Human scalp hair samples were collected in three cities (Winston-Salem, Clemmons, and Kernersville) in Forsyth County, North Carolina, USA. Forsyth County is 409.60 square miles in area, with a population of 350,670 [57]. Females and males account for 52% and 48%, respectively of the population. In addition, people under 18 years of age account for 24.6% of the sample population while those older than 18 years are make-up of 73.4% of the population in Forsyth County. The demographic racial composition of Forsyth County consists of Caucasians (majority 62.3%) and non-Caucasians (minority, 37.7%). Non-Caucasians were made up of African Americans, 26%; Hispanics, 11.9%; and Asians, 1.9%. Scalp hair samples were collected from 125 participants using stainless steel scissors during regular haircut periods at various barber shops and beauty salons over a period of three weeks in June 2011. The sample population consists of 60 males and 65 females aged between 5 and 70 years old. In addition, the hair samples were collected from different races and demographics consisting of Caucasian (n = 87) and non-Caucasians (n = 38). Non-Caucasians were made up of African Americans, Hispanics, and Asians. Only eighteen of the people from whom hair samples were collected declared themselves as tobacco product smokers while 107 identified themselves as nonsmokers.    

The collected hair samples were immediately placed in cleaned polyethylene plastic bags, properly labeled, and transported to a dust free laboratory, where the samples were sequentially washed with acetone and ethanol, and thoroughly rinsed with deionized water to remove pharmaceutical products, particulates, and other exogenous materials. The washed hair samples were subsequently dried in an oven at 90˚C for approximately 20 minutes and stored in pre-nitric acid washed Teflon sample containers. Approximately 1.0 g of dried hair sample from each participant was accurately weighed and digested with 15 mL HNO₃ (trace metal grade, purity, 99.999%) in a digestion flask. Hair sample digestions were initiated at relatively low temperatures to prevent a violent reaction until all the hair samples were fully dissolved in the nitric acid. The temperature was then subsequently increased until the solution turned pale yellow to ensure complete sample digestion. The nitric acid digested hair samples were filtered into a 25 mL volumetric flask using a Whatman filter paper (ashless) and diluted to the mark with deionized water.  

Working range standard solutions used to construct calibration curves for each element were prepared by serial dilution of 1000 ppm standard stock solutions of each element purchased from Fisher Scientific. The calibration curve for each element was constructed by plotting the absorbance obtained from FAAS analysis of the standard solution versus the element concentration. The nitric acid digested hair samples were subjected to trace element analysis using a flame atomic absorption spectrophotometer (Shimadzu, AA-6300) using a pre-mixed burner air-acetylene flame. The flow rates of the fuel and oxidant gases were always carefully optimized for each metal analysis. Other routine instrumental checks and calibrations were always performed on the spectrometer before use to ensure the accuracy, reliability and consistent performance of the spectrometer. Each sample was analyzed in triplicate and the averages of the trace element concentrations in the hair samples were calculated using the constructed calibration curves and listed in Table 1. All necessary precautions were also observed to ensure the accuracy of the results of this study. First, the acetone and ethanol used for the hair wash were of spectroscopic grade. Second, the highest purity HNO₃ acid (purity, 99.999%) was used for the hair digestion and standard solution preparations. Also, all glassware including the sample containers, digestion flasks, and volumetric flasks were pre-soaked in 6M nitric acid for three days and thoroughly rinsed with deionized water before use to remove impurities and contaminants. Additionally, all trace element sample analyses were blank subtracted.

A recovery study was also performed for each metal to further evaluate the accuracy and reliability of the results of the heavy metal concentrations obtained in the analysis. The recovery study was performed by randomly spiking ten previously analyzed hair samples with a known concentration of metal standard solution. The spiked samples were then subjected to HNO₃ digestion and metal analysis procedures under the same experimental conditions as previously used for hair sample analysis. The recovery of the metals in the spiked samples was evaluated by comparing the known concentration of the spiked metal with the concentration detected using FAAS spectrometer.

Linear regression and multivariate principal component analysis for pattern recognition of the levels of the metals in human scalp hair was performed using chemometric software (The Unscrambler, CAMO Inc., 9.4).

Table 1 presents the results of the calibration curve parameters constructed for the investigated trace elements in the hair samples. The table indicates the wavelength used for the AAS elemental analysis, the limits of detection (LOD), and limits of quantitation (LOQ) of the trace elements analyzed. The LOD, which is defined as the minimum detectable amount of metal, was calculated using the equation: LOD = 3 s/m; where s is the signal of the blank and m is the slope of the calibration curve. The limit of quantitation, defined as the lowest measurable concentration of analytes (trace elements in this study), was evaluated using the formula: LOQ = 10 s/m. The parameters used to calculate LOQ were as previously defined for LOD. The high values of the correlation coefficients (r) obtained in Table 1 demonstrate good linear correlation of the absorbance with trace element concentrations.