Subjects
The Young Hearts Project (YH) is an ongoing longitudinal study evaluating the prevalence of CVD risk factors in young people living in Northern Ireland. Sampling procedures and methods used in the first two, school-based screening phases of the YH study have been described fully elsewhere [5]. Briefly, the initial screening (YH1), conducted in 1989/1990, surveyed 1015 adolescents (12-year old boys, n 251;12-year old girls, n 258; 15-year old boys, n 252; and 15-year old girls, n 254) randomly selected from post-primary schools. At that time, the resulting cohort represented a 2% sample of each of the two age populations in Northern Ireland. In 1992/93, a follow up study (YH2) was undertaken, in which subjects from the original 12 year old cohort were reassessed using procedures identical to those used in YH1. The response rate in the follow-up study was 90%. Between October 1997 and October 1999, all YH1 subjects were invited to participate in the third, hospital-based screening phase (YH3), and a 48.2% response rate was achieved. Reasons for the low response rate in YH3 have been described in full elsewhere [12]. Briefly, non-attenders reported that they were 'too busy', 'living outside Northern Ireland', 'busy with new job', 'couldn't be bothered' or 'didn't feel that the study was relevant to them'. As described by [12] and [13], attempts were made to determine the representativeness of the YH3 cohort by comparing the baseline YH1 data for those who participated in YH3, with the data obtained for those who declined to participate. YH3 participants tended to be from families with higher socio-economic status, and had lower BMI at baseline (YH1) than non-participants. Furthermore, males who declined to attend for screening at YH3, were fatter and reported a greater saturated fat intake at YH1 than YH3 male participants.
The analyses reported in the current paper are restricted to males (n 245) and females (n 231) for whom there were complete data sets at age 15y (either from YH1 or YH2) and at young adulthood [mean (SD) age 22.0 (1.6)y]. Ethical approval for each phase of the study was obtained from the Medical Research Ethical Committee of The Queen's University of Belfast, and written informed consent was obtained from all subjects prior to participation.
Anthropometry
Each subject's height, weight and skinfold thicknesses were measured at all study timepoints. Standing height was measured to the nearest millimetre using a Harpenden portable stadiometer (Holtain, UK), and body weight was measured to the nearest 0.1 kg using an electronic balance (Seca, Germany; 200 kg × 0.1 kg). For both measurements, subjects wore light indoor clothing and no shoes. Body mass index (BMI) was then calculated as weight (kg)/ [height (m)]2. Skinfold thicknesses were measured to the nearest millimetre using Harpenden callipers at four sites (biceps, triceps, subscapular, suprailiac). Two measurements were taken at each site and the average was recorded. The sum of the four skinfolds thicknesses was then calculated for each subject.
Dietary intake
At all study timepoints, dietary data were obtained using the diet history method [14]. This consisted of a detailed, open-ended one-to-one interview, the purpose being to ascertain the habitual weekly food intake of each subject. The diet history method was used for two reasons. Firstly, in subjects aged 15y, the diet history has been shown to provide more valid estimates of energy intake at the group level than weighed records [15]. Secondly, given that a complete diet history can be obtained from a subject in approximately one hour, it was the most feasible and cost-effective method for obtaining detailed dietary information from the YH1 and YH2 school-based cohorts. The method was used again in YH3 in order to maintain continuity. Reported energy and macronutrient intakes were calculated using computerised databases based on UK food composition tables as previously described [16, 12]
Physical activity
At age 15y, habitual physical activity was assessed by self-report questionnaire, and scored according to the method of [17]. This method assessed the extent of daily participation in activities that were based around a typical school day. Each activity was assigned a score from 1–100, based on its frequency, intensity and duration.
As the school-based questionnaire was not relevant to the young adult subjects, a modification [18] of the Baecke questionnaire was used in YH3 to quantify habitual work activity, sports activity and non-sports leisure activity. For each of the three activity components, scores based on a five-point Likert scale were calculated and summed, giving total possible scores ranging from 3–15.
Aerobic fitness
Aerobic fitness at age 15y was assessed by the 20 metre shuttle test (20MST). In order to estimate maximal aerobic capacity, or VO2max (ml/kg/min), the number of laps completed by each subject in this maximal endurance test was entered into a sex-specific regression equation, based on data obtained in the Northern Ireland Fitness Survey [17].
As it was not feasible to conduct the 20MST at young adulthood (due to a lack of space in the hospital setting), VO2max was assessed using the Physical Work Capacity at a heart rate of 170 beats per minute (PWC170) cycle ergometer test [19]. PWC170 was calculated as the workload corresponding to a heart rate of 170 bpm, and expressed per kg body weight. The volume of oxygen consumed and the heart rate were monitored throughout the test (Quinton Metabolic Cart, Quinton, USA). For each subject, a straight line was fitted to three pairs of data (heart rate in bpm, VO2 in ml/kg/min), and this was used to estimate VO2max at the age-adjusted maximum heart rate [12].
Statistical analyses
All data were analysed using SPSS version 11.0.1 (SPSS Inc, Chicago, USA). Means and standard deviations were used to summarise the data for physical characteristics, aerobic fitness, physical activity scores and energy and macronutrient intakes of males and females at age 15y and young adulthood.
Tracking of each of these variables over time was assessed by determining the extent to which subjects who were placed into low, medium and high categories at age 15y, maintained their ranking in young adulthood. Owing to the fact that different techniques were used to measure physical activity and aerobic fitness at each timepoint, a method based on ranks, rather than actual measurements, was employed for assessing the tracking of these variables. Tracking of the other variables was also assessed using the rank based method because of its relative simplicity, and its ability to show the numbers of subjects making the transition between low, medium and high categories [20]. For example, in order to study the tracking of physical activity in females from age 15y to young adulthood, the group of 225 girls aged 15y was divided into three classes by physical activity score: lowest 25% (L1); middle 50% (M1); highest 25% (H1). Rather than using pre-determined fixed values, each class was defined by the first and third empirical quartiles. In young adulthood, the female group was divided into three similar classes; L2, M2 and H2. Using these two sets of classifications, a 3 × 3 tracking matrix was constructed; the entry in a specific cell being the number of subjects belonging to the corresponding classes at age 15y and at young adulthood (see Figure 1 for examples). This approach provides a broad picture of the relative changes in a particular variable over time, such that a matrix with relatively small off-diagonal elements provides evidence of 'good' tracking. For the purposes of this study, the degree of tracking was summarised by a weighted kappa (κ) value, and interpreted according to [21] as follows: κ ≤ 0.20, poor tracking; κ 0.21–0.40, fair; κ 0.41–0.60, moderate; κ 0.61–0.8, good; κ 0.81–1.0, very good. This procedure was undertaken separately for males and females to assess the tracking, between age 15y and young adulthood, of energy and macronutrient intakes, height, weight, BMI, skinfold thicknesses, aerobic fitness and physical activity scores.