as reputed antihypertensives, cardioprotectants, and anti-inflammatories, among other medical claims. However, oysters and mussels can become dangerously toxic after they ingest poisonous microscopic algae. If these molluscs are sold as food in Australia, they are subject to the Food Standards Code,1 under which their sale is prohibited if biotoxin levels per kilogram of wet shellfish meat exceed 800 g of paralytic shellfish poisons, 200 mouse units of neurotoxic shellfish poisons, 200 g of diarrhoetic shellfish poisons, or 20 mg of amnesic shellfish poisons. Shellfish capsules can be simply manufactured by milling dried meat and encapsulating the powder, a process unlikely to degrade shellfish biotoxins, which are stable to heat, pressure and freeze-drying. 2,3 Such capsules may then become subject to regulation by the Australian Therapeutic Goods Administration (TGA), which distinguishes therapeutics from food, on the basis of whether there is a “tradition of use as a food in the form presented”, especially if there is an associated health claim. Such complementary medicines can be either “registered” or “listed”. Registered medicines require extensive safety, quality and efficacy data. Listed medicines are considered to pose a lower risk than registered medicines, and regulations allow product sponsors to “self-assess” products. Listing is a route commonly taken for complementary medicines. A pertinent example is the TGA listing of therapeutic goods containing dried green-lipped mussel (Perna canaliculus).4 Where manufacturers of shellfish capsules have undertaken the responsibility of ensuring product safety, they invariably adopt existing biotoxin testing protocols developed for food safety. However, as the allowable biotoxin level is based on wet weight, and the dry weight of bivalve shellfish is only 10%–15% of the wet weight,5 safety limits for shellfish meat as food are incorrect by an order of magnitude, and potentially more for capsules containing extracts of shellfish meat. While important for acute exposure to these toxins, this may be even more relevant in chronic exposure. Okadaic acid, the cause of diarrhoetic shellfish poisoning, is a tumour promoter,6 and epidemiological studies suggest that rates of cancer have increased in regions with regular dietary exposure to low levels of this toxin.7 Capsules are available that contain 500 mg of dried shellfish meat, which may equate to 5 g of wet shellfish meat.5 Unlike a shellfish meal, which may be considered equivalent to a single acute exposure, recommended doses for shellfish capsules can be as many as five capsules a day for many weeks, if not months, therefore magnifying the risk of chronic exposure. It is known that different classes of biotoxins can co-occur in shellfish, adding to the potential hazard outlined here.8 Further complications arise because some shellfish capsules include other natural extracts (such as ginseng) or pharmaceutical formulations that might affect toxin uptake. While this situation needs to be subjected to risk assessment, testing products in accordance with an inappropriate standard can make them seem safe when they might not be. This is especially so for products which are usually self-prescribed, and where patients can exceed recommended doses in the belief that more is better. For products such as shellfish capsules that straddle the food/therapeutic divide, it is better for manufacturers to test the final consumer product and not the raw supply.

