The Relationship Between Physical Fitness and Physical Activity During Adolescence and Cardiovascular Disease Risk Factors at Adult Age. The Amsterdam Growth and Health Longitudinal Study

The purpose of this study was to analyse the relationship between physical activity and physical fitness during adolescence (between 13 and 16 years of age) and cardiovascular disease (CVD) risk factors at adult age (32 years). The following CVD risk factors were considered: lipoprotein levels (total serum cholesterol (TC), high density lipoprotein cholesterol (HDL), the TC:HDL ratio, systolic and diastolic blood pressure, and indicators for body fatness and body fat distribution (sum of four skinfolds, waist to hip ratio, waist circumference). The data were derived from the Amsterdam Growth and Health Longitudinal Study, an observational longitudinal study in which eight repeated measurements were carried out over a period of 20 years. Physical fitness (measured by maximal oxygen uptake per kg bodyweight and by maximal slope reached on a treadmill test) during adolescence was related to a healthy CVD risk profile at the age of 32 years (i. e. inversely to the sum of four skinfolds, waist circumference and to total serum cholesterol). Physical activity during adolescence was not related to a healthy CVD risk profile at the age of 32 years.


Introduction
In the Western society the financial burden of treatment of chronic diseases at adult age is enormous. One of the most important chronic diseases in the developed countries is cardiovascular disease (CVD). Even though the clinical symptoms of CVD do not become apparent until much later in life, it is known that the origin of CVD lies in early childhood [13]. It is therefore often argued that prevention of CVD has to start as early in life as possible. A possible preventive strategy with respect to the prevention of CVD at adult age could be a change towards a healthy lifestyle during youth. Physical activity and physical fitness are recognised as important components of such a healthy lifestyle. For these components especially the adolescent period seems to be important. It is known that in the Western world the amount of habitual physical activity is decreasing dramatically during this age period [3,15]. However, the scientific evidence showing a positive relationship between physical activity and physical fitness during adolescence on the one hand and health outcomes at adult age on the other hand is scarce.
This important information should be obtained from longitudinal studies in which lifestyle parameters during youth are related to the prevalence of CVD at adult age. Performing such a long term prospective study is very difficult, because the population under study has to be very large and the follow-up period has to be very long. This kind of study is therefore merely retrospective in design [14]. The problem, however, is that retrospective infor-The Relationship Between Physical Fitness and Physical Activity During Adolescence and Cardiovascular Disease Risk Factors at Adult Age.

The Amsterdam Growth and Health Longitudinal Study
Study, an observational longitudinal study in which eight repeated measurements were carried out over a period of 20 years. Physical fitness (measured by maximal oxygen uptake per kg bodyweight and by maximal slope reached on a treadmill test) during adolescence was related to a healthy CVD risk profile at the age of 32 years (i. e. inversely to the sum of four skinfolds, waist circumference and to total serum cholesterol). Physical activity during adolescence was not related to a healthy CVD risk profile at the age of 32 years.

S 8
mation about lifestyle during youth is not very reliable [7]. Another research strategy is not to relate lifestyle during youth to the actual occurrence of CVD later in life, but to relate lifestyle during youth to "indirect" health parameters at adult age, i. e. to the levels of important risk factors for CVD like lipoprotein levels, blood pressure, body fatness and body fat distribution.
The Amsterdam Growth and Health Longitudinal Study [10] is a prospective longitudinal study, which started with subjects aged 13 years. Over a period of 20 years, longitudinal measurements were carried out. With these data it is possible to answer the following research question: Is there a relationship between physical activity and physical fitness during youth and CVD risk factors at adult age?

Methods
This study is part of the Amsterdam Growth and Health Longitudinal Study (AGHLS) [10]. Over a period of 20 years, eight repeated measurements were carried out. In general, during the first four years of the study four consecutive measurements were carried out. The initial age of the subjects at the beginning of the longitudinal study was 13.1 ( 0.8) years. In 1985 the study continued with a fifth measurement. The subjects returned in 1991 for a sixth measurement (age 27.1 [ 0.8] years) and in 1993 for a seventh measurement at the age of 29.1 ( 0.9) years. In 1996/ 1997 the subjects were remeasured for the eighth time at the age of 32.4 ( 1.1) years.
The present study uses two subpopulations of the AGHLS. The first subpopulation (132 males and 145 females) consists of all subjects with a measurement at the age of 13 years and a measurement at the age of 32 years. The second subpopulation (80 males and 96 females) consists of all subjects with at least three measurements during the adolescent period between 13 and 16 years of age and a measurement at the age of 32 years (see Fig. 1).
Besides the description of the natural longitudinal development of growth, health and lifestyle from adolescence into adulthood, the purpose of the AGHLS is to relate lifestyle parameters in youth to health status at adult age. Therefore at each year of measurement not only anthropometric parameters (body height, body weight, body composition), but also biological parameters (lipoprotein levels, blood pressure, physical fitness) and lifestyle parameters (nutritional habits, smoking behaviour, daily physical activity) were measured.

