Gender aspects in type 2 diabetes mellitus and cardiometabolic risk

Men are well known to have a higher risk than women for cardiovascular disease. In recent years, however, studies show adult men also have higher risk for type 2 diabetes, an observation which has important clinical implications, particularly in the public health arena. This chapter explores the relevant data underlying this observation, examines potential mechanisms including life course changes in insulin resistance and role of adiposity, and discusses relevant clinical implications and solutions.

Introduction

It has become apparent that middle-aged men are at significantly higher risk of type 2 diabetes than women in several different populations.1, *2, *3 For example, whilst the prevalence of diabetes was over 19% in men over 50 years of age by 2005 in a population-based study in Ontario, Canada, the corresponding prevalence for women was just under 16% [Fig. 1].¹ Of interest, this gender difference in type 2 diabetes was not apparent in the 20–49 year old groups, and only emerged in middle-aged Canadians. Similarly, a recent Korean study reporting data from 2005 showed type 2 diabetes prevalence in above 30 year olds to be around 7.9% in women but 10.2% in men with the biggest differences in prevalence's apparent in the 40–59 year old brackets, where, remarkably, male diabetes prevalence was around double that in females.² Likewise, higher males diabetes rates have been noted in many studies around the world, as summarised in a recent meta-analysis paper.³ Thus, it is clear type 2 diabetes is more common in middle aged men than middle aged women, but further research would be useful in other ethnicities to extend findings.

In a recent paper,⁴ we explored the possibility that men have to gain less weight to develop type 2 diabetes than do women. To do so, we examined the relationship between age and BMI at diagnosis among 51,920 men and 43,137 women included in a population-based diabetes register in Scotland for whom an index BMI measurement was taken within 1 year of diabetes diagnosis. The findings confirmed the original hypothesis and, as shown in Fig. 2, demonstrated that men do indeed need to attain a lower average BMI to be diagnosed with type 2 diabetes than do women. Mean BMI closest to date of diagnosis of type 2 diabetes mellitus was 31.83 kg/m² in men and 33.69 kg/m² in women. However, it was notable that mean BMI difference by gender was most marked at younger ages and narrowed with advancing age. Furthermore, as HbA1c levels within 1 year of diagnoses were broadly similar in men and women, we were able to confirm that men were not being diagnosed any later in the course of their disease than women. These Scottish observations⁴ were quickly confirmed by analysis of data from the UK General Practice Research Database (GPRD).⁵ In this latter study, which included data from within 6 months of diabetes diagnosis, the age-adjusted average BMI at diagnosis was higher in women by 1.8 (95% CI 1.7, 1.9) kg/m² (p < 0.01), almost identical to Scottish findings. Furthermore, both studies confirmed a pattern of lower BMIs at diagnosis with advancing age.*4, 5

Linked to the above observations, we wondered whether the greater increase in adiposity in women needed to develop diabetes may be relevant to their greater relative increase in cardiovascular risk compared to men. This fact was most recently confirmed in the Emerging Risk Factor Collaboration (ERFC) analysis,³ in which the Hazard Ratio (HR) for cardiovascular (CVD) disease in women with type 2 diabetes (compared to women without) was 2.59 (95% CI 2.29, 2.93) whereas the comparable HR was lower at 1.89 (1.73, 2.06) in men. To test our hypothesis, risk factor patterns in men and women with and without diabetes were examined in two large studies of older adults, the British Regional Heart Study (NRHS) and British Women's Health and Heart Study (BWHHS).⁶ We confirmed that type 2 diabetes was associated with increased CVD risk factors in both genders, but diabetic women showed greater relative increase in many CVD risk factors than diabetic men. Moreover, we showed that these greater risk factor changes were to some or a large extent explained by greater increase in adiposity and insulin resistance associated with type 2 diabetes in women than in men.⁶ We therefore proposed the hypothesis that women have to undergo greater metabolic deterioration to develop type 2 diabetes than do men, and as such, many insulin resistance risk factors must change to a greater extent. This observation must therefore at least partially explain the increased relative risk of CVD in women with diabetes compared with men. As recently discussed,⁷ the clinical ramifications of this observation are also important and suggest that women with type 2 diabetes should receive similar risk factor treatment as men for prevention of CVD.

