European Journal of Epidemiology

, 22:771

Commuting physical activity is favourably associated with biological risk factors for cardiovascular disease

Authors

    • Research Centre for Prevention and HealthGlostrup Hospital
  • Knut Borch-Johnsen
    • Steno Diabetes Centre
  • Torben Jørgensen
    • Research Centre for Prevention and HealthGlostrup Hospital
Cardiovascular Disease

DOI: 10.1007/s10654-007-9177-3

Cite this article as:
von Huth Smith, L., Borch-Johnsen, K. & Jørgensen, T. Eur J Epidemiol (2007) 22: 771. doi:10.1007/s10654-007-9177-3

Abstract

Background Little is known about the effects of commuting physical activity on biological cardiovascular risk factors although such knowledge may form an important basis for interventions aimed at reducing cardiovascular disease (CVD) by increasing physical activity. We examined the associations between commuting, leisure time and total physical activity and biological risk factors for CVD. Design A cross-sectional study of men and women, who participated in a health screening programme. Methods The study population comprised persons aged 30–60 years from a population-based random sample, response rate 53% (n = 6,906). Weight, height, waist circumference and blood pressure were measured and blood samples were collected. Physical activity was assessed by a self-administered questionnaire. Results Time spent on commuting, leisure time and total physical activity was positively associated with high-density lipoprotein cholesterol and negatively associated with low-density lipoprotein cholesterol, triglycerides, waist circumference and body mass index. Time spent on total physical activity was negatively associated with total cholesterol and diastolic blood pressure. Among men there was no relationship between time spent on physical activity and systolic blood pressure. Time spent on commuting physical activity and total physical activity was negatively associated with systolic blood pressure among women. Conclusion Commuting physical activity, independent of leisure time physical activity, was associated with a healthier level of most of the cardiovascular risk factors. An increase in commuting physical activity in the population may therefore reduce the incidence of CVD.

Keywords

ExerciseEpidemiologyPreventionCardiovascular diseaseRisk factors

Abbreviations

CVD

Cardiovascular disease

HDL

High-density lipoprotein

LDL

Low-density lipoprotein

BP

Blood pressure

BMI

Body mass index

FFQ

Food frequency questionnaire

E%

Energy percentage

Introduction

Physical inactivity increases the risk of cardiovascular disease (CVD) and mortality [16]. A minimum of 30 min of moderate physical activity on most, preferably all days of the week is recommended [7], but in many countries a large part of the population is physically inactive at leisure. It is suggested that in the EU an average of 41% of the adult population do not meet the physical activity recommendations [8].

Lack of time is an important barrier for being physically active [9]. Therefore some individuals might prefer activities that can be incorporated into daily living activities. Physical activity in transportation to and from work could be such an activity for some individuals, as some might find it less time consuming and more convenient to be physically active during commuting than at leisure.

The relationship between leisure time physical activity and cardiovascular risk factors has been investigated in a number of studies. However, only three studies [1012] have investigated the relationship between commuting physical activity and biological cardiovascular risk factors. These studies were not adjusted for diet, which potentially is an important confounder. Two of the studies were conducted in China and this limits the generalizability of the results to European countries. The association between commuting physical activity and cardiovascular mortality has been investigated only in one study, which showed an inverse relationship among women but no association among men [6]. Another study showed that bicycling to and from work was associated with reduced all-cause mortality [13].

In this study we investigated the relationship between self-reported commuting physical activity and leisure time physical activity and serum lipids, blood pressure and anthropometric measurements taking into account potential confounders including dietary components. Our findings may support that public health strategies aiming to reduce CVD by increasing physical activity could benefit from focusing on commuting physical activity besides leisure time physical activity.

