Age- and Gender-Specific Reference Intervals for Fasting Blood Glucose and Lipid Levels in School Children Measured With Abbott Architect c8000 Chemistry Analyzer

Abstract

Reference intervals for pubertal characteristics are influenced by genetic, geographic, dietary and socioeconomic factors. Therefore, the aim of this study was to establish age-specific reference intervals of glucose and lipid levels among local school children. This was cross-sectional study, conducted among Saudi school children. Fasting blood samples were collected from 2149 children, 1138 (53%) boys and 1011 (47%) girls, aged 6 to 18 years old. Samples were analyzed on the Architect c8000 Chemistry System (Abbott Diagnostics, USA) for glucose, cholesterol, triglycerides, HDL and LDL. Reference intervals were established by nonparametric methods between the 2.5th and 97.5th percentiles. Significant differences were observed between boys and girls for cholesterol and triglycerides levels in all age groups (P < 0.02). Only at age 6–7 years and at adolescents, HDL and LDL levels were found to be significant (P < 0.001). No significant differences were seen in glucose levels except at age 12 to 13 years. Saudi children have comparable serum cholesterol levels than their Western counterparts. This may reflect changing dietary habits and increasing affluence in Saudi Arabia. Increased lipid screening is anticipated, and these reference intervals will aid in the early assessment of cardiovascular and diabetes risk in Saudi pediatric populations.

Introduction

The evaluation of pediatric diseases relies on the measurement of a spectrum of disease biomarkers in clinical laboratories to guide important clinical decisions. Physicians rely on the availability of suitable and reliable reference intervals to accurately interpret laboratory test results in conjunction with data collected via medical histories and physical examinations. However, critical gaps currently exist in accurate and up-to-date reference intervals (normal values) for accurate interpretation of laboratory tests performed in children and adolescents [1]. These gaps in the available pediatric laboratory reference intervals have the clear potential to contribute to erroneous diagnoses or misdiagnoses of many diseases of childhood and adolescence. Most of the available reference intervals for laboratory tests were determined over two decades ago using older instruments and technologies and are no longer relevant, considering the current testing technology used by clinical laboratories. It is thus critical and of the utmost urgency to establish a more reliable and comprehensive database [1].

Establishing age-specific reference values for children is challenging. Many international and national institutes have sought to address these challenges. One of the most promising initiatives is the Canadian initiative, CALIPER (Canadian Laboratory Initiative on Pediatric Reference Intervals), which was established to create a comprehensive database for both traditional and emerging biomarkers of pediatric diseases [2–4]. We believe that provision of age-matched reference values for all applicable tests is critical for the proper care of pediatric patients, and we are committed to providing the most accurate reference values possible.

Many institutions have conducted a literature review to determine whether published studies provide useful and reliable interpretive guidance based on analytic method comparability, patient selection criteria and sample size, specimens employed, and statistical tools used. When it is impossible to identify applicable reference values in the published literature, or if published values are deemed to be incomplete or based on insufficient data, de novo pediatric reference values have to be established. In some cases, as for example, with sex steroids, such values might require subject populations who have undergone clinical puberty staging using the Tanner classification.

With the increasing incidence of cardiovascular disease, diabetes and metabolic syndrome in the Middle East, testing of lipid levels would be helpful in assessing individuals with risk factors for these diseases. Although adult reference intervals for local populations are well established and validated with other literature-derived values, there is not much reference data available for the pediatric population. It was therefore necessary to conduct this study to establish a reference database and to evaluate the clinical relevance of lipid tests in our pediatric population.

Approximately 5–10% of Arab children may be at risk for coronary heart disease (CHD). Approximately 25% of Arab children have a higher level of total serum cholesterol than the desirable target value of 4.4 mmol/l set forth by the NCEP [5].

Current recommendations from the National Cholesterol Education Program Adult Treatment Panel III suggest that initial screening of patients for levels of total cholesterol, low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, and triglycerides should be carried out when patients are in a fasted state [6–9]. The guidelines also state that total cholesterol, HDL cholesterol, and non-HDL cholesterol may also be measured in nonfasting sample populations.

Thus, our study was designed to be a landmark study, whose aim was to determine and establish the lipid reference intervals based on age and gender among Saudi Arabian National Guard school children and to compare them with other ethnic groups locally and internationally.

