Can neck circumference predict insulin resistance in older people ? A cross-sectional study at primary care in Brazil

This study focused on the correlation between neck circumference and other anthropometric measurements and cardiovascular risk factors and insulin resistance in the elderly in Niterói, Rio de Janeiro State, Brazil. This was a cross-sectional study in 411 patients seen in primary care. Anthropometric measurements including neck circumference, body mass index (BMI), waist circumference (WC), and waist-hip ratio (WHR) were recorded. Bioelectrical impedance analysis (BIA) estimated the percentage of total body fat (%TBF). Insulin resistance was estimated by HOMA-IR (homeostasis model assessment IR index). The data were analyzed by the Mann-Whitney test and the correlation between numerical variables by Spearman’s test. Receiver operating characteristic (ROC) was used to assess the anthropometric measurements’ predictive capacity to diagnosis insulin resistance. Neck circumference was positively correlated with WC, BMI, WHR, %TBF, and HOMA-IR in both sexes. In women, neck circumference showed higher AUC (area under the curve) for insulin resistance. In men, WC showed higher AUC, followed by BMI and neck circumference. The current study in Southeast Brazil suggests that neck circumference can predict insulin resistance, an important marker of cardiovascular risk in the elderly population treated in primary care.


Introduction
Obesity has become epidemic all over the world, which is a cause of concern, since it is a public health problem related to cardiovascular diseases, metabolic syndrome, cancer, and other comorbidities.Moreover, in older persons, obesity can exacerbate the age-related decline in physical function and lead to frailty 1 .
Upper-body distribution of fat is more predictive of metabolic and cardiovascular diseases than whole body obesity.Body mass index (BMI), a traditional anthropometric marker used in the majority of the epidemiological studies 2,3,4 , is not the best method to evaluate central obesity 5,6,7 .Computed tomography is the most accurate method to measure body fat, specifically visceral fat (VF), however, high cost and radiation exposure limit its use in clinical practice 8 .
After the age of 60, modifications in the body began to be noticed, such as composition with lean mass reduction and increase of fat mass, which tends to be redistributed with advancing age toward more VF 9 .As a result, the anthropometric measures of central adiposity, such as BMI or abdominal circumference may not be accurate in older people 10 .Therefore, is of great interest to find a simple and practical anthropometric measures of central obesity for that group.Neck circumference (NC), a simple and alternative anthropometric measure, was proposed as an index of central obesity.NC is associated with other cardiovascular risk factors, such as dyslipidemia, arterial hypertension, hyperuricemia, and insulin resistance (IR) 11,12,13 .
To date, there are no published epidemiological data on the clinical significance of NC among individuals over 60 years old in Brazil.
Therefore, the aim of this study was to investigate the correlation of NC, other anthropometric measures, and body composition analysis with cardiovascular risk factors and insulin resistance in individuals aged over 60 years attending a primary care center.

Methods
This cross-sectional study was performed at a Geriatric and Internal Medicine Outpatient Clinic, a primary care center, in the city of Niterói, State of Rio de Janeiro, Brazil, including 411 individuals who visited the Outpatient Clinic for any reason from March 2011 to March 2013.All participants signed an informed consent form.The study protocol was approved by the Ethics Committee of the Medical School of Universidade Federal Fluminense (0183.0.258.10).
Exclusion criteria were: clinical or laboratory evidence of cardiac, renal, or liver disease, goiter or other neck lesions, severe cognitive deficits, or individuals using corticosteroids or immunosuppressants.
All the subjects underwent a detailed physical examination.Blood pressure was measured at the left arm, with the subjects sitting after five minutes of rest, using a calibrated automatic sphygmomanometer (OMRON HEM 742INT; OMRON Healthcare, Bannockburn, USA).The same researcher performed anthropometrical measures, while the subjects were wearing light clothes and no shoes.Weight and height were measured using a calibrated anthropometric scale (Filizola, São Paulo, Brazil) with capacity for 150kg.BMI was calculated as weight (in kilograms) divided by the square of the height (in meters) (kg/m²).Circumferences were measured using an inelastic tape (SANNY, São Paulo, Brazil).Waist circumference (WC) was measured at the midpoint between the lowest rib and the iliac crest, with the patient standing and at the end of an exhalation.Hip circumference (HC) was measured at the widest point between the waist and the thigh.Waist-to-hip ratio (WHR) was calculated by dividing WC by HC.Neck circumference was measured at the base of the neck, below the laryngeal prominence, with the subject standing upright, looking straight ahead, with shoulders down.
The percentage of total body fat (%TBF) was assessed by a tetrapolar bioimpedance device (BIA) (Model 310, RJL; Biodynamics Corp., Seattle, USA), according to the protocol established by Lukasky et al. 14 .
Insulin resistance was assessed using the homeostasis model assessment IR index (HOMA-IR), which was calculated using the following formula: HOMA-IR = [fasting glucose (mmol/L) x fasting insulin (µU/L)]/22.5.Insulin resistance was defined by a HOMA-IR > 2.71, which is the cut-off value determined for the Brazilian population, for both sexes 15 .