Methods

Subjects and study design

We used data from the baseline investigation in the Inter99 study, which is a lifestyle (smoking, physical activity and diet) intervention study aiming to reduce the incidence of ischaemic heart disease in the general population. The data used in this study was collected before the intervention. The study population comprised 61,301 individuals born in 1939–1940, 1944–1945, 1949–1950, 1954–1955, 1959–1960, 1964–1965 and 1969–1970 and living in 11 municipalities in the south-western part of the Copenhagen County. The individuals were drawn from the Civil Registration System in which a unique 10-digit number registers all inhabitants in Denmark. An age- and sex-stratified random sample of 13,016 individuals was drawn from the study population. Eighty-two persons had died or could not be traced. Of the remaining 12,934 persons 6,906 (53.4%) participated in the study. Of these, 122 were excluded because of alcoholism, drug abuse or linguistic barriers, leaving 6,784 (52.5%) for analyses (3,482 women, 51.3 %). When comparing the responders to the 13,016 individuals in the random sample there was a larger proportion of women among the responders. Further, there was a lower proportion of men aged 40 years or less and a lower proportion of women aged 30 years among the responders. There was no difference in hospital contacts due to all causes between responders and non-responders. Compared to non-responders responders had less hospital contacts because of ischaemic heart disease, CVD and diabetes. The Inter99 study has been described in detail elsewhere [14]. Written informed consent was obtained from all the participants. The Inter99 study was approved by Ethics Committee.

Assessment of dependent variables

The subjects participated in a health screening at Research Centre for Prevention and Health. They provided fasting blood samples for assessment of total cholesterol, high-density lipoprotein (HDL) cholesterol and triglycerides. The blood samples were stored in a refrigerator and were sent daily to the laboratory for analysis. Total cholesterol, HDL cholesterol and triglycerides were measured by enzymatic procedures (Boeringer Mannheim, Germany). Low-density lipoprotein (LDL) cholesterol was calculated by Friedewald’s formula [15]. Blood pressure (BP) was measured twice with a mercury sphygmomanometer after 5 min of rest in a lying position; the mean value was used in the analyses. Height was measured without shoes to the nearest cm, weight without a coat to the nearest kg and body mass index (BMI) was calculated (kg/m2). Waist circumference was measured midway between the lower rib margin and iliac crest to the nearest cm. The number of persons with missing data on the dependent variables ranged from 1 (systolic BP) to 128 (LDL cholesterol).

Assessment of physical activity and other independent variables

The independent variables were assessed with a self-administered questionnaire. The participants were asked the following question on commuting physical activity: “How much time do you spend walking, cycling or running on your way to and from work?”, and the answer categories were: less than 15 min; 15–30 min; 30 min to 1 h; 1 h or more; and I do not work at the moment. Those who answered that they did not work were excluded in the variable commuting physical activity. Commuting physical activity was missing for 422 persons and 542 persons were excluded from the variable commuting physical activity as they did not work, leaving 5820 for analysis.

Leisure-time physical activity was assessed with the question: “How many hours a week are you physically active? (include walking, bicycle rides, gardening but exclude transportation to and from work)”. The answer categories were: 0 min; approx 1/2 h/week; approx 1 h/week; approx 2–3 h/week; approx 4–6 h/week; and 7 h/week or more. Leisure-time physical activity was missing for 130 persons leaving 6654 persons for analysis.

We calculated total physical activity by summing responses to the questions on commuting physical activity (converted into minutes per week using a five-day working week) and leisure time physical activity (converted into minutes per week). When the answer categories were intervals the middle value was used, and when the answer categories were open-ended the lowest value was used. The answer category “I do not work at the moment” was assigned the value 0 min in order not to exclude participants who did not work from the variable total physical activity. The variable total physical activity was grouped into four categories: 0–113 min/week (0–2 h/week); 143–225 min/week (2–4 h/week); 255–420 min/week (4–7 h/week); 450–720 min/week (7–12 h/week). Total physical activity was missing for 480 persons leaving 6304 for analysis.

The participants also answered questions about alcohol consumption and smoking habits, and they answered a 198-item food frequency questionnaire (FFQ) from which sodium intake, dietary fibre intake and energy percentage (E%) saturated fat was calculated [16].

Statistical analysis

Data were analysed using SAS statistical software, version 8.2 (SAS Institute Inc., Cary, NC, USA). The associations between the risk factors for CVD and commuting, leisure time and total physical activity were examined using multiple linear regression analysis. All outcome variables were logarithmically transformed to improve the variance homogeneity and the normal distribution of the residuals. Back-transformation of the model estimates was performed to provide proportional differences.

