Impact of HbA1c criterion on the definition of glycemic component of the metabolic syndrome: the China health and nutrition survey 2009

The metabolic syndrome (MetS), characterized by abdominal obesity, hypertriglyceridemia, low high-density lipoprotein cholesterol (HDL-C) level, high blood pressure (BP), and increased fasting glucose level, predisposes individuals to a high risk for cardiovascular disease (CVD) and diabetes [1–4]. The MetS is a common disorder in China. According to data from the International Collaborative Study of Cardiovascular Disease in ASIA, 64 (15.1%) million adults aged 35–74 years have the MetS according to the modified Adult Treatment Panel III (ATP III) criteria [5]. The ATP III criteria were slightly modified by lowering the threshold for fasting plasma glucose (FPG) to 5.6 mmol/l in 2004 by the American Heart Association/National Heart, Lung, and Blood Institute (AHA/NHLBI) [6, 7] to be consistent with the American Diabetes Association (ADA) criteria for impaired fasting glucose (IFG) [8]. Most recently, several major organizations, led by the International Diabetes Federation (IDF) and AHA/NHLBI, proposed a unified definition of MetS in 2009 [9], while the ADA recommended the use of glycated hemoglobin (HbA1c) as an alternative to IFG to define the category of increased diabetes risk in 2010 [10]. Studies from the USA and Europe showed that HbA1c can be used instead of FPG in identifying individuals with MetS. However, evidence has suggested that HbA1c value might differ according to ethnic origin [11]. Whether HbA1c can be used instead of FPG in the definition of MetS in Chinese population remains unknown. Hence, we took advantage of the large representative sample of Chinese adults who participated in the China Health and Nutrition Survey (CHNS) 2009 to compare the uses of HbA1c ≥ 5.7% or FPG ≥ 5.6 mmol/l in the definition of the glycemic component of MetS. In addition, as prior evidence provided mixed results as to whether the presence of MetS aggravates cardiovascular risk in diabetic patients [12–14], we repeated the analysis in participants without diabetes (FPG ≥7.0 mmol/l or HbA1c ≥6.5%).

In the present study, we compared HbA1c and FPG in the definition of glycemic component of MetS. We saw low magnitude of overlap and poor agreement between FPG- and HbA1c-based identification of MetS. Similar significant trends were noted in each evaluated subgroup. The prevalence of HbA1c-based diagnosis of MetS was significantly higher than FPG-based diagnosis of MetS. Screening MetS by introduction of the new HbA1c criterion in addition to assessment of the FPG could contribute to identification of more people with MetS.

Our results are in line with the study from an European country with respect to a higher prevalence of MetS measured by HbA1c than by FPG [17]. Evidence has suggested that there were a large proportion of Chinese with isolated postprandial hyperglycemia [18]. Although postprandial glucose were not available in the present study, it is possible that even with normal fasting glucose, individuals may have postprandial hyperglycemia. HbA1c represents chronic exposure to basal and postprandial hyperglycemia. Hence, HbA1c can identify more individuals with MetS than FPG.

In the present CHNS cohort, HbA1c-based diagnosis of MetS illustrated limited overlap with FPG-based diagnosis of MetS. This deviates from the findings of the cohort studies from United States and European countries in which a good agreement between HbA1c- and FPG-based identification of MetS was observed [17, 19]. Hence, a discrepancy in the degree of overlap might be more likely to happen in Chinese than in US and European adults. The inconsistency between the cited studies and our present study may be attributed to the discrepancies in ethnicity, socio-demographic or personal characteristics, prevalence of CVD risk factors such as hypertension, obesity, hypertriglyceridemia, and lifestyle differences. Another possible explanation for the inconsistency might be due to the fact that FPG differentially correlated with HbA1c among different ethnic populations as several recent reports have compared the relationship of HbA1c to FPG in various populations and have found significant discordance [11, 20]. HbA1c correlates better with insulin resistance than FPG [21]. To date, differences in the relation of FPG and HbA1c to insulin resistance and insulin secretion have been little described in Chinese population. One study conducted in Japanese population indicated that high normal HbA1c levels (5.4–5.8%) were associated with impaired insulin secretion without marked insulin resistance and elevated HbA1c levels (≥5.9%) were associated with substantial impairment in insulin secretion, insulin sensitivity and β-cell dysfunction [21], while one Chinese population based study suggested that increased FPG is predominately induced by the decline in insulin sensitivity and insulin secretion, with the insulin secretion more pronounced [22]. Taken together, it is possible that how well HbA1c performs as compared with FPG for the identification of MetS will depend on the target population.