Lyndon E Llewellyn · Cedric E Robillot · Andrew P Negri

Terry J Coyne,* Michael G Findlay,† Torukiri I Ibiebele,‡ David W Firman§ * Senior Lecturer, School of Population Health, University of Queensland, Public Health Building, Medical School, Herston Road, Herston, QLD 4029; † Acting Senior Analyst, ‡ Assistant Analyst, Epidemiology Services Unit, Queensland Health, Brisbane; § Team Leader, Surveys and Social Statistics, Office of Economic and Statistical Research, The Treasury, Queensland Government, Brisbane.t.coyneATsph.edu.au To the Editor: While the rates of overweight and obesity among Australian adults, as determined by the Australian Diabetes, Obesity and Lifestyle Study (AusDiab),1 may be alarming to some, they may in fact be underestimates of the true prevalence of overweight and obesity. The AusDiab study design2 and its low response rates indicate that the results will need to be interpreted with caution. Firstly, the AusDiab study design excluded rural and predominantly Indigenous census collection districts (CDs). In Queensland, all CDs selected were capital city or other major urban centres (Rural and Remote Areas Classification, categories 1 and 2);3 thus, people living in major rural centres (such as Rockhampton or Bundaberg) or major remote centres (such as Mt Isa) were excluded (ie, in Queensland, about 20% of the population were excluded). Secondly, another potential bias may have been introduced by the Socio-Economic Indexes for Areas (SEIFA) scores of the CDs sampled in the AusDiab study. For example, the overall SEIFA score for the CDs included in Queensland was 1035 (73rd percentile), well above the state average. Finally, the response rates in the AusDiab study were low: only 29% of those estimated to be eligible, and only 52% of those invited, actually completed the study. Our analysis of risk factors of Queensland-AusDiab participants suggests that these participants may have been more health conscious than the general Queensland population. Rates of smoking reported for men and women were considerably lower in the Qld-AusDiab cohort compared with those in the Queensland phase of the 2001 National Health Survey4 (17.3% and 14.5% v 28.4% and 19.8%, respectively). Compared with results of a Queensland Omnibus telephone survey5 conducted at about the same time, higher proportions of Qld-AusDiab participants reported greater intakes of vegetables (≥ 4 serves/day: 27.4% Qld-AusDiab v 16.4% Omnibus) and fruit (≥ 2 serves/day: 28.9% v 24.3%), and less frequent consumption of fast foods (> 1 day/week: 37.3% v 49.5%). Given the low response rate and possible selection bias in the AusDiab study, we suggest that the overweight and obesity data should be interpreted with caution. Several indicators suggest that these data could be underestimates of the true prevalence of overweight and obesity, and that the AusDiab population may have been of higher socioeconomic status, more health conscious (lower rates of smoking, better dietary intake), and more willing to participate in a lengthy examination than the general Australian population. These factors may all be associated with lower rates of overweight and obesity, and therefore future national surveys will need to take these factors into consideration to obtain more accurate estimates of important determinants of health.

Terry J Coyne · Michael G Findlay · Torukiri I Ibiebele · David W Firman

Adrian J Cameron,* Paul Z Zimmet,† David W Dunstan,‡ Jonathan E Shaw§ * Epidemiologist, † Director, ‡ Research Fellow, § Physician in Diabetes, and Director, Clinical Research; Epidemiology Department, International Diabetes Institute, 250 Kooyong Road, Caulfield, VIC 3162. acameronATidi.org.au In reply: Coyne suggests that, based on comparisons within Queensland, the national prevalence of obesity in our article1 is an underestimate. It should be noted that the Australian Diabetes, Obesity and Lifestyle Study (AusDiab) was designed primarily to produce national, not state-specific, data. Forty-two census collection districts (CDs) were selected Australia-wide, with only six CDs selected within each state. The primary objective of this sample selection was to obtain a nationally representative population, not necessarily one representative of each state. Coyne states that none of the Queensland CDs were in major provincial centres. Of the six Queensland CDs, four were outside Brisbane. From the national sample, 17 of 42 CDs (40.5%) were outside capital cities. As a comparison, 36% of the Australian population lives outside capital cities.2 Regarding selection of CDs, we excluded only those in Statistical Local Areas defined as 100% rural, and those where the Indigenous population made up 10% or more of the overall population.3 This excluded only 5.8% of the total eligible population. If the prevalence of obesity among this group was double the overall prevalence, this would not significantly alter the national rate. While the smoking rates in AusDiab were lower than reported elsewhere, the prevalence of obesity, hypercholesterolaemia and hypertension were in line with trends in a series of surveys over the past 20 years.4 In an extensive analysis of food consumption between AusDiab and the 1995 National Nutrition Survey, the rates of fruit and vegetable consumption were within 4% between the surveys for those most commonly eaten. Since our conclusion was that obesity has increased, the possibility of an underestimate only reinforces our message.