Fundamentally in all individuals, regardless of ethnicities, excess weight is the strongest risk factor for type 2 diabetes and, accordingly, body mass index (BMI) forms a major aspect of all risk factor scores for diabetes. When individuals gain weight they become more insulin resistant but do so at different rates or at different BMI thresholds dependent upon several factors such as their gender (as shown above), ethnicity and family histories of diabetes. In terms of gender, adult men are more insulin resistant than women,⁸ and clinically, many women can remain highly insulin sensitive despite considerable weight gain as they appear to have an excellent ability to expand (safer) subcutaneous far stores.⁹ By contrast, as subcutaneous storage capacity is lower in men (predominantly to do with reproductive needs of women driven, partially/predominantly, by differential sex hormone settings), then with weight gain excess fat is placed more rapidly into ‘other’ tissues in men. This so-called ‘ectopic fat’ appears to accumulate in many places including the intra-abdominal and peri-vascular cells, and critically into skeletal muscle, liver and possibly the pancreas. Rising waist circumference is a relatively simple clinical sign associated with ectopic fat gain, whilst rising liver enzymes, in particular ALT (more so than AST) and GGT, especially when seen in conjunction with parallel triglyceride changes, indicate liver fat gain.¹⁰ Ectopic fat, in turn, renders organs insulin resistant by mechanisms which might include rising tissue levels of metabolically toxic lipid derivatives such as ceramides or diglycerides or other products, which, in turn, interfere with insulin signalling pathways.¹¹ Interestingly, excess peri-vascular fat may also be relevant to diabetes risks by altering nutrient flow via altered ‘vasocrine’ signalling.¹²

In terms of risk factors, several parameters hint at men being more insulin resistant than women (Table 1). Middle-aged men generally have higher fasting glucose levels, higher triglyceride and lower HDL-cholesterol levels than women of a similar age even after adjusting for BMI.*3, 13 Fat distribution differs by sex and, in general, men have greater visceral and hepatic fat compared with women.⁸ Indeed, in keeping with greater baseline liver fat levels, normal ranges for the liver enzyme, ALT, a predictor of diabetes, is higher in men than women. Other liver associated markers associated with hepatic insulin resistance (e.g. SHBG) also display sexual dimorphism, with levels being much lower in men and related to diabetes risk in both genders.¹⁴ At the adipokine level, lower circulating adiponectin levels in men¹⁵ is consistent with their greater insulin resistance, whereas women tend to have higher leptin levels, consistent with their greater total percent fat mass.¹⁶ Whether adiponectin is a cause or consequence of greater insulin resistance in general remains a subject of debate, however.¹⁷ Going forwards, further research is needed to determine role of pancreatic and other fat depots in explaining gender differences in insulin resistance and diabetes, though clearly higher waist circumferences in men appear to be an obvious factor.

Whilst levels of type 2 diabetes are clearly greater in middle age men than women, the reverse pattern appears to be true in children. There is evidence (albeit predominately from studies on fasting insulin) that pre-pubertal girls (perhaps even from birth) are more insulin resistance than age-matched boys,¹⁸ and robust evidence that type 2 diabetes is more common in adolescent females than males.¹⁹ The latter observation is also in keeping with a greater hike in insulin resistance in girls during puberty than in boys.²⁰ However, beyond puberty, it is tempting to speculate that differential patterns of fat accumulation with greater subcutaneous fat capacity in women along with lower waist circumference and accompanying better insulin sensitivity, results in a reversal of the risk so that type 2 diabetes rates are higher in men.1, *2 Fig. 3 gives a pictorial representation of this change over the life course but further research would be useful to corroborate such a pattern, and extend it to other ethnicities.

There is clear evidence for greater diabetes risk in many ethnicities including South Asians and African Caribbeans relative to European whites, but research to define why this might be the case, and, in particular whether mechanisms for higher diabetes rates might vary by gender, remains sparse. Using the Southall And Brent REvisited (SABRE) cohort, a tri-ethnic prospective study, we demonstrated that insulin resistance (measured by HOMA) and truncal obesity accounted for the twofold excess incidence of diabetes in Indian Asian and African Caribbean women, but not men.²¹ Indeed, age-adjusted subhazard ratios remained around two for type 2 diabetes risk even after adjusting for insulin resistance and truncal obesity in South Asian and men African Caribbean men, suggesting alternative mechanisms (e.g. impaired beta cell function) may be at relevant. Clearly, given the rapidly rising type 2 diabetes rates around the world, particularly in developing countries and in certain ethnicities (i.e. South Asians, Arabs and Chinese), the need for further research in these areas is paramount.