New and clinically useful markers of risk are of essence in high risk states such as type 2 diabetes mellitus. The major cause of death in type 2 diabetes mellitus is cardiovascular disease. One of the strongest predictors of such manifestations of arteriosclerosis is the pulse wave velocity in aorta, which is a non-invasive measure of the degree of arterial calcification [1]. Yet another non-invasive estimation of the degree of vascular disease can be done by evaluating the thickness of the intima and media of carotid arteries by ultrasound [2]. Both these clinical physiological examinations are non-invasive and hence risk-free to perform. However, both these measures of risk require skilled personnel and the equipment needed also carries a considerable cost. Leptin is a hormone that is secreted by adipose tissue that regulates metabolism and reduces appetite. Levels of leptin in humans are log-linearly increased in relation to obesity [3] and lack of leptin induces severe obesity, demonstrating that leptin signaling is necessary for normal feedback control of body weight [4]. The feedback information to the central nervous system regarding adipose tissue size by leptin is influenced by so called leptin resistance [5] which likely is a reason for poor effect to reduce body weight with recombinant leptin in humans [6]. Potential adverse effects of this hormone includes increased activity of the sympathetic nervous system and of the adrenal glands to secrete corticosteroids [7] and leptin thus not only increases metabolic rate but it also elevates blood pressure according to experimental studies [8-11]. Leptin also has immune-modulatory effects and it has been suggested that this hormone links obesity with inflammation and hence with cardiovascular disease by such mechanisms [12-14].

Since leptin apparently interact with many of the features of cardiovascular risk, such as blood pressure and inflammation, it might be a useful marker of prognosis that gives information that is independent of a cluster of classical clinical risk factors. Thus we prospectively studied leptin levels in relation to ischemic heart disease in middle aged patients with type 2 diabetes mellitus. In summary we found that levels of leptin were positively related to the hazard ratio (HR) of ischemic heart disease in both men and women and this was independently of age, HbA1c, BMI, systolic blood pressure and LDL-cholesterol/HDL-cholesterol ratio.

We analyzed baseline data from 244 women and 476 men who participated in a community-based cohort, CArdiovascular Risk factors in patients with DIabetes – a Prospective study in Primary care, acronym CARDIPP, in which serum leptin was measured at the baseline investigation. This prospective observational study was launched in November 2005 and recruitment of the participants was completed in 2008. The patients all had type 2 diabetes mellitus and were between 55–66 years old when they were consecutively recruited from 22 different primary health care centers in the counties of Östergötland and Jönköping, Sweden. Detailed information about the structure and results from CARDIPP has been described previously [15]. The centres were located in different demographic areas and differed in size. However, the model of treatment and care of diabetes was organized similarly and all the primary care centres adhered to the same national guidelines of diabetes care. The size of the cohort and age of participants was chosen to allow collection of endpoint data within 6 years, when assuming similar incidence of cardiovascular disease as in other cohorts of diabetes in Sweden [16].

We found that high levels of leptin were related to increased risks of ischemic heart disease independently of traditional risk factors in men and that this prognostic information also remained significant when carotid IMT and aortic PWV were both included in the Cox regression. Importantly these analyses included BMI as a measure of obesity and this implies that leptin gave additional information independently of the degree of adiposity. This is in line with the idea that leptin resistance, an insensitivity to feedback signaling of adipose stores to the CNS, is linked with increased risk for ischemic heart disease in patients [17]. However, since our study did not include an intervention aimed to affect leptin levels, this does not provide definite proof of causality. Elevated levels of CRP have been suggested to participate in induction of leptin resistance by direct interaction with the leptin molecule [18]. The positive correlation between leptin and CRP in both men and in women in our study is in line with such a theory. Also leptin is a regulator of the sympathetic nervous system with effects that, according to experimental studies, include increases of metabolic rate, angiogenesis and arterial tone [9]. This could theoretically have impact on ischemic heart disease incidence that goes beyond many traditional markers of cardiovascular risk, including BMI [19] which is in line with our findings in men with type 2 diabetes mellitus. High circulating leptin levels have earlier been reported to be associated with diabetes incidence and also with development of other components of the metabolic syndrome [20]. However, leptin levels were not found to be related to cardiovascular morbidity and mortality in a study of non-diabetic elderly people [21].