Adrian J Cameron · Paul Z Zimmet · David W Dunstan · Jonathan E Shaw

Overweight and obesity are very common in Australian adults (56%) and children (27%). Rates of overweight and obesity are snowballing and will place greater burdens on health services for the treatment and care of chronic diseases. Prevention is urgently required from health, social and economic perspectives, but the response to date has been inadequate. A long-term, sustained action plan starting with a focus on young people is needed. This should particularly address the “obesogenic” environments causing the epidemic. Although whole-of-government action is required, support from and involvement by parents, carers, community leaders, healthcare professionals, teachers, childcare workers, urban planners, recreation managers, food manufacturers, employers, advertisers, and communicators is essential. The health sector should take the lead, but success will only come from concerted and integrated action across the whole of society. There are now signs of political commitment to addressing overweight and obesity. Doctors should get behind this and help mobilise community support.

John C Catford DM, FRCP, FAFPHM · Ian D Caterson MB BS, PhD, FRACS

To the Editor: We read with interest Abraham’s recent article on bodyweight issues facing young women.1 Abraham notes that this group is at risk of extreme weight-loss behaviours, including excessive exercise and use of slimming tablets. However, recent data from our studies show young women’s high risk of substantial weight gain and obesity.2,3 Obesity entails significant health and social costs for young women. There are also likely to be substantial financial costs associated with efforts to manage weight. The low levels of satisfaction with their body and poor self-esteem among young women, in conjunction with their heightened risk of weight gain, make them prime targets for the slimming industry. A number of studies have estimated what obesity costs and its impact on the healthcare system,4,5 and some have estimated expenditure by consumers on weight-loss products.6 However, none have quantified the financial costs of weight management for women. We recently (January 2002) investigated weight-management strategies among a randomly selected, nationwide sample of 445 women aged 18–32 years. Our study was approved by the Deakin University Human Research Ethics Committee. Women reported their use of and expenditure on nine methods “to lose weight, prevent weight gain, or control body shape” in the previous 12 months. Thirty-one per cent of the sample were overweight or obese (body mass index > 25 kg/m2), and 61% had used at least one weight-loss method in the past 12 months. Box 1 shows the proportions using each method. Some women spent more than $3000, with a mean expenditure of $441 per woman among those using a weight-loss method, or $251 per woman across all women in the sample (Box 2). Extrapolating these results to the population of women in this age group, this equates to almost $414 million per annum spent by young women to manage their weight. Thus, young women are investing considerable amounts of money to manage their weight. Increasing rates of obesity suggest that young women’s efforts to manage their weight are ineffective. Promotion by general practitioners of safe, low-cost weight-management strategies — including low-fat healthy eating, as well as walking for exercise — could help alleviate the substantial health and economic costs of obesity and weight control. 1: Proportion of women using weight management strategies 2: Mean expenditure by women using weight management strategies