The results were adjusted for potential confounding by including age, sex, smoking, alcohol intake, dietary fibre intake and E% saturated fat as covariates in all models. Additionally, sodium intake was included in models with BP as the outcome, and BMI was included in models with waist circumference as the outcome. The analyses of commuting physical activity were adjusted for leisure time physical activity and vice versa. In the analyses of serum lipids and BP adjustment for BMI and waist circumference was performed only in secondary analyses, as they are potential intermediate factors in the causal chain between physical activity and CVD risk factors.

We assessed linearity between continuous variables (alcohol intake, dietary fibre intake, E% saturated fat, sodium intake) and the outcome variables by including quadratic and cubic terms in the analyses. They were kept in the model if they were significant. Interaction terms between physical activity and sex, and physical activity and age were also examined. We only observed significant interaction for systolic BP (P = 0.02 for the interaction between sex and total physical activity; P = 0.05 for the interaction between sex and commuting physical activity); therefore, only the analyses of systolic BP were stratified by sex. We used a 5% significance level.

We excluded persons who used lipid-lowering medication (n = 90) in the analyses of serum lipids and persons who used blood-pressure-lowering medication (n = 470) in the analyses of BP. They were excluded as their artificially lowered values could bias the associations.

We also excluded persons with high triglyceride values (>10 mmol/l, n = 13), high diastolic BP (≥95 mmHg, n = 974) and high systolic BP (≥160 mmHg, n = 503) in the analyses of triglycerides, diastolic BP and systolic BP, respectively, as the association with physical activity might be different in the upper range of these CVD risk factors.

Results

Population characteristics according to commuting physical activity are shown in Table 1. Almost two thirds (63.6%) of the participants spent less than 15 min per day on commuting physical activity. More women than men were physically active during commuting. HDL cholesterol increased and triglycerides decreased with increasing level of commuting physical activity. There was a tendency towards lower levels of LDL cholesterol, total cholesterol, diastolic BP, waist circumference and BMI with increasing commuting physical activity. Participants with a high level of commuting physical activity had a higher sodium intake, ate more fibre and less saturated fat, and fewer were smokers compared to those with a lower level of physical activity.
Table 1

Population characteristics according to commuting physical activity among participants in the Inter99 studya

Characteristics

Commuting physical activity

<15 min/day

15–30 min/day

30–60 min/day

≥60 min/day

No., (%)

3701 (63.6)

1332 (22.9)

549 (9.4)

238 (4.1)

Age, year

45.5 (7.7)

45.3 (7.9)

46.0 (7.8)

46.1 (7.5)

Women, (%)

44.9

62.9

56.7

52.5

High-density lipoprotein cholesterol, mmol/lb

1.4 (0.4)

1.5 (0.4)

1.5 (0.4)

1.6 (0.4)

Low-density lipoprotein cholesterol, mmol/lb

3.5 (1.0)

3.4 (0.9)

3.4 (1.0)

3.4 (0.9)

Total cholesterol, mmol/lb

5.5 (1.1)

5.4 (1.0)

5.4 (1.1)

5.4 (0.9)

Triglycerides, mmol/lb

1.3 (0.9)

1.2 (0.8)

1.2 (0.8)

1.1 (0.6)

Systolic blood pressure, mmHgc

126.8 (13.3)

124.8 (13.5)

126.0 (13.4)

126.4 (13.7)

Diastolic blood pressure, mmHgc

79.1 (8.2)

78.2 (8.5)

78.5 (8.4)

78.4 (8.5)

Waist circumference, cm

87.8 (13.2)

83.7 (13.0)

84.8 (12.3)

83.1 (11.2)

Body mass index, kg/m2

26.5 (4.5)

25.7 (4.7)

26.0 (4.3)

25.3 (3.6)

Saturated fat, energy %

12.7 (3.5)

12.2 (3.4)

11.9 (3.6)

11.9 (3.5)

Dietary fibre intake, g/day

24.8 (10.2)

25.8 (11.4)

28.0 (11.2)

28.7 (11.6)

Alcohol intake, units/week

10.6 (12.8)

9.2 (12.5)

9.1 (10.4)

10.3 (11.6)

Sodium intake, g/day

3.8 (1.4)

3.7 (1.5)

3.8 (1.3)

4.0 (1.6)

Daily smokers, (%)

