In reply: We are pleased that our article stimulated debate regarding the inappropriate measurement of insulin levels in clinical practice.
Lane, representing the Polycystic Ovarian Syndrome Association of Australia (POSAA), presents an impassioned plea for more effective diagnosis and treatment of polycystic ovary syndrome (PCOS). Her concerns focus on the general lack of know-ledge about diagnostic criteria and the condition itself. We support the wider recognition of this condition, the greatest cause of infertility in this country. However, PCOS cannot be diagnosed or measured in any way by insulin levels, even though about 80% of patients are insulin resistant. Lane also calls for Australian guidelines for diagnosis of PCOS; these are not necessary, as simple, widely accepted international guidelines exist.1 As pointed out by Hutchison et al, estimates of insulin resistance are not required for diagnosis of PCOS.
PCOS is common and costly, both in absolute fiscal terms (eg, in-vitro fertilisation) and in quality of life and other, inestimable “human” terms. We recognise the great suffering of women with PCOS, and the heartbreaking difficulties of infertility, with its intrusive and expensive management. These factors make these women vulnerable as consumers, so it is important to inform POSAA and other consumer groups of useless measures that have no evidence base in diagnosis and treatment. Rightly, Lane expects medical practitioners and departments of health to agree on diagnostic guidelines for PCOS, and Hutchison et al anticipate the approval of metformin therapy in PCOS with clinical indications such as anovulatory infertility. The role of metformin in treating PCOS is not in dispute here. However, independent of symptoms or signs of PCOS, prescribing metformin after “diagnosis” of insulin resistance based on insulin levels is negligent. We agree unanimously with Hutchison et al that “use of metformin should be based on clinical indications”.
Kidson agrees that insulin measures are unreliable. His referenced comments highlight that insulin measures only have an evidence base in epidemiology. Again, we invite evidence for utility of measuring insulin levels in clinical practice, if it “can ever be presented”.
Tran points to the dominant role of the obesity epidemic, the overwhelmingly large elephant in the room we have thus far ignored. Obesity causes (and worsens) insulin resistance, and causes diabetes, heart disease, stroke and some cancers. With 60% of the adult Australian population now overweight or obese, we can expect a greater frequency of insulin resistance in the community. Tran presents a convincing, well researched argument against measuring insulin levels, either fasting or during an oral glucose tolerance test. Measures of central abdominal obesity (eg, waist circumference) have been shown in long-term studies to be the best predictors of heart disease, diabetes, cancer and all-cause mortality.
Any strategy that assists obese people to lose weight will reduce insulin resistance and components of the metabolic syndrome, particularly diabetes and heart disease. Motivating patients in lifestyle change is a difficult and perpetual challenge for the clinician. Nevertheless, we find it astonishing that clinicians use insulin levels to enhance motivation, as suggested by Strakosch. This is truly invoking phantoms. We encourage all clinicians in our difficult task of counselling and motivating lifestyle change. The creation of a facilitating environment to offset the Australian obesity and diabetes epidemic is a high political priority.2
We also thank Gale for emphasising the importance of lifestyle management in diabetes and related pre-diabetes conditions. However, his comment that the cost of measuring insulin levels can be justified is not supported by any evidence, and leads to a question as to who should bear the cost.
Who is bearing the cost of measuring insulin levels? This burden falls mainly on the Health Insurance Commission (HIC). If patients were made to bear the cost, they might demand greater clinician scrutiny of its validity. Is it appropriate for the Austra-lian taxpayer and the precious medical budget to fund an unvalidated and unreli-ably poor estimate of an entity that, by best practice, does not need to be measured? We acknowledge that insulin levels have a role in epidemiology and research — but only there, and the HIC has very clear guidelines that it is inappropriate to fund research through Medicare.
- 1. Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril 2004; 81: 19-25.
- 2. Zimmet PZ, James WPT. The unstoppable Australian obesity and diabetes juggernaut. What should politicians do [editorial]? Med J Aust 2006; 185: 187-188. <MJA full text>
Most of the authors have received ad-hoc honoraria for delivering lectures on their research or clinical interests at general practitioner or specialist educational meetings; some have also received travel assistance to attend international scientific meetings.