Physical activity and fitness
Daily physical activity was measured by a structured interview [24], which covered a period of the previous three months. The total time spent on all daily physical activities (activities at school and/or work, organized sports activities, unorganized sports activities, other leisure time activities etc.) was combined with the intensity of the different activities to form a total weighted activity score (expressed in METs/week). A distinction was made between medium heavy physical activity (average of 5.5 METs), heavy physical activity (average of 8.5 METs) and vigorous physical activity (average of 11.5 METs).
From the activity interview the following physical activity indicators were derived: 1) total amount of physical activity in minutes/week, 2) the total amount of physical activity in METs/week, 3) the amount of heavy physical activity plus vigorous physical activity (in minutes/week) and 4) the amount of only vigorous physical activity (in minutes/week). Physical fitness was measured by a running test on a treadmill at 8 km/hour with increasing slope until exhaustion. From this measurement, maximal oxygen uptake, absolute (V O 2 max), V O 2 max per kg bodyweight and maximal slope were used as determinants in the present study [9].

Risk factors for CVD
For the determination of lipoprotein levels {i. e. the concentration of total serum cholesterol ([TC]) and high density lipoprotein cholesterol ([HDL])} in blood serum, approximately 10 ml of venous blood was taken from the forearm vein with subjects in a non fasting state. External quality control took place with samples from a WHO reference laboratory. TC/HDL ratio was derived from the above measurements. Blood pressure (BP) was measured with an indirect method. A standard pressure cuff was placed around the left upper arm. With a sphygmomanometer diastolic blood pressure (DBP) and systolic blood pressure (SBP) were measured twice and the lowest value was recorded. Body fatness was operationalized as the sum of four skinfolds (SSF); i. e. biceps, triceps, subscapular and suprailiac. Skinfold thickness was measured with a Harpenden skinfold caliper and expressed in mm according to Weiner and Lourie [25]. Body fat distribution was operationalized as the ratio between the waist circumference and the hip circumference (WHR) and as the waist circumference.
For extensive information regarding the methods applied see Kemper [10].

Analysis
The aim of the analysis was to relate physical activity and fitness during adolescence (i. e. between 13 and 16 years of age) to CVD risk factors at adult age (i. e. at the age of 32 years). For this purpose linear regression analyses were used, with the physical activity and fitness indicators as separate determinants, correcting for gender and biological age. For each indicator two analyses were carried out: 1) the relationship between physical activity and fitness at 13 years of age and CVD risk factors at 32 years of age and 2) the relationship between "maintained" exposure to calendar age (years) 13 16 32 physical activity and fitness during adolescence (between 13 and 16 years of age) and CVD risk factors at adult age (see Fig. 1). "Maintained" exposure was defined as the average value of the particular parameter over the first four annual measurements during adolescence. After assessing the main effects, interactions between the determinants (physical activity and fitness) and gender were added to the linear regression models. When an interaction showed a p-value < 0.10, separate analyses were carried out for males and females.
All determinants were analysed as continuous determinants in the linear regression analyses, which were carried out with SPSS [19].
Results Table 1 shows descriptive information for the CVD risk factors measured at the age of 32 years and in Table 2 the same information is given for the lifestyle determinants. For the latter the value at the age of 13 years is given as well as the average value for the total adolescent period between 13 and 16 years of age. In Tables 3 -5 the results of the linear regression analyses relating physical activity and physical fitness during adolescence to CVD risk factors at adult age are given. For the lipoproteins ( Table 3) it was shown that all activity measures at the age of 13 years were inversely related to HDL levels at the age of 32 years. When the average value between 13 and 16 years was considered, the standardized regression coefficients were much lower. The aver-age fitness levels (expressed as V O 2 max per kg bodyweight and as maximal slope) between 13 and 16 years of age were inversely related to TC levels and to the TC:HDL ratio. When fitness parameters at 13 years of age were considered, the standardized regression coefficients were lower and the relationships were only found for males and not for females. Table 4 shows the relationships between physical activity and physical fitness during adolescence and systolic and diastolic blood pressure at adult age. In general the standardized regression coefficients were rather low and the results show no consistent pattern. The only significant relationship was a positive relationship between the average V O 2 max per kg bodyweight between 13 and 16 years of age and systolic blood pressure at 32 years of age and an inverse relationship between V O 2 max (absolute) and systolic blood pressure for males.
In Table 5 the results of the relationship between physical activity and physical fitness during adolescence and body fatness and body fat distribution at 32 years of age are given. Most remarkable are the inverse relationships between V O 2 max per kg bodyweight and maximal slope on the one hand and the sum of four skinfolds on the other hand. This inverse relationship is also found with waist circumference, but the standardized regression coefficients were lower and more pronounced for females. For the physical activity parameters, positive relationships were found with the waist circumference, but no relationships were observed with the sum of four skinfolds and the waist to hip ratio, except for an inverse relationship between heavy and vigorous physical activity between 13 and 16 years of age and the waist to hip ratio. This relationship was however only found for females.