35.8

34.2

25.6

25.7

aIf not specified otherwise data are presented as mean ± SD

bParticipants who report use of lipid-lowering medication (n = 90) are excluded. For triglycerides participants with triglycerides > 10 mmol/l (n = 13) are also excluded

cParticipants who report use of blood-pressure-lowering medication (n = 470) are excluded. For systolic BP participants with values ≥ 160 mmHg (n = 503) are also excluded and for diastolic BP participants with values ≥ 95 mmHg (n = 974) are excluded

HDL cholesterol, LDL cholesterol and triglycerides all showed significant and favourable associations with time spent on commuting, leisure time and total physical activity in adjusted analyses (Table 2). Compared to the least active individuals spending more than 60 min a day on commuting physical activity had on average a 8% higher HDL cholesterol, a 3% lower LDL cholesterol and a 13% lower triglycerides. There was no association between total cholesterol and time spent on commuting physical activity, but a negative and significant association between total cholesterol and time spent on total physical activity. Additional adjustment for BMI and waist circumference attenuated the associations between commuting, leisure time and total physical activity and HDL cholesterol and triglycerides, whereas the associations with LDL cholesterol and total cholesterol disappeared (data not shown).
Table 2

The proportional difference in serum lipids associated with physical activitya

 

Serum lipidsb

High-density lipoprotein cholesterol (mmol/l)

Low-density lipoprotein cholesterol (mmol/l)

Total cholesterol (mmol/l)

Triglycerides (mmol/l)

β (95%CI)a

β (95%CI)a

β (95%CI)a

β (95%CI)a

Commuting physical activity (min/day)

N = 5400

N = 5345

N = 5400

N = 5391

    <15

1

1

1

1

    15–30

1.04 (1.02–1.05)

0.98 (0.96–1.00)

0.99 (0.98–1.00)

0.96 (0.93–0.99)

    30–60

1.03 (1.00–1.05)

0.98 (0.95–1.00)

0.99 (0.97–1.00)

0.96 (0.91–1.00)

    ≥60

1.08 (1.04–1.11)

0.97 (0.94–1.01)

0.99 (0.96–1.01)

0.87 (0.81–0.93)

    P-valuec

P < 0.0001

P = 0.0089

P = 0.2963

P < 0.0001

Leisure time physical activityd (h/week)

N = 5400

N = 5345

N = 5400

N = 5391

    0–1/2

1

1

1

1

    1

1.02 (0.99–1.05)

1.02 (0.98–1.05)

1.01 (0.98–1.03)

0.95 (0.89–1.01)

    2–3

1.03 (1.00–1.06)

1.00 (0.97–1.04)

1.00 (0.98–1.02)

0.90 (0.95–0.95)

    4–6

1.06 (1.03–1.09)

0.99 (0.96–1.02)

0.99 (0.97–1.01)

0.85 (0.80–0.90)

    ≥7

1.09 (1.06–1.12)

0.98 (0.94–1.01)

0.98 (0.96–1.01)

0.81 (0.76–0.86)

    P-valuec

P < 0.0001

P = 0.0120

P = 0.0630

P < 0.0001

Total physical activity (h/week)

N = 5842

N = 5779

N = 5842

N = 5832

    0–2

1

1

1

1

    2–4

1.03 (1.01–1.06)

0.99 (0.97–1.01)

1.00 (0.98–1.01)

0.93 (0.89–0.97)

    4–7

1.07 (1.05–1.10)

0.96 (0.94–0.98)

0.98 (0.97–1.00)

0.87 (0.83–0.90)

    7–12

1.12 (1.09–1.15)

0.95 (0.92–0.98)

0.98 (0.96–1.00)

0.79 (0.75–0.83)

P-valuec

P < 0.0001

P < 0.0001

P = 0.0063

P < 0.0001

Adjusted for age, sex, smoking, alcohol intake, dietary fibre intake, E% saturated fat, leisure time physical activity (only when the independent variable is commuting physical activity) and commuting physical activity (only when the independent variable is leisure time physical activity)

aThe β-coefficients and 95% confidence intervals are back-transformed coefficients from multiple linear regression models using log-transformed dependent variables

bParticipants who report use of lipid-lowering medication (n = 90) are excluded. For triglycerides participants with triglycerides > 10 mmol/l (n = 13) are also excluded

