Article

Influence of Individual Determinants on Physical Activity at Work and During Leisure Time in Soldiers

A Prospective Surveillance Study

Quantified physical activity is an important parameter for evaluating the risk of the incidence of internal and musculoskeletal disorders. The objective of this study was to evaluate the physical activity of German Soldiers on duty and during leisure time with regard to individual determinants and to evaluate if factors associated with the risk of the incidence of internal or musculoskeletal disorders are of relevance for physical activity.

For this purpose, we conducted activity measurements on 169 subjects (142 male and 27 female). The accelerometer-based activity sensor was worn for 7 consecutive days. The number of steps taken was evaluated as a physical activity marker. Furthermore, the state of health of the subjects was recorded.

We observed that a high body mass index and a large waist circumference were associated with a low activity level. Women were found to be more active than men, particularly during leisure time. Personnel under 25 years of age were more physically active than those between 25 and 50 years of age. Subjects with underlying musculoskeletal disorders were less active than those who had internal disorders or were healthy. Men and overweight people run a higher risk of developing musculoskeletal and internal disorders.

Health promotion should focus on raising the physical activity level with the aim of exerting a positive influence on the associated risk factors.

Individual body dimensions and derived characteristics are reflected in the physical activity level (PAL). For example, it has been shown that the body mass index (BMI) and body fat percentage of children correlate negatively with their physical activity.1 If the PAL is changed by a daily target number of steps being set and the use of pedometers, parameters like BMI, body fat percentage, and weight can be influenced.2 Data have been collected on the classification of physical activity and its importance for a healthy lifestyle.3 It has been postulated that adults over 30 years of age should take an average of more than 10,000 steps a day to reduce their risk of developing disorders associated with inac- tivity.3 Various disorders, e.g., the risk of cardiovascular diseases, can be influenced positively by an adequate activity level.4 Woolf et al have shown that, in addition to physical activity, age and diet have a significant influence on body dimensions and that there is a specific interdependence among these factors.5

Recommendations regarding exercise should be adjusted to specific factors. For example, it is recommended that children take up to 12,000–15,000 steps a day in order to be rated as active and to have a positive effect on their state of health.6 Appropriate guidelines have also been published for older people.7 Moreover, evidence suggests that the positive effects of exercise on body dimensions and health risks also differ between office workers and workers who do physical work.8 Irrespective of this, it was shown that exercise during leisure time correlates with the level of training. Workers who do physical work and craftsmen exercise less during their leisure time than office workers.9 Attitudes towards physical activity differ with gender. They are particularly associated with sociocultural influences. For example, men in southeast Asia show a higher level of physical activity, while no standard gender differences have been observed in Western industrial nations due to different factors influencing working and private lives.9,10 In addition, office workers and executive personnel usually have a lower level of physical activity during working hours, and this entails health risks.11 Soldiers need high PAL in their physical training to be prepared for deployment.12 Studies from the US Army showed that physical activity is also dependent on the specific type of training the Soldier is conducting.13 During deployment, Soldiers showed de­creased physical activity compared to that experienced during predeployment preparation.14

The extent to which physical activity of Soldiers is influenced by factors such as age, BMI, waist circumference, gender, and job is not yet fully clear. For this reason, we conducted this study to evaluate the physical activity of a personnel cohort on duty and during leisure time with regard to individual physical determinants. Our study was also intended to evaluate how far above mentioned risk factors known to influence the incidence of musculoskeletal disorders, injuries, and internal disorders are of relevance in the case of Soldiers.


Methods

The study was conducted with 169 subjects (142 male, 27 female). Twenty-five subjects were officers, 89 were noncommissioned officers (NCO), and 55 were junior enlisted personnel. Ninety-seven subjects mainly did of- fice work, while 72 mainly did physical work. The mean age of the subjects was 27.5±8.3 years, their mean weight was 84.3±16.4 kg, their mean BMI was 26.6±4.3 kg/m2, and their mean waist circumference was 90.2±13.3 cm. The subjects underwent a medical examination. Only Soldiers who did not show any acute limiting disorder that prevents him or her from doing regular service were included. Health records of each participant were evaluated with regard to musculoskeletal disorders, injuries, internal disorders, and regular use of medication.

