Results demonstrate that breast cancer survivors, who are both centrally obese and insulin resistant, are characterised by greater metabolic dysfunction in terms of dyslipidemia, glucose dysmetabolism, raised CRP and a higher prevalence of the MetSyn compared to women who are centrally obese in the absence of insulin resistance. This altered metabolic profile was accompanied by higher absolute and adjusted RMR levels but was not associated with differing habitual physical activity levels. These results are consistent with others which have demonstrated deteriorations in lipid-lipoprotein profile, inflammatory status and measures of glucose-insulin homeostasis when central obesity exists concurrently with insulin resistance [31, 32]. In these studies, participants who were centrally obese and not insulin resistant were characterised by metabolic profiles that were similar to lean controls.
In the present study inflammatory status was measured by CRP, an acute-phase reactant secreted from the liver in response to pro-inflammatory cytokines (TNF-α and IL-6). Pro-inflammatory cytokines are secreted from the visceral adipose tissue in response to infiltration by macrophages to expanding adipose tissue and act on the liver to stimulate gluconeogenesis, resulting in increased production of hepatic glucose and consequently increased pancreatic secretion of insulin leading to the development of insulin resistance . CRP levels >3 mg.L-1 are indicative of chronic low-grade inflammation . Therefore in the current study, CRP values suggestive of low-grade inflammation (4.98 (3.46) mg.L-1) were observed in those who were insulin resistant. These values are comparable to those reported by Thomson and colleagues  who described a mean high sensitivity CRP of 5.1 (5.3) mg.L-1 in an overweight cohort of breast cancer survivors. Interestingly, in the present study mean CRP values did not exceed the 3 mg.L-1 cut-point in any of the non-insulin resistant groups. Among breast cancer survivors, significant associations have been documented between acute inflammatory markers (CRP and serum amyloid A (SAA)) and a number of anthropometric and lifestyle related prognostic indicators including BMI, increasing waist circumference and decreasing physical activity levels  in addition to all-cause mortality following breast cancer .
Chronic inflammation may also adversely influence lipid-lipoprotein profiles resulting in dyslipidemia . Consequently, in the current study HDL-C was significantly lower and TG significantly higher in the insulin resistant group compared to all other groups. There is a lack of consensus regarding the presentation of dyslipidemia in obese, insulin resistant adults. Nieves and colleagues  reported in insulin resistant, non-obese individuals, that the presentation of dyslipidemia could be explained by differing levels of visceral fat accumulation. In contrast, Piché and colleagues  suggested that variations in the presentation of dyslipidemia in centrally obese adults were driven by insulin resistance. In the current study, only those who were insulin resistant showed evidence of dyslipidemia, however causality cannot be inferred.
It should be acknowledged that the HOMA-IR scores observed in this study (mean 2.09 (range: 0.47 – 6.36) were lower than observed in similar cohorts for example the HEAL study which reported mean HOMA-IR scores of 2.55 (range: 0.25 – 40.16) . The HEAL study is a multicentre, multiethnic trial based in the United States of America. To the authors’ knowledge, the current study is the first to examine metabolic characteristics in an Irish cohort of breast cancer survivors. One of the limitations of this cohort is a lack of diversity variables such as BMI scores. The mean BMI of participants in the current study was 27.02 (range 19.7-39.9) kg.m-2 compared to 27.3 (range 16.2 – 53.3) kg.m-2 in those involved in HEAL . While the mean values are comparable, the range and maximum values were lower in the current study. As obesity is the predominant underlying cause of insulin resistance , the lower HOMA-IR scores observed in the current study may be attributed to lower range of BMI values. The prevalence of obesity in Ireland is increasing steadily , albeit at a slower rate than observed in North America . Nonetheless, the metabolic and physiological health consequences of obesity in Ireland are becoming increasingly evident and there is a need for intervention to prevent further decline.
A relationship between physical activity and metabolic markers was not observed in the current study. This is in contrast other studies in breast cancer survivors who reported correlations between physical activity and several metabolic variables including fasting insulin , HOMA-IR , CRP , leptin , IGF-I  and IGFBP-3 . Furthermore habitual physical activity levels did not vary between groups. Sedentary behaviour was prevalent across each of these groups representing 50% (Range: 44% in Group 1 to 54% in Group 3) of waking hours. In contrast, moderate and vigorous intensity activity accounted for only 7% of waking hours (Range: 5% in Group 3 to 10% in Group 1) with less than one-third of any group achieving recommended physical activity guidelines. It could be suggested that the high percentage of sedentariness made it difficult to identify differences between equally sedentary groups in this small sample. These results, like others [7, 28], represent the extent of the problem of inactivity among breast cancer survivors and may also represent the problem of inactivity at population level. Recent reports suggest that 31% of adults globally are inactive (not achieving recommended physical activity guidelines), with inactivity levels highest in Europe and America and greatest amongst women . Excess sedentary behaviour is particularly concerning among breast cancer survivors however due to the association between physical inactivity and prognosis. Physical activity undertaken following breast cancer is associated with up to a 35% reduction in all-cause mortality [19, 40] while inactivity has been linked to a four-fold increase in mortality risk .
RMR comprises the largest component of energy expenditure and therefore is an essential determinant of energy balance. Mean RMR levels (1297.53 (203.13) jK.day-1 (equivalent to 1297.53 (203.13) kcal.day-1 in the current study were comparable to those reported previously in a similar cohort (1300 (183.9) kcal.day-1) . In addition, both absolute and adjusted RMR were significantly higher in the insulin resistant group compared to all others and demonstrated moderate-to-weak correlations with a number of metabolic variables. These results are consistent with others which report that adjusted RMR is typically increased in individuals with T2DM relative to healthy counterparts [42, 43]. Weyer and colleagues  suggested that increases in RMR occur early in the development of T2DM in conjunction with the development of progressive metabolic abnormalities including an increase in gluconeogenesis and increased hepatic glucose productions. In contrast, there is a lack of consensus as to whether RMR is increased or decreased in individuals with clinically defined MetSyn compared to metabolically healthy counterparts [44, 45]. Despite the exclusion of individuals with T2DM from the current study, results are consistent with the development of underlying glucose dysregulation. In contrast, non-insulin resistant groups with lower RMR values may be at risk of continued weight gain .
The aim of the current study was to identify known modifiable prognostic risk factors that may respond favourably to intervention to improve breast cancer survival. This evaluation did not compare findings to a matched control group however it is highly likely that the presented metabolic characteristics are comparable to population norms. Population statistics suggest that “normal” is characterised by obesity , inactivity  and sedentariness ; factors which are all associated with reduced breast cancer outcome and will contribute to metabolic decline. The challenge therefore with rehabilitation in the cancer context is the need for cancer survivors to adopt health promoting lifestyle behaviours which may be different from pre-diagnosis levels. Rehabilitation which focuses on a return to “normal” function or baseline levels is no longer adequate. Therefore it is suggested that metabolic screening to identify key modifiable risk factors and inform exercise prescription in this cohort may be more valuable than comparison to a matched control group.