cThe P-value is for test for equality

dThis does not include commuting physical activity

There was a no association between diastolic BP and commuting and leisure time physical activity, but a significant negative association between diastolic BP and time spent on total physical activity (Table 3). Individuals who were physically active 7 to 12 h per week had a 2% lower diastolic BP compared to those who were active less than 2 h per week. Among women there were significant negative associations between systolic BP and commuting and total physical activity, but no association with leisure time physical activity (Table 3). Compared with the least active women spending more than 60 min a day on commuting physical activity had a 2% lower systolic BP. There were no significant associations between systolic BP and time spent on physical activity among men. Additional adjustment for BMI and waist circumference removed the significant associations with diastolic and systolic BP, and the association between total physical activity and systolic BP among men became significant but positive (data not shown).
Table 3

The proportional difference in blood pressure associated with physical activitya

 

Blood pressure (mmHg)b

Diastolic blood pressure (mmHg)

Systolic blood pressure (mmHg)

β (95%CI)a

Women β (95%CI)a

Men β (95%CI)a

Commuting physical activity (min/day)

N = 4542

N = 2465

N = 2405

    <15

1

1

1

    15–30

1.00 (0.99–1.00)

0.99 (0.98–1.00)

1.00 (0.99–1.01)

    30–60

0.99 (0.98–1.00)

0.99 (0.98–1.01)

1.00 (0.99–1.02)

    ≥60

0.99 (0.97–1.00)

0.98 (0.96–1.00)

1.02 (1.00–1.03)

    P-valuec

P = 0.1150

P = 0.0121

P = 0.4706

Leisure time physical activityd (h/week)

N = 4542

N = 2465

N = 2405

    0–1/2

1

1

1

    1

1.00 (0.98–1.01)

1.00 (0.98–1.02)

1.01 (0.99–1.03)

    2–3

1.00 (0.98–1.01)

1.00 (0.98–1.02)

1.00 (0.98–1.02)

    4–6

0.99 (0.98–1.00)

1.00 (0.98–1.02)

1.01 (0.99–1.02)

    ≥7

0.99 (0.98–1.00)

1.00 (0.98–1.02)

1.01 (0.99–1.02)

    P-valuec

P = 0.1695

P = 0.9338

P = 0.3963

Total physical activity (h/week)

N = 4876

N = 2668

N = 2552

    0–2

1

1

1

    2–4

1.00 (0.99–1.01)

1.00 (0.98–1.01)

1.00 (0.98–1.01)

    4–7

0.99 (0.98–1.00)

0.99 (0.98–1.01)

1.00 (0.99–1.01)

    7–12

0.98 (0.97–0.99)

0.98 (0.96–0.99)

1.01 (1.00–1.03)

    P-valuec

P = 0.0075

P = 0.0159

P = 0.1291

Adjusted for age, sex (only diastolic BP), smoking, alcohol intake, sodium intake, dietary fibre intake, E% saturated fat, leisure time physical activity (only when the independent variable is commuting physical activity) and commuting physical activity (only when the independent variable is leisure time physical activity)

aThe β-coefficients and 95% confidence intervals are back-transformed coefficients from multiple linear regression models using log-transformed dependent variables

bParticipants who report use of blood-pressure-lowering medication (n = 470) are excluded. For systolic BP participants with values ≥ 160 mmHg (n = 503) are also excluded and for diastolic BP participants with values ≥ 95 mmHg (n = 974) are excluded

cThe P-value is for test for equality

dThis does not include commuting physical activity

Significant negative associations between commuting, leisure time and total physical activity and BMI and waist circumference were observed (Table 4). Individuals spending between 7 and 12 h a week on physical activity had a 4% lower BMI and a 2% lower waist circumference compared to the least active.
Table 4

The proportional difference in anthropometric measurements associated with physical activitya

 

Anthropometric measurements

Body mass index (kg/m2)

Waist circumference (cm)

β (95%CI)a

β (95%CI) a

Commuting physical activity (min/day)

N = 5484

N = 5478

    <15

1

1

    15–30

0.98 (0.97–0.99)

1.00 (0.99–1.00)

    30–60

0.99 (0.97–0.99)

1.00 (0.99–1.00)

    ≥60

0.97 (0.95–0.99)

0.99 (0.98–1.00)

    P-valueb

P = 0.0005

P = 0.0038

Leisure time physical activityc (h/week)

N = 5484

N = 5478

    0–1/2

1

1

    1

0.98 (0.96–1.00)

1.00 (0.995–1.011)

    2–3

0.96 (0.94–0.97)

1.00 (0.99–1.00)

    4–6

0.94 (0.92–0.96)

0.99 (0.99–1.00)

    ≥7

0.94 (0.92–0.96)

0.99 (0.98–1.00)

    P-valueb

P < 0.0001

P < 0.0001

Total physical activity (h/week)

N = 5944

N = 5938

    0–2

1

1

    2–4

0.98 (0.96–1.01)

0.99 (0.98–1.00)

    4–7

0.98 (0.96–1.01)

0.99 (0.98–1.00)

    7–12

0.96 (0.93–0.99)

0.98 (0.97–0.99)

    P-valueb

P = 0.0022

P = 0.0069

Adjusted for age, sex, smoking, alcohol intake, dietary fibre intake, E% saturated fat, leisure time physical activity (only when the independent variable is commuting physical activity) and commuting physical activity (only when the independent variable is leisure time physical activity). Waist circumference is also adjusted for BMI

aThe β-coefficients and 95% confidence intervals are back-transformed coefficients from multiple linear regression models using log-transformed dependent variables

bThe P-value is for test for equality

cThis does not include commuting physical activity

Discussion

In this study HDL cholesterol, LDL cholesterol and triglycerides were favourably associated with commuting physical activity among men and women. Our results are in accordance with findings in a Finnish study [12], which observed a positive association between commuting physical activity and HDL cholesterol among men and women. LDL cholesterol and triglycerides were not included in their study. In contrast to our results, a Chinese study by Hu et al. [10] found negative associations between commuting physical activity and LDL cholesterol and triglycerides only among men, and a positive association with HDL cholesterol only among women. These results were adjusted for BMI but not for dietary habits, which together with differences in the self-reported measures of physical activity might explain the inconsistencies between our findings and those by Hu et al. [10]. The observed effect of commuting physical activity on serum lipids is important as it supports that being physically active during commuting is likely to protect against CVD.

The study showed that total physical activity and leisure time physical activity had a beneficial effect on HDL cholesterol, LDL cholesterol and triglycerides, and that total physical activity was negatively associated with total cholesterol. This is in accordance with results from training studies, which frequently show an increase in HDL cholesterol, but also a small decrease in LDL cholesterol, triglycerides and total cholesterol [17, 18]. In epidemiological studies the most consistent association is also observed between physical activity and HDL cholesterol [12, 1925] but negative relationships with triglycerides [21, 24, 26] and total cholesterol [12, 24] are also demonstrated. However, contrary to our results, many epidemiological studies do not show an association with LDL cholesterol [10, 23, 27, 28].

We found no associations between commuting and leisure time physical activity and diastolic BP, but a negative association between total physical activity and diastolic BP. Unexpectedly, we observed negative associations between commuting and total physical activity and systolic BP among women only. Both training studies [29] and an observational study by Wareham et al. [30] using a validated objective measurement of energy expenditure have shown that physical activity reduces systolic BP among both men and women. The mechanisms for the blood pressure lowering effect of physical activity are still uncertain, but reduced sympatic activity and increased insulin sensitivity are possible pathways. It is difficult to explain the observed gender differences; although the mechanisms could be different for men and women we find this unlikely. The differences may be caused by gender related measurement errors when assessing physical activity.

Our results on commuting physical activity and BP differ from the findings by Hu et al. [11]. Unexpectedly, they observed that commuting physical activity is positively associated with diastolic BP among women and with systolic BP among both men and women. The inconsistent findings between the studies might be explained by differential adjustment for BMI in the analyses. Our findings on the relationship between commuting physical activity and diastolic BP are in accordance with results from a study by Barengo et al. [12], however, they differ from their findings on the association between commuting physical activity and systolic BP as they also observe a negative relationship for men. Another study [31] showed no associations between commuting physical activity and diastolic and systolic BP.

The finding that commuting physical activity was negatively associated with BMI corresponds with results from a study by Hu et al. [11] and a study by Barengo et al. [12], but another study showed that use of bicycle for transportation was negatively correlated with BMI only among men [31]. Consistent with previous studies [6, 19, 20, 24, 32, 33], the present study found a negative association between total and leisure time physical activity and BMI.

