Diabet. Med. 28, 1436–1437 (2011)

Seventy-five years ago The Lancet published a manuscript by Sir Harold Himsworth entitled ‘Diabetes mellitus: its differentiation into insulin-sensitive and insulin-insensitive types’ [1]. He described a functional test ‘for distinguishing these two types of diabetes’ and concluded that ‘the insulin-sensitive type appears to be caused by a deficiency of insulin’, whereas the insulin-insensitive type is ‘apparently due not to a lack of insulin’. The phrase ‘a seminal contribution’ has become a cliché, but if ever a scientific contribution merited that accolade, it is this publication.

Despite Himsworth’s standing as a pre-eminent clinical scientist, and his publications over the next few years in the world’s most distinguished medical journals, his findings seemed to have had little long-term impact. In 1970, we published what we believed to be the first experimental method to directly quantify insulin-mediated glucose, leading to the conclusion that insulin resistance was a characteristic finding in patients with glucose intolerance [2]. At that time, conventional wisdom was that diabetes was another example of a hormone-deficiency disease. We were unaware of Himsworth’s studies showing that there was more than one form of diabetes and that the majority of patients belonged to the insulin-insensitive form. Indeed, based upon the reception our findings received, ignorance was not confined to our research group. It took approximately another decade before the importance of insulin resistance in the pathogenesis of Type 2 diabetes was no longer questioned and the importance of insulin resistance in the development of many clinical syndromes no longer ignored [3]. This tribute to the 75th anniversary of the first demonstration that insulin resistance played a role in the pathogenesis of a clinical syndrome will include manuscripts focusing on the association between insulin resistance and aspects of polycystic ovary syndrome, obstructive sleep apnoea and cognitive dysfunction. In addition, as what made the contributions of Himsworth so credible was the ingenuity of his experimental methods, approaches currently used to study the link between insulin resistance and pathophysiology will be evaluated. Finally, Himsworth ended his studies of diabetes by addressing possible mechanisms accounting for insulin resistance and the last contribution summarizes the information we have gained in this context from genetic studies.

I am embarrassed to confess that my awareness of the incredible series of manuscripts concerning the pathophysiology of diabetes mellitus published by Himsworth only occurred approximately 40 years later, and was a function of my concern over the recommendation by the American Diabetes Association (ADA) that patients with diabetes should follow a low-fat/high-carbohydrate diet to decrease risk of cardiovascular disease. One of the putative benefits of this dietary approach was purported to be its ability to enhance insulin sensitivity, citing manuscripts published in the 1930s by Himsworth, an author unknown to me at the time. As this point of view was in conflict with my own research findings, I dug these papers out of the archives of our library and, much to my delight, found a series of incredible publications, as well as evidence that his findings had been somewhat misinterpreted. At the simplest level, Himsworth pointed out that ‘with progressive restriction of dietary carbohydrate there is progressive impairment of sugar tolerance’ [4], accounting for why as a medical house-officer I was told not to schedule a glucose tolerance test unless the subject had an adequate carbohydrate intake for at least the preceding 3 days.

More to the point of the low-fat/high-carbohydrate diet advocacy of the American Diabetes Association, results published in 1939 [5] suggest that, although such a diet might make sense for Type 1 diabetes (insulin-sensitive), it does not seem to be the best choice for those with Type 2 diabetes (insulin-insensitive). To begin with, ‘increasing the dietary carbohydrate from 150 g to 200 g resulted in only an 8% improvement in sensitivity to insulin or in glucose tolerance’ in healthy persons without diabetes. The effects of increasing carbohydrate intake several-fold in patients with diabetes varied as a function of their differences in insulin action. Thus, this manipulation had little or no effect on fasting glucose concentration, but increased insulin action in insulin-sensitive patients. In contrast, similar increases in carbohydrate intake in patients with insulin-insensitive diabetes were associated with higher fasting glucose concentrations, without any improvement in insulin action. These observations hardly seem to support the enthusiasm that the American Diabetes Association seems to have for low-fat/high-carbohydrate diets in patients with Type 2 diabetes.

Having provided substantial evidence that there were at least two forms of diabetes, Himsworth and colleagues built on the findings of Houssay [6] that ‘anterior pituitary lobe extract counteracts the action of insulin’, by studying the effect of a pituitary extract (glycotropic factor) that seemed to be free of thyrotropic or gonadotropic effects on insulin action in rabbits [7]. The results of these studies indicated that, under conditions that excluded the liver from the circulation, ‘the pituitary glycotropic factor inhibits the action of insulin in accelerating the removal of blood glucose by the peripheral tissues’.

Himsworth’s valedictory comments on the syndrome of diabetes mellitus were published in The Lancet in 1949 [8]. The final paragraph in that manuscript states, ‘It thus appears that we should accustom ourselves to the idea that a primary deficiency of insulin is only one, and then not the commonest cause of the diabetic syndrome’. I feel compelled to insert a personal, and somewhat disturbing, comment at this point. I entered medical school in 1949, and at no time in medical school, internship, medical residency or endocrine fellowship, was I introduced to the work of Himsworth and colleagues; diabetes was one disease and had only one cause—insulin deficiency.

The 1949 publication is a masterpiece, worth reading by anyone wanting to be exposed to the thinking of an incredible clinical investigator. I hope the following three extended quotes may whet someone’s appetite.

  • 1
     ‘Broadly speaking, patients giving the insulin sensitive response to the insulin-glucose test tend to be young and thin with healthy arteries, and normal blood pressure; in them the disease is severe and of severe onset but readily responsive to insulin; they readily develop ketosis, and slight excess of insulin produced hypoglycaemia. The type with the insensitive response tends to be older and obese, with arteriosclerosis and a high blood pressure, in him the disease is milder, of insidious onset, and often surprisingly insusceptible to insulin; ketosis rarely develops, and he can tolerance excess of insulin without symptoms.’

I was taught the same clinical description of juvenile-onset and maturity-onset diabetes in medical school, but without any consideration of the aetiological importance of the differences in insulin sensitivity.

  • 2
     ‘But when the insulin sensitivity of normal peoples is tested, it is found that intermediate degrees occur in the apparently healthy and are found with increasing frequency in the older age-groups. It appears that insensitivity to insulin must exceed a certain degree for diabetes to develop. As a consequence all insensitive diabetics have high degrees of insensitivity.’

Can it be said any more clearly ∼60 years later?

  • 3
     ‘But, most striking of all, reducing the patient’s weight by any dietary means not only removes the symptoms and signs of diabetes, but also restores the sugar tolerance curve to normal; and this result is all the more remarkable because it is achieved by the very measure—semi-starvation—which impairs carbohydrate tolerance and insulin sensitivity in normal people. In the sense that there is no detectable abnormality of carbohydrate metabolism—as long as obesity is prevented, such cases can be regarded as cured.’

Notice the phrase ‘by any dietary means’; consistent with all of the many recent studies comparing diets of different macronutrient content.

In conclusion, I plead ignorance to being aware of Himsworth’s contributions before I began my own efforts, but not to hyperbole when I contemplate his scientific achievements. He was truly ‘a man for all seasons’. Hip, hip, hooray!

Competing interests

  1. Top of page
  2. Competing interests
  3. References

Nothing to declare.


  1. Top of page
  2. Competing interests
  3. References
  • 1
    Himsworth HP. Diabetes mellitus: its differentiation into insulin sensitive and insulin insensitive types. Lancet 1936; 1: 127130.
  • 2
    Shen S-W, Reaven GM, Farquhar JW. Comparison of impedance to insulin-mediated glucose uptake in normal and diabetic subjects. J Clin Invest 1970; 49: 21512160.
  • 3
    Reaven GM. Why syndrome X? From Harold Himsworth to the insulin-resistance syndrome Cell Metab 2005; 1: 914.
  • 4
    Himsworth HP. On the mechanism of diabetes mellitus. Lancet 1939; 2: 155.
  • 5
    Himsworth HP, Kerr RB. Insulin-sensitive and insulin-insensitive types of diabetes mellitus. Clin Sci 1939; 4: 119152.
  • 6
    Houssay BA. Carbohydrate metabolism. N Eng J Med 1936; 214: 971986.
  • 7
    Himsworth HP, Scott DBM. The action of Young’s glycotropic factor of the anterior pituitary gland. J Physiol 1938; 92: 183207.
  • 8
    Himsworth HP. The syndrome of diabetes mellitus and its causes. Lancet 1949; 2: 467473.