Statistical analyses
Analyses were performed using IBM SPSS version 21.0 for Windows (IBM Corp., Armonk, USA).All data were analyzed and reported by sex, as fat metabolism is different between sexes due to the regional fat distribution 16,17 .Normality was verified using the Shapiro-Wilk test.The results are presented as medians with interquartile range for continuous variables.Associations between continuous variables and sex were determined by means of the Mann-Whitney U test.The Spearman correlation test (simple and partial) was adopted and adjusted for age and the diagnosis of diabetes.The correlation between variables was classified according to the categorization by Dancey & Reidy 18 .Receiver operating characteristic (ROC) curves were constructed to evaluate the performance of anthropometric parameters in order to identify insulin resistance.The areas under ROC curves (AUC) were calculated using a range of 95%.P-values < 0.05 were considered statistically significant.

Results
The study group consisted of 411 subjects (304 women and 107 men) with a mean age of 70 years.Age, BMI, systolic blood pressure (SBP), TG, fasting blood glucose, fasting insulin, and HOMA-IR were not significantly different between the male and female groups.Women had a significantly greater percentage of %TBF, higher TC, and HDL-c, but lower WC, NC, diastolic blood pressure (DBP), and uric acid (Table 1).
The AUC values for different anthropometric measures were statistically significant.In women, NC presented the largest AUC for IR, which is similar to WC.However, in men, WC showed the largest AUC, followed by BMI and NC (Figure 1).Using the ROC curve analysis, NC ≥ 39.4 cm for men and ≥ 33.7 cm for women were determined to be the best cut-off values for identifying subjects with IR with 67.3% sensibility for men and 66.9% sensibility for women.

Discussion
In this study, NC correlated significantly with traditional anthropometric indicators of obesity such as BMI and WC.Also, NC correlated with cardiovascular risk factors and %TBF.The performance of NC to identify IR was similar to WC in women and inferior to WC and BMI in men.
Several studies have shown that NC is an innovative and representative anthropometric measure of excess body fat or its abnormal distribution, and can be used to identify individuals with cardiovascular risk 11,12,13 .Different mechanisms have been implicated in the correlation of NC and cardiovascular risk factors.One possible mechanism is that free fat acid storage process in the fed state favors upper-body fat accumulation in both sexes 19 .In addition, the increased release of free fatty acids from VF could be an important mechanism contributing to or leading to IR and endothelial Cad.Saúde Pública 2017; 33(8):e00060916   ROC (receiver operating characteristic) curves for body mass index (BMI), neck circumference (NC), waist circumference (WC), and waist-to-hip ratio (WHR) for the diagnosis of insulin resistance (homeostasis model assessment insulin resistence index: HOMA-IR) in women and men (p-value < 0.001).
dysfunction 20 , which would render the measure of NC as an important tool in identifying individuals with increased cardiovascular risk.In a study by Zhou et al. 21, NC significantly correlated with all outcomes of cardiometabolic risk in both sexes.In this study, NC showed a significant correlation with BMI, WC and WHR, similar to the study in Turkish adults with mean age of 55 years 22 .In Brazilian young adults, a significant correlation between NC, BMI, WC, and WHR was also demonstrated 11 .
Neck circumference, BMI, WC, and WHR were positively correlated with %TBF, similar to previous studies that demonstrated significant correlation between NC and VF measured by CT or ultrasound 11,13 .The bioimpedance analysis is a validated method with satisfactory sensitivity and specificity to predict VF in older population 23 .Li et al. 24 , comparing neck and lower abdomen CT areas, found significant correlation between NC and VF in Chinese adult men and women, what may be accounted by the fact that neck fat was significantly correlated with VF.
Considering the cardiovascular risk factors, NC was positively correlated with fasting blood glucose, TG, uric acid, fasting insulin, and HOMA-IR, while there was a negative correlation between NC and HDL-c, in accordance with previous studies 12,19,24 .Several of these risk factors are components of the metabolic syndrome, which increases the risk of macrovascular disease 5,25 .
This study showed a significant correlation between NC and IR, measured by HOMA-IR, in agreement with previous studies 24,26,27 .Moreover, a previous analysis of the Framingham Heart Study demonstrated that NC is associated with IR, hypertension, and dyslipidemia, regardless of VF 12 .Stabe et al. 11 , using the same cut-off value for HOMA-IR in younger adults, showed that NC was the best anthropometric measure on the prediction of IR in women, similar to the results we found for older patients.Therefore, in addition to WC, NC could also be used as a simple, quick method for identifying insulin resistance, especially in the older population, who can present health conditions that affect the abdominal wall, organs, and cavity, as well as patients restricted to wheelchairs.NC, as an index of upper body obesity, can be used as a marker of IR in this population.This study has a few limitations, such as the small sampling size and the fact that the subjects were attending a primary care clinic.Also, the admixture of races in the Brazilian population may limit the external validity of these data to other populations.Further studies are necessary to identify the relationship between NC and cardiovascular risk factors in the general population.

Conclusions
This cross-sectional study conducted in the Southeast region of Brazil suggests that NC can predict insulin resistance, which is a marker of cardiovascular risk, in the older population attending a primary care center.

Contributors
M. A. N. Saad participated in the conception and design of the study; acquisition, analysis, and interpretation of data, drafting and critical revision of the intellectual content, final approval of the version to be published.M. L. G. Rosa contributed in the analysis and interpretation of data, drafting and critical revision of the intellectual content, final approval of the version to be published.G. B. Lima participated in the interpretation of data, critical revision of the intellectual content, final approval of the version to be published.R. A. Cruz Filho collaborated in the conception and design of the study, interpretation of data, critical revision of the intellectual content; final approval of the version to be published.All authors agreed to be accountable for all aspects of this study.

Table 1
Characteristics of subjects according to gender.Niterói, State of Rio de Janeiro, Brazil.