Methods

The study, using a cross-sectional design, was carried out among National Guard school children in Riyadh, Saudi Arabia. These children are of National Guard members. A total of 2,149 school children, comprising 1,138 (53%) boys and 1,011 (47%) girls, were selected for the study using a cluster sampling strategy. Informed consent and child assent forms were obtained at least two weeks prior to data and blood collection. The parents filled out a self-administered questionnaire that provided information regarding their socioeconomic status and their child’s health history. The study protocol was approved by the Research and Ethics Subcommittee at the King Abdullah International Medical Research Center (KAIMRC) in Riyadh.

The study was conducted between January and June 2008. Each child was examined by the data collection team in his or her school clinic or a specifically designated area. Height was measured using a wall-mounted stadiometer, and the measurement was recorded to the nearest 0.1 cm. Weight was measured to the nearest 0.1 kg with a beam-balance scale. All measurements were collected by a nurse on the data collection team in the child’s school clinic. A general and systematic examination was done by hospital family-medicine clinicians, paying particular attention to the presence of any physical or endocrinological abnormalities. Healthy subjects were selected based on absence of diabetes, hypertension, or coronary heart disease and a history of nonconsumption of any medication for lowering lipid levels. Children were excluded for any chronic or acute illness or medication use, abnormal blood pressure, recent or current surgery or hospitalization, obesity, oral contraceptive use, pregnancy or alcohol consumption or tobacco use.

Fasting blood samples were collected in serum separator tubes (Greiner bio-one, Germany), labeled, transported, lifted to be clotted for more than 15 min and then centrifuged for 10 min at 3,000 rpm using a Multifuge 35R. Serum samples were then separated and immediately assayed by the clinical chemistry analyzer Architect Chemistry System (Abbott, USA); otherwise, they were stored in a deep freezer at −70°C until further testing, but not for more than three months because of the instability of lipid samples in vitro. A glucose test and four lipid tests, i.e., total cholesterol, triglycerides, low density lipoprotein (LDL), and high density lipoprotein (HDL), were performed. Three levels of normal and abnormal quality control materials were used (Bio-Rad, USA) for each assay. The patient results were transmitted and stored in the Laboratory Information System (LIS) (Cerner, USA), which interfaced with an Architect Chemistry analyzer. The data were then retrieved from the LIS for subsequent statistical analysis.

Descriptive statistics were determined by calculating the number and percent for categorical variables and the mean and standard deviation for continuous variables. Stratified analyses by age for height, weight, and body mass index were also carried out. Moreover, the mean, standard deviation, and 95% confidence intervals for age and gender were also calculated. Data reduction and analyses were carried out using the SPSS program (version 15).

The intervals reported reflect the central 95% confidence intervals for the populations tested. The low and high values were calculated at the 2.5th and 97.5th percentiles. As a result, reference intervals were reported either as separate ages or separate genders.

Results

Table 1 summarizes the range of lipid and glucose test results between the 2.5th and 97.5th percentiles in boys and girls in different age groups ranging from 6 to 18 years old. Significant differences in cholesterol, triglycerides, HDL, LDL and glucose levels (P < 0.0001) were observed between both boys and girls for each age group. Therefore, specific age reference intervals should be used for each age group, and combined reference intervals cannot be used here.

Table 1.

Lower (2.5th) and upper (97.5th) percentiles for tested lipid and glucose levels in Saudi children between 6 and 18 years of age