Kylie Ball PhD · Sari Andajani-Sutjahjo PhD · David Crawford PhD

Judy A Seal,* Eric M Johnson,† Zelda Doyle,‡ Kelly Shaw§ * State Nutrition Officer, † State Food Officer, § Public Health Registrar, Public and Environmental Health, Department of Health and Human Services, GPO Box 125, Hobart, TAS 7001; ‡ Field Officer, Broad Street Consultants, Tasmanian Iodine Monitoring Program, Lauderdale, TAS. judy.sealATdhhs.tas.gov.au To the Editor: Tasmania has been recognised for many years as an area of endemic iodine deficiency.1 According to the World Health Organization, populations are considered iodine sufficient if population median urinary iodine (UI) levels exceed 100 μg/L, with less than 10% of the UI levels below 50 μg/L.2 Two random surveys (1998–99 and 2000–01) of Tasmanian school children aged 4–14 years suggest mild iodine deficiency. Median UI levels were 75 μg/L and 77 μg/L, with 13% and 21%, respectively, of the UI levels below 50 μg/L.3 In response to these findings, an iodine supplementation program was introduced in October 2001. Tasmanian bakeries were encouraged to switch to using iodised salt in place of regular salt. The program is voluntary, with participating bakeries asked to sign a memorandum of understanding. Industry advice suggests that bakeries that have signed the memorandum produce about 80% of the bread available for consumption in Tasmania. The Tasmanian Iodine Monitoring Program commenced in July 2002. Its objectives are to determine the effect of iodine supplementation of bread on the general population and on high-risk groups, and to identify any negative health effects associated with the program. Preliminary results from the monitoring are encouraging. Children were selected using a random cluster sampling approach. The sampling frame included all Grade 4 classes in all government, Catholic and independent schools in Tasmania. To date, 148 urine samples have been collected, with results from 124 available (test completion rate, 84%). The median UI level from the preliminary results is 97 μg/L (95% CI, 90–109 μg/L), with 10.5% below 50 μg/L. Ongoing monitoring will provide a more rigorous evaluation of the effects of the iodine supplementation program. Early indications suggest the supplementation program may be achieving its goal of improving the iodine status of the Tasmanian population. The monitoring program will continue for the next 4 years, with regular surveys to detect any changes in the population’s iodine status. It will be challenging to retain the ongoing participation of the bread industry if, in the future, there is increased reliance on premixed and ready-to-bake products from outside Tasmania. Maintaining bread supplementation in Tasmania would then require cooperation from interstate suppliers to ensure iodine supplemention of these premixes and ready-to-bake products. Given that recent research has shown mild iodine deficiency in other parts of Australia and New Zealand, perhaps it is time for a bi-national solution to the problem.4,5

Judy A Seal · Eric M Johnson · Zelda Doyle · Kelly Shaw

Objective: To determine the prevalence of iodine deficiency in primary school children in an Australian urban population.Design and setting: A cross-sectional survey of school children aged 5–13 years attending a public school on the Central Coast of New South Wales in November 2000.Participants: 324 (70%) of the 465 children enrolled in the school (180 boys; 144 girls).Main outcome measures: Thyroid volumes compared with World Health Organization/International Council for the Control of Iodine Deficiency Disorders (WHO/ICCIDD) thyroid volume reference values. Iodine status based on WHO/ICCIDD urinary iodine concentration (UIC) categories (normal, ≥ 100 μg per litre of urine [μg/L]; mild iodine deficiency, 50–99 μg/L; moderate deficiency, 20–49 μg/L; severe deficiency, < 20 μg/L); not more than 20% of the population should have a UIC below 50 μg/L.Results: Median UIC for school children was 82 μg/L, and 14% of children had UICs below 50 μg/L. Thyroid volume reference values indicated a prevalence of goitre of zero. In girls, only four (3%) and one (1%) had thyroid volumes above the WHO/ICCIDD medians by age and body surface area (BSA), respectively (P < 0.001). In boys, three (2%) and one (1%) had thyroid volumes above WHO/ICCIDD medians by age and BSA, respectively (P < 0.001).Conclusion: Despite the median UIC being less than ideal, most children were not goitrous. This underscores the importance of using physiological outcome measures in areas where iodine deficiency is marginal before concluding the need for iodine supplementation based purely on median UIC. We call for a systematic national survey to determine iodine status using a combination of iodine deficiency indicators.