Discussion
The purpose of this paper was to analyse the relationship between physical activity and physical fitness during youth and CVD risk factors at adult age. There are three possible mechanisms involved in the relationship between activity and fitness during youth and health at adult age: 1) Activity and fitness during youth are related to health status during youth. This can be important because from the literature it is known that health status during youth is an important predictor for health status    at adult age [13,21,22]. 2) Activity and fitness during youth are related to activity and fitness at adult age. This can be important because there is extensive evidence that activity and fitness at adult age is related to health status at adult age [8,11,16]. 3) Activity and fitness during youth are directly related to health status at adult age. The latter theoretical mechanism is investigated in the present study.
In general the results of the present study are rather disappointing. Only the two fitness parameters V O 2 max (expressed per kg bodyweight) and maximal slope were related to a healthy CVD risk profile. They were both inversely related to the sum of skinfolds, inversely related to waist circumference and inversely related to total serum cholesterol. This finding is in agreement with the study of Blair et al. [6] in which low physical fitness is found to be an important determinant of all causes of mortality in adult populations.
The strongest evidence for the beneficial effects of physical fitness on cardiovascular health in the present study was found in the relationship with the sum of four skinfolds (i. e. body fatness). The observed relationships were more or less expected regarding the literature in this field [17]. The fact that in the present study no relationships were observed among physical fitness indicators and diastolic and systolic blood pressure is also in agreement with the literature [1,17]. Based on experimental studies it is sometimes argued that a healthy effect of physical fitness on blood pressure is only observed in subjects with elevated blood pressure [21]. Because the population of the AGHLS consists of healthy subjects, this hypothesis could not be investigated in the present study. For the relationship between physical fitness and lipoprotein levels, the literature is quite ambiguous. There is not much evidence of a relationship between physical fitness and lipoprotein levels. Probably the most consistent finding is the beneficial effect of physical fitness on HDL-levels [2,17]. However, this evidence is mostly based on cross-sectional studies and is not confirmed in the present longitudinal study.
The reason for not finding very strong relationships between physical fitness and CVD risk factors can be the relatively high fitness level of the research population. Especially from training studies it is known that the relationship between an increase in fitness and CVD risk factors is stronger for relatively low levels of fitness.
For physical activity no relationships with a healthy CVD-risk profile were found. In fact, for all physical activity indicators measured at 13 years of age an inverse relationship was observed with HDL-levels. This is rather surprising, because it is generally accepted that if there is a positive effect of physical activity on cardiovascular health, it is the positive relationship with HDLlevels [2,17]. One of the possible explanations for this unexpected finding is that the amount of physical activity decreases dramatically after the age of 13 years [15]. So maybe the subjects with the highest levels of physical activity at the age of 13 years are also the subjects with the steepest decrease in physical activity over the period under study. This explanation is in agreement with the fact that the average physical activity levels between 13 and 16 years of age were not significantly inversely related to HDL-levels. To investigate this possible explanation for all four physical activity indicators, the change in activity between 13 years of age and 32 years of age was calculated. This change was then related to HDL levels at the age of 32 years. Two analyses were carried out: Firstly the change was treated as a continuous determinant and secondly this change was divided into four equal groups and this categorical variable was used as a determinant in the linear regression analysis. From the results of these additional analyses (see Table 6) it can be seen that the explanation makes sense. There is an inverse relationship between a decrease in activity and HDL levels, which indicates that the subjects with the steepest decrease in physical activity levels between 13 and 32 years of age show the lowest HDL-levels at the age of 32 years. This is an important finding which must be taken into account when evaluating the results of other studies relating physical activity during youth and health indicators at adult age, especially because it is known that physical activity levels are decreasing enormously from childhood to adult age. Regarding total cholesterol and the TC:HDL ratio no relationships were found with any of the physical activity indicators. This is not really surprising, because the results from other (both cross-sectional and longitudinal) studies are not consistent [2,17,23].
In the present study daily physical activity during youth was not significantly related to both systolic and diastolic blood pressure. In the literature there are some contradictory results regarding this relation [20]. It is sometimes argued that (high intensity) physical activity is associated with an increase in blood pressure, which may be caused by an increase in muscle mass due to an increase in physical activity. An increase in muscle mass is associated with an increase in blood pressure. Baranowski et al. [5] argued that the possible lowering effect of physical activity on blood pressure only holds for subjects with hypertension and not for subjects with normal values of BP. It was already mentioned that most of the subjects of the AGHLS showed normal BP values at adult age, so this hypothesis can not be investigated.
Daily physical activity during adolescence was not related to the sum of four skinfolds, to the waist to hip ratio and to the waist circumference at adult age. In the literature there is strong evidence that physical activity is related to both body fatness and to a positive body fat distribution [4,12,18]. However, most of the evidence is based on cross-sectional studies and it is shown that this relationship is rather short-term. In fact, when investigating the long-term relationship between physical activity and both fatness and body fat distribution, not finding any relationship as in the present study is in agreement with the literature [5].

Conclusions
This study does not show strong relationships between physical activity and physical fitness during adolescence and CVD risk factors at adult age. Only physical fitness seems to be related to a healthy CVD risk profile. For physical activity it was shown that a decrease in activity levels between adolescence and adulthood was related to unhealthy HDL-levels. A decrease in activity levels seems to be worse than a stable low activity lifestyle during this age period.