Age	Cholesterol		Triglycerides		HDL		LDL		Glucose
	Boys	Girls	Boys	Girls	Boys	Girls	Boys	Girls	Boys	Girls
6–7 years n (boys) 102 n (girls) 108	1.8–5.1 (3.8)	3.2–7.3 (4.5)	0.2–1.5 (0.6)	0.4–1.5 (0.7)	0.6–2.0 (1.3)	1.0–2.0 (1.4)	0.9–3.4 (2.2)	1.7–5.1 (2.8)	2.6–5.5 (4.2)	3.5–5.2 (4.3)
P value	<0.0001		<0.022		<0.001		<0.0001		<0.178
8–9 years n (boys) 165 n (girls) 161	2.4–5.6 (4.1)	3.1–5.9 (4.3)	0.3–1.4 (0.7)	0.4–1.9 (0.8)	0.8–2.1 (1.4)	0.9–2.0 (1.4)	1.4–3.6 (2.4)	1.6–4.1 (2.6)	3.2–5.3 (4.4)	3.5–5.2 (4.3)
P value	<0.003		<0.001		<0.483		<0.016		<0.620
10–11 years n (boys) 242 n (girls) 221	2.9–5.6 (4.1)	1.7–5.5 (4.1)	0.3–1.6 (0.7)	0.3–2.1 (0.8)	1.0–2.0 (1.4)	0.7–1.9 (1.4)	1.3–3.9 (2.4)	0.8–3.7 (2.4)	3.6–5.3 (4.5)	2.2–5.7 (4.5)
P value	<0.969		<0.003		<0.543		<0.845		<0.631
12–13 years n (boys) 234 n (girls) 183	2.8–5.7 (4.1)	1.7–5.7 (4.1)	0.3–1.7 (0.7)	0.4–1.8 (0.8)	0.9–1.8 (1.3)	0.4–1.8 (1.3)	1.4–4.0 (2.5)	0.9–4.1 (2.5)	3.8–5.4 (4.7)	1.8–5.5 (4.5)
P value	<0.679		<0.001		<0.075		<0.879		<0.034
14–15 years n (boys) 219 n (girls) 192	2.9–5.4 (3.8)	2.9–5.3 (4.0)	0.3–2.1 (0.8)	0.3–1.4 (0.8)	0.8–1.8 (1.2)	0.8–1.8 (1.3)	1.3–3.7 (2.3)	1.4–3.5 (2.4)	3.9–5.5 (4.6)	3.5–5.6 (4.6)
P value	<0.03		<0.091		<0.051		<0.018		<0.15
16–18 years n (boys) 176 n (girls) 146	2.6–5.1 (3.8)	2.9–5.3 (4.1)	0.4–2.1 (0.9)	0.4–1.6 (0.7)	0.8–1.5 (0.8)	0.9–1.9 (1.3)	1.3–3.7 (2.3)	1.5–3.5 (2.4)	3.4–5.7 (4.6)	3.5–5.6 (4.6)
P value	<0.005		<0.0001		<0.0001		<0.231		<0.077

The values in parentheses denote the mean. All lipid and glucose values are in units of mmol/l. A P value <0.05 was considered statistically significant

There were significant differences between boys and girls at 6 to 7 years of age for cholesterol (P < 0.0001), triglyceride (P < 0.022), HDL (P < 0.001), and LDL levels (P < 0.0001), but not for glucose levels (P = 0.178). At 8 to 9 years of age, only cholesterol, triglyceride and LDL levels showed significant differences between boys and girls.

Significant differences between boys and girls were observed in triglycerides at 10 to 11 years of age (P < 0.02) and at 12 to 13 years of age (P < 0.0001). At the onset of puberty between 14 and 16 years of age, there were significant differences between boys and girls in cholesterol (P < 0.0001), triglycerides (P < 0.0001), HDL (P < 0.0001), and LDL (P < 0.001). Therefore, separate reference intervals should be used for each gender at this age. Only the triglyceride level differed significantly in boys and girls at 10 to 11 years of age. Similarly, at 12 to 13 years of age, triglyceride (P < 0.001) and glucose (P < 0.034) levels differed significantly between boys and girls. However, at 14 to 15 years of age, only cholesterol and LDL levels were significantly different in boys and girls, with p-values of P < 0.03 and P < 0.018, respectively. Finally, at 16 to 18 years of age, cholesterol, triglyceride, and HDL levels were significantly different in boys and girls with P values of P < 0.005, P < 0.0001 and P < 0.0001, respectively.

Discussion

Our study shows that the cholesterol levels in prepubescent boys and girls under 10 years of age were slightly lower than those found in Western countries [10, 11], with the exception of girls of preschool age, but higher than the values reported by Hicks et al. [12]. In contrast, we found slightly higher cholesterol levels at puberty. This may be explained by the different analytical methods used for the nonfasting sample populations. We also found that HDL levels were the same at prepubescent ages for both boys and girls and were lower in pubescent boys above the age of 16. This is consistent with other published results [10, 11]. However, Kottke et al. [12, 13] have reported lower levels of HDL in their pediatric population. This may also be attributed to differences in the analytical method used in their study [13]. Our LDL data were higher and similar to those reported by others [14, 15] for prepubescent children but were lower for children in puberty.