We observed a negative relationship between commuting physical activity and waist circumference, which is in accordance with findings by Barengo et al. [12]. Our results demonstrate a negative association between leisure-time physical activity and waist circumference. Population-based studies on this association are sparse [12, 21, 23, 34] but a longitudinal study by Sternfeld et al. [34] and the study by Barengo et al. [12] support our results.

The relationship between physical activity and BMI and waist circumference could be opposite to what we suggest in this cross-sectional study, so that a high BMI reduces the participation in physical activity. It appears plausible that BMI and physical activity in some cases interact, resulting in a downward spiral, where a low level of physical activity results in increased BMI, which leads to a lower level of physical activity.

Overall, our study showed that commuting physical activity was strongest associated with HDL cholesterol and triglycerides, and in women also systolic BP. Compared to the least active, persons spending at least 60 min per day on commuting physical activity had an 8% higher HDL cholesterol, a 13% lower triglyceride value and a 2% lower blood pressure. Using average values of 1.5 mmol/l for HDL cholesterol, 1.2 mmol/l for triglycerides and 120 mmHg for systolic BP the effect of a commuting physical activity level corresponds to an increase of 0.12 mm/l in HDL cholesterol, a decrease of 0.156 mm/l in triglycerides and a decrease of 2.4 mmHg in systolic BP. The effect of commuting physical activity on HDL observed in this study is likely to have a positive impact on public health as Gordon et al. [35] showed that a 0.26 mmol/l increase in HDL cholesterol decreased CVD mortality with 3.7% in men and 4.7% in women. However, when evaluating the overall public health value of commuting physical activity, the harmful health effects from air pollution and traffic accidents should be taken into consideration.

This study has some limitations. As in other epidemiological studies of physical activity inaccurate assessment of physical activity is an important problem, which is likely to have attenuated the associations. The cross-sectional design is a weakness of the study, as we cannot be sure of the direction of the observed associations. On the other hand, as the study is cross-sectional physical activity and CVD risk factors are measured at the same time, and therefore bias due to seasonal variation in physical activity is reduced. The response rate was relatively low, but as we studied associations this is unlikely to have affected the internal validity of our findings. The low response rate might, however, affect the generalizability of the results. We excluded participants using lipid lowering or blood pressure lowering medication as well as participants with high levels of cholesterol and with high blood pressure from the analyses. This strengthens the validity of the results and thereby also the generalizability of the findings. Consequently, we believe that the findings can be generalised to adults whose blood pressure and lipids are within the normal range.

The large number of participants and the random sampling from the general population are important strengths of the study. It is also a strength that we could adjust for important potential confounders such as diet and smoking. Although socioeconomic status could confound the relationship between physical activity and biological CVD risk factors, the analyses were not adjusted for this as we believe that socioeconomic status is likely to affect the association through diet, smoking and alcohol consumption, which we have adjusted for. Although data is from a lifestyle intervention study, the intervention does not influence the results as both CVD risk factors and physical activity were measured at baseline before the intervention. Furthermore, it is strength of the study that the relationship between commuting physical activity and CVD risk factors was investigated, as most of the previous studies only examined the association with leisure time physical activity. As two thirds of the participants were not physically active in relation to transportation to and from work, focus on this domain might have a public health impact.

The most important findings emerging from this study are that commuting physical activity, independent of leisure time physical activity, was associated with a healthier level of HDL cholesterol, LDL cholesterol, triglycerides, BMI, waist circumference and among women also systolic BP. Therefore, increased commuting physical activity in the population is likely to have a positive impact on the prevention of CVD. The challenge for public health professionals is in collaboration with town planners, employers and others to develop strategies to increase commuting physical activity.

Acknowledgements

We thank the participants in the Inter99 study and staff from the Research Centre for Prevention and Health. Steering committee of the Inter99 study: Torben Jørgensen, DMSc, Knut Borch-Johnsen, DMSc, Hans Ibsen, DMSc, Troels Thomsen, Ph.D, Charlotta Pisinger, PhD and Charlotte Glümer, PhD.

Copyright information

© Springer Science+Business Media B.V. 2007