In a study of patients with type 2 diabetes mellitus by Brennan et al. leptin levels were reported not to be related to cardiovascular disease in 1194 women that were followed for 12 years [22]. This finding was in contrast to our data which do support prognostic use of leptin in serum when added to traditionally used risk factors, to estimate risk for ischemic heart disease in women. However, statistical significance was lost in the women in our trial when IMT and PWV were included in the equations. It is possible that less efficient treatment of complications in diabetes might have obscured specific actions of leptin, not acting through traditional risk factors, in the study by Brennan et al., since data were analyzed during 12 years up to the year 2002 [22], i.e. ending 3 years before we made our baseline investigation. In line with this LDL-cholesterol levels were 3.4-3.7 mmol/l across quintiles of leptin in the study by Brennan et al. while corresponding mean value was 2.77 in the women in the study presented herein. This is suggestive of more efficient lipid lowering therapy in our study which has been a trend in treatment of type 2 diabetes mellitus in general during the last decade [23]. It is also possible that age of participants is crucial in analyses of the risks related to leptin since at higher age obesity is not strongly linked with morbidity and mortality. Indeed, the term “obesity paradox” has been applied to the phenomenon that moderate obesity might even be beneficial in older age [24]. Indeed, better prognosis in obese as compared with leaner subjects in patients with diabetes > 65 years of age has been demonstrated [25] and patients in the study by Brennan et al. became older during follow up [22] than those in our cohort.

In males the relationship between leptin and IHD was independent of traditional risk factors which was in line with earlier trials in men with high risk but not necessarily restricted to type 2 diabetes [26-29]. However, a study of patients which also included a large proportion of patients with diabetes and renal impairment did not find an independent relationship with cardiovascular events of leptin when analyzed among many other cytokines [30]. We acknowledge that we had no information about cytokines such as adiponectin or macrophage migration inhibitory factor as in the paper by Schöttker et al. [30] but we did find leptin to be independently related to ischemic heart disease also in combination with two specific measures of arteriosclerosis, IMT and PWV. This might seem contradictory if leptin indeed mainly relates to arteriosclerotic disease per se. Potential explanation could be that PWV and IMT do not capture all information about true burden of arteriosclerosis. Another potential explanation is that although IMT and PWV together carry much information about arteriosclerosis in large vessels, we do not gain much knowledge from these measures of risk for plaque rupture which is the main cause of acute ischemic heart disease and which is to a large extent dependent on whether a plaque is vulnerable or stable [31]. Indeed, leptin has been found to be related to endothelial dysfunction in diabetes [32].

Strengths in our study include use of registry data with high quality and in terms of mortality and morbidity incidence no lost to follow-up. Weaknesses include a rather short follow-up time and consequently low rates of mortality and morbidity. However the relatively short follow-up period suggests clinical relevance insofar that patients were indeed being cared for in a manner that is in accordance with current standards. This was confirmed by frequent treatment with statins and antihypertensive medications, and levels of HbA1c and blood pressures were also generally close to targets recommended by guidelines. Indeed, it is possible that our finding of an association between circulating leptin and risk for ischemic heart disease was statistically significant after correction for traditional risk factors in men and women was made possible by reasonably good control of glycaemia, blood pressure and cholesterol that increased the chance to detect the risk related leptin, that might act through other mechanisms than mediated by the traditional risk factors that are known to the provider of care. The positive relationship between serum leptin levels and CRP, confirmed in our study, is an example of such a risk factor that is not targeted in routine clinic care, as of yet.

This prospective observational data support the use of serum leptin for estimation of risk for ischemic heart disease independently of traditional risk factors. In the men serum leptin also provided information about risk that was independent of two strong non-invasive markers of cardiovascular disease, carotid IMT and carotid-femoral PWV. The results of our study support further assessment in larger cohorts of the role of measuring leptin levels as a predictive marker if ischemic heart disease in the care of patients with type 2 diabetes.