Some earlier studies of serum cholesterol in children in Saudi Arabia were conducted on hospital-based populations. Rafii et al. [16] reported that the mean serum cholesterol concentration in Saudi boys and girls younger than 14 years old was 3.88 ± 0.83 mmol/l. They found no statistical differences between girls and boys, which was inconsistent with our data for the prepubescent and pubertal stages.

There are several reports showing that girls of preschool and school age have higher total serum cholesterol concentrations than boys [17, 18]. This is consistent with our findings, in which a gender-determined difference was seen at early school and adolescent ages. In addition to total cholesterol, differences were also observed in other lipid tests with respect to the adolescent stage.

Serum cholesterol concentrations have been reported to rise during childhood [19, 20], although some investigators found no such increase [21, 22]. A decrease in serum cholesterol after the adolescent growth spurt has also been reported [23, 24]. In our study, we observed a decrease in the mean cholesterol level in the prepubescent and pubertal stages.

We also found age-related differences between boys and girls in all of the lipid tests. No age-related differences were observed between boys and girls at any age except the age of 12–13. This may be attributed to the early hormonal changes in girls but not in boys. Yip et al. [25] reported similar findings, in which the lipid levels have age- and gender-related differences, particularly in young adults following puberty. They found that the concentrations of total cholesterol and LDL remain relatively constant throughout childhood but decrease for males in early adulthood. Triglyceride levels increase gradually throughout childhood and adolescence, and along with the reference range for cholesterol, the upper limits of these reference intervals exceed the recommended lipid concentrations in children. However, they observed no age- or gender-related differences in HDL until after puberty, when the values for boys decrease slightly.

Lopez et al. [26] measured the lipid levels in 2,153 children of both sexes from birth until 18 years of age. They did not report any significant differences in total cholesterol, HDL or LDL with respect to gender up to 12 years of age; in their study, these levels were significantly higher in adolescent girls than in adolescent boys. In contrast, triglyceride and very low density lipoprotein (VLDL) levels were higher in girls than in boys up to 12 years of age, but there were no gender-related differences in these levels in adolescents. The authors proposed that the changes in the lipid profile in adolescence were related to the hormones produced at puberty. In contrast, we found a significant gender-specific difference in HDL and LDL levels at 6 to 7 years of age, but not at age 14 and above, using a different chemistry analyzer system (the Architect c8000).

Yip et al. [25] assessed lipids in 525 serum or plasma samples from an outpatient pediatric population younger than 20 years of age using the VITROS 5,1 FS Chemistry System. The lipid levels demonstrate age- and gender-related differences, particularly during the first year of life and in young adults following puberty. They found that the concentrations of total cholesterol and LDL are lowest in the 12 months after birth and remain relatively constant throughout childhood but decrease in males in early adulthood, which is consistent with our findings. Triglyceride levels increase gradually throughout childhood and adolescence. This is also consistent with our results in boys, but we found that triglyceride levels decreased in girls after puberty. With respect to HDL, no age- or gender-related differences were observed in their study until puberty, when the levels in boys decreased slightly. In contrast, our results show that there were gender-related differences in postpuberty HDL levels.

Chan et al. [2] assessed lipids in 1,459 blood samples from children younger than 20 years old who visited select outpatient clinics and were deemed to be metabolically stable. The samples were analyzed on an ARCHITECT ci8200 chemistry analyzer. They observed that the upper limit of the normal range of cholesterol and triglycerides in their study population was clearly higher than the current cut-offs in lipid guidelines, but this may simply reflect the high prevalence of lipidemia in the general pediatric population. It is also important to note that the high triglyceride levels were likely due to the fact that the samples collected were from nonfasting subjects, which is inconsistent with our study design.

In our study, blood samples were collected during the months of April and May so that the effect of seasonal variation in the serum cholesterol concentration was eliminated [27]. The effect of meals could not be determined because samples from nonfasting subjects could not be collected after discontinuing fasting conditions. A recent study suggests that the consumption of a meal has negligible effects on serum cholesterol levels and no effect on triglycerides and glucose levels [8, 21].

We conclude that, unlike children in other developing countries, Saudi children do not have lower serum cholesterol levels than their Western counterparts. We believe that these findings reflect changing dietary habits and increasing affluence in Saudi Arabia. We also conclude that our reported reference intervals in Saudi children and adolescents provide an important update for the clinical evaluation of lipids. Increased lipid screening is anticipated, and these reference intervals will aid in the early assessment of cardiovascular and diabetes risk in Saudi pediatric populations.