Kamala Guttikonda MB BS, FRACP · Steven Boyages FRACP, PhD · Cheryl A Travers BSc · Peter R Lewis MB BS, FAFPHM

Jennifer A O'Dea Senior Lecturer, Faculty of Education, University of Sydney, Building A35, Sydney, NSW 2006. j.o'deaATedfac.usyd.edu.au To the Editor: As part of a large, national nutrition study, height and weight were measured among 4441 students from 38 schools randomly selected from lists of all state and territory schools in Australia in 2000. Public, private and Catholic schools, in both rural and urban areas, were represented. Schools were categorised as being of low or middle/high socioeconomic status (SES),1 based on direct measurement of parental income. Parental consent was obtained, and the study was approved by the University of Sydney Ethics Committee and all state departments of education. Overweight and obesity, as defined by an international standard definition,2 were identified in 17.3% and 6.4% of participants, respectively. These characteristics showed a trend towards greater prevalence among students from low-SES backgrounds compared with those from middle/high-SES backgrounds for the total group (19% v 16.8% overweight [P = 0.09]; 8.9% v 5.8% obese [P = 0.02]), females (19.7% v 17.2% overweight [P = 0.2]; 6.9% v 6.2% obese [P = 0.56]), and males (18.5% v 16.3% overweight [P = 0.23]; 9% v 5.5% obese [P = 0.003]), although not all differences were statistically significant. After controlling for SES differences in age and height, mean body mass index (BMI) was significantly higher among low-SES than middle/high-SES participants for the total group (20.3 kg/m2 [95% CI, 20.1–20.5 kg/m2] v 19.7 kg/m2 [95% CI, 19.6–19.9 kg/m2]; P < 0.001), females (20.4 kg/m2 [95% CI, 20.1–20.7 kg/m2] v 19.8 kg/m2 [95% CI, 19.6–19.9 kg/m2]; P < 0.001), and males (20.2 kg/m2 [95% CI, 20.0–20.5 kg/m2] v 19.6 kg/m2 [95% CI, 19.5–19.8 kg/m2]; P < 0.001). A breakdown of results by SES, sex and school level is shown in the Box. Low-SES primary school children were also 1–2 cm shorter, on average, than middle/high-SES primary school children (boys: mean 141.5 cm [95% CI, 140.6–142.5 cm] v 143.5 cm [95% CI, 143.0–144.0], P < 0.001; girls: mean 141.0 cm [95% CI, 140.8–142.6 cm] v 143.3 cm [95% CI, 142.5–143.6 cm], P = 0.01). The average proportions of overweight and obese children and adolescents in the study were similar to those found in other Australian studies.3-5 The results suggest that SES is a factor in the development of overweight and obesity among Australian school children. This may be a relatively recent trend, as these data were obtained in late 2000. Low SES in children may also be associated with nutritional deprivation and height retardation. Further research should clarify these relationships among children from low, middle and high SES backgrounds, as well as examining the combined impact of both SES and ethnicity. School students classified as overweight or obese* according to socioeconomic status (SES), school level and sex Males (n = 2232) Females (n = 2209) Low SES (n = 574) Middle/high SES (n = 1658) Low SES (n = 508) Middle/high SES (n = 1701) Primary school students (grades 1–6; ages 6–13 years) Overweight students 19.4% (42/216) 16.2% (110/680) 23.2% (51/220) 17.8% (136/766) Obese students 6.9% (15/216) 5.3% (36/680) 6.4% (14/220) 5.7% (44/766) High school students (grades 7–12; ages 13–18 years) Overweight students 17.6% (63/358) 16.4% (160/978) 17.0% (49/288) 16.8% (157/935) Obese students 10.1% (36/358) 5.6% (55/978) 7.3% (21/288) 6.5% (61/935) * Overweight and obesity are classified according to the international standard definition.2

Jennifer A O'Dea

Many young women feel that body image and exercise are important for their self-esteem, want to lose weight, are afraid they might gain weight, and feel fat. Interventions that improve self-esteem, encourage communication and help adolescents to be supportive of each other may prevent some of these women from developing eating disorders. If an eating disorder is suspected, it may be useful for physicians to ask about fear of loss of control over the body, eating, weight and shape; and preoccupation with food, eating, nutrition, body weight and shape, as these issues may differentiate those at greater risk.

Suzanne F Abraham MSc, PhD