The limited overlap between HbA1c- and FPG-based diagnosis of MetS highlights different facets of glucose metabolism and multifactorial pathophysiologic mechanisms for glucose dysfunction [23–25]. In contrast to daily pre-prandial glucose snapshot offered by FPG, HbA1c captures chronic hyperglycemia, including postprandial glucose spikes. Increased FPG is predominately induced by liver insulin resistance and a defect in the early phase of insulin secretion [23, 25], whereas increased HbA1c is dominated by a combination of hepatic insulin resistance, muscle insulin resistance and impaired insulin secretion [24, 25]. Although the explanations for the significantly higher proportion of HbA1c-based diagnosis of MetS in women compared with men observed in the present study remain to be elucidated, it is probably related to gender differences in glucose absorption patterns [26]. The difference in gender-specific glucose absorption pattern is a consequence of the relatively higher dose of glucose given to women compared with men when seen in relation to their body size [27, 28]. Several studies have shown that men in general have a higher prevalence of isolated impaired fasting glucose than do women [27, 28]. In contrast, women often exhibit a higher prevalence of isolated impaired glucose tolerance [27, 28]. Monnier et al. found that the relative contribution of postprandial hyperglycemia accounted for ~70% of overall glycemic exposure in patients with HbA1c <7.3%, while the relative contribution of FPG increased gradually with increasing levels of HbA1c [29]. In the present study, we noted that only 8 (0.9%) individuals had HbA1c ≥7.3% among the 872 individuals with HbA1c-based definition of MetS (data not shown), indicating a greater contribution of postprandial versus FPG to HbA1c levels. Hence, the HbA1c-based definition of MetS reflects more of the postprandial glucose-based diagnosis of MetS. Taken together, it is possible that women carry a higher prevalence of HbA1c-based diagnosis of MetS than do men. A greater contribution of postprandial versus FPG to HbA1c levels may also in part explain the limited overlap between FPG and HbA1c-based definition of MetS.

Subgroup differences were noted in the present study, with a slightly higher magnitude of overlap between FPG and HbA1c-based definition of MetS among persons with older age, elevated BMI, WC, and BP. One possible explanation for this relatively low discrepancy in older age might be the fact that the prevalence of MetS was more frequent among older persons [5, 30]. The relatively low discrepancy in worsening BMI, WC, and BP status might be attributed to the fact that obesity, enlarged waist, and hypertension accompany the insulin-resistant state that underlies the MetS.

In this study, individuals with FPG-based identification of MetS had different characteristics compared with individuals with HbA1c-based identification of MetS. The different characteristics might affect progression to CVD mortality and morbidity within each case of discordantly diagnosed MetS. Most components of the MetS are required to be conducted in the fasting state, although an overnight fast is not required for HbA1c measurement, it is possible that clinicians can have access to both FPG and HbA1c in most of their patients. In this context, and based on our data, the MetS rate may increase by approximately 70% if both FPG and HbA1c are measured. Although there is controversy whether the diagnosis of MetS conveys additional cardiovascular risk in subjects with diabetes [12–14], a recent meta-analysis evidenced that in the absence of type 2 diabetes mellitus, the MetS was associated with an increased risk of cardiovascular mortality (RR: 1.75; 95% CI: 1.19 to 2.58) [3]. Our present study showed that in participants without diabetes, even more limited overlap between HbA1c- and FPG-based diagnosis of MetS was observed. The identification of MetS is of clinical importance because appropriate interventions can clearly prevent or delay the CVD mortality and morbidity, especially in subjects without diabetes. Although evidence suggested that there was no linear relationship between FPG and CVD mortality and morbidity [31], the risk for diabetes and CVD mortality and morbidity increased gradually with rising FPG and HbA1c [32, 33]. Two prospective studies demonstrated that the combined use of FPG and HbA1c were more predictive of the risk of future diabetes [34, 35]. Our finding that individuals with Mets based on both FPG and HbA1c were older, more obese, hypertensive, dyslipidemic, and insulin resistant and had higher CRP compared with individuals with Mets based on FPG alone or HbA1c alone further supported the notion. Therefore, individuals with Mets based on both measure might allow for assessment of a substantially increased risk of CVD mortality and morbidity. We believe that introduction of HbA1c into FPG is appropriate for detection of MetS, and will contribute to cover a larger number of people being submitted to aggressive drug regimens to prevent future risk for diabetes and CVD mortality and morbidity. Further large-scale longitudinal studies are, however, required to address this issue.

Our study has several strengths including a well-established cohort of nationally representative sample of the Chinese adult population, a vigorous quality assurance program, the same strict methodology used to ensure the quality of the data collection over the entire study period, and the centralization of laboratory measurements. We also benefited from the CHNS 2009′s availability of broad range of demographic, clinical, and biomarker data collected by trained staff. Nevertheless, the current study was subjected to several limitations. First, we got availability of only single measures of all biochemical variables, a common limitation to most large epidemiologic studies. Second, since the current study is a cross-sectional design, we cannot explore whether diagnosis of MetS based on different measures have the same predictive capacity in the risk of future diabetes and CVD mortality and morbidity. Third, estimates across subgroups should also be interpreted with caution because of limited sample size.

In summary, limited overlap and poor agreement are demonstrated between the FPG- and HbA1c-based diagnosis of MetS. Screening MetS through introduction of HbA1c in addition to FPG could contribute to identification of more people with MetS who would otherwise have been missed.