To be as precise as possible, an activity profile was produced which distinguishes for time spent walking, standing, or lying/sitting covering the duty hours and leisure time of the subjects 24 hours a day over a period of 7 consecutive days.15 Physical activity was measured by using the wearable sensor 

activPAL™ (PAL Technologies, Glasgow, UK). The activPAL™ is an accelerometric sensor that can detect a person’s body position, any changes in this position, the steps taken, and the step frequencies (measuring frequency: 10 Hz), as well as calculate the energy the person consumed on the basis of these data. Reliability of this sensor was tested by Dahlgren et al.16 Following the manufacturer’s instruction, after activation, the sensor was taped onto the midline of the right anterior thigh of the subjects. The participants were also asked to keep a daily activity log, which was correlated with the readings. These results were assigned to distinct measuring points in the activity log, e.g., start of duty, end of duty, breaks, night rest, and exercise. The study was approved by the ethical committee of the University of Rostock (Reference No. A 2009 36) and we obtained a written informed consent from each participant.


Statistical Analysis

Descriptive statistics were determined (mean, standard deviation, minimum, and maximum). The statistical significance test bet­ween the groups was based on an analysis of variance (ANOVA and post hoc least signifi- cant difference test). Paired comparisons were checked using an independent sample t test. The level of significance was established at P ≤ 0.05. F-statistic and eta2 were reported. All the data were computed using the SPSS statistical program version 21.0 (SPSS Inc, Chicago, IL, USA).

Additionally, a bipartite analysis was conducted. First, the categorisation of the study cohort examined according to the various parameters (waist circumference, BMI, age, state of health, gender, and job) was conducted as a basis for assessing whether there was a difference in activity between built up groups. In a second statistical analysis, the relationships between the above-mentioned parameters and significant predictors of physical activity were determined in subsequent correlation and regression analyses. Therefore, the individual parameters were correlated with one another or whether there were specific predictors of physical activity. For this purpose, a correlation analysis was conducted, ie, the calculation of Pearson’s correlation coefficient (r) for metric readings and Spearman’s correlation coefficient for catego- ry readings (CS). A multiple linear regression analysis was conducted. The coefficient of determination (R2), the standardized regression coefficient (ß), and the non-standardized regression coefficient (b) of the predictors together with the Pearson correlation coefficient and semipartial correlation squared (sr2) were reported. The effect size was calculated with G*Power. According to Cohen, the effect size f2 was interpreted as follows: f2 = 0.02 minor effect, f2 = 0.15 moderate effect, f2 = 0.35 major effect.17


Results

The mean awake time for the 169 participants was 16.5 (±1) hours per day. As a mean, they passed 8,500 steps during that time. They were sitting or lying 65.6% (±7.5%) of that time. Only 10.2% (±2.91%) of the time awake was spent walking. and 24.2% (±5.6%) of the time awake was spent standing. Only 41 Soldiers reached the recommended step count of 10,000 steps/day.


Influence of Individual Physical Determinants

Age

Photo


Subjects were divided into 5 groups by age in years: <25, 25 to 30, 31 to 40, 41 to 50, >50. This classification was used to produce a distribution of participants such that there were enough participants in every group. Consid- ering the subjects’ ages, participants under 25 years of age were significantly more active than those between 25 and 50 years of age as shown in Figure 1 and Table 1. There were no significant differences among the subjects aged 25 years or more in the various age groups.

Photo


Table 1 and Figure 1 showed that subjects over 50 years of age were just as active as those under 25 years of age. Increased activity during duty hours correlated positively with subjects of a young age (r = 0.24; P = 0.002). A negative correlation was observed between a high PAL, particularly during leisure time, and older age (r ≤ -0.177; P ≤ 0.021). Older age correlated with a higher weight (r = 0.314; P < 0.001), a larger waist circumference (r = 0.444; P< 0.001), and a higher BMI (r = 0.361; P < 0.001).


Job Position at Work

The number of steps taken per hour during work time decreased among subjects who had a higher or senior job position, shown in Table 2. During leisure time, officers were more active than NCOs, and junior enlisted personnel tended to be more active than NCOs (Table 1). Mean BMI of the NCOs was 26.7 kg/m2, and thus tendentially higher than that of officers (BMI 25.5 kg/m², P = 0.071) and significantly higher than that of junior en- listed personnel (BMI 25.3 kg/m2, P = 0.011). This trend is also evident with regard to waist circumference (officer and junior enlisted, 87 cm; NCO, 92.5 cm; P = 0.015 and P = 0.004, respectively).

Photo


Low-level positions at work correlated negatively with the incidence of internal disorders (CS = -0.339; P < 0.001) and disorders of the locomotor system (CS = -0.302, P < 0.001) and positively with high activity levels on duty (CS = 0.278; P < 0.001.


Gender

Altogether, women showed a higher activity level than men (F = 3.91; P = 0.007) as shown in Figure 2. No relevant difference was observed during duty hours (F = 1.834; P = 0.607). However, during leisure time the number of steps taken was significantly higher among the women than in the male comparison group (F = 13.459; P <0.001) (Figure 2). Female gender correlated with an increase in leisure activity (CS = 0.205; P =.001). Furthermore, a high BMI correlated negatively with female gender (CS =- 0.306; P <.001).

Photo



BMI

As shown in Figure 3, there was a significant difference in activity between Soldiers of normal weight (BMI<25 kg/m2) and overweight Soldiers (BMI between 25 and 30 kg/m2). Subjects of normal weight achieved a higher step count than overweight subjects (F = 2.131; e2 = 170760.611; P = 0.04).

Photo


Participants of normal weight got up to 42 minutes more sleep a day than obese subjects (P = 0.048) and 24 minutes more than overweight subjects (P = 0.012).


Waist Circumference

Soldiers with a large waist circumference were less physically active. Additionally, as shown in Figure 4 and Table 3, significant differences in the number of steps taken during duty hours and leisure time were observed between the groups studied. The difference in leisure time activity was particularly clear. Soldiers with a waist circumference below 90 cm were markedly more active during their leisure time than those with a larger waist circumference. No significant differences in physical activity were observed among sub- jects with a waist circumference above 90 cm. A large waist circumference correlated negatively with female gender (CS = -0.412; P < 0.001) and fixed-term employment contracts (CS = -0.292; P < 0.001). A larger waist circumference correlated positively with a higher BMI (CS = 0.744; P < 0.001), older age (CS = 0.45; P < 0.001), the incidence of internal disorders (CS = 0.327; 

Photo

Photo


P < 0.001) or orthopedic disorders (CS = 0.302, P < 0.001), and use of medication (CS = 0.241; P < 0.002). A negative correlation was observed between a high PAL, particularly during leisure time, and older age (r ≤ -0.177; P ≤ 0.021), waist circumference (r ≤ -0.253; P ≤ 0.001), and BMI (r ≤ -0.158; P ≤ 0.04).


State of Health

The analysis showed that the activity levels of subjects with preexisting chronic internal disorders (hypertension, hyperuricemia (n=22)) did not differ significantly from those of healthy subjects. Subjects with chronic orthopedic (n=24) or chronic internal and orthope- dic disorders (n=15) took approximately 100 steps per hour less than healthy subjects (F=2.981; e2=155.091; P = 0.017) or subjects who only had preexisting internal disorders (P = 0.03).


Subjects with a history of acute injury to the locomotor system (distortion of the upper ankle joint, muscle injuries) were less physically active. Subjects without a history of injury (n=135) were physically active (time spent waking) for 12.5% of their duty hours, while this figure was 10.5% for subjects with a history of injury (n=34) (F = 0.597; P = 0.01). Regarding leisure activity, there were no significant differences between subjects with and without a history of injury. Short-term employment contracts (12 to 24 months) correlated negatively with internal disorders (CS = -0.459; P < 0.001), injuries suffered due to exercise (CS = -0.447, P < 0.001), use of me­dication (CS = -0.302; P < 0.001), and BMI (CS = -0.299; P < 0.001). There was no significant correlation between an increase in body dimensions (BMI and waist circumference) or low levels of physical activity and the number of visits to a doctor. The amount of exercise the subjects performed also did not correlate with the number of visits to a doctor. However, increased physical activity on duty correlated negatively with the incidence of injuries (CS = -0.152; P < 0.049).


Correlation and Regression Analysis of the Determinants-Predictors of Activity

In the following sections we describe the results of the correlation and regression analysis of the influence of known risk factors of internal and orthopaedic diseases on physical activity. We investigated the effect of the risk factors on step count per day as well as on step count per hour in mean, leisure time and on duty. In order of appearance, we identified waist circumference, BMI, age, and gender to be the main predictors of physical activity.


Step Count Per Hour (Mean)

The statistic model was significant at 

F = 4.645 (P = 0.001), and the model equation correlated to R = 0.321 with the criterion variable (R2 = 0.103; R2 adjusted = 0.081; 

f2 = 0.115; power = 0.99). The variance in the mean step count per hour was predicted to a significant extent by the variable 

waist ­circumference (ß = -0.511; b = -6.374; SE = 2.059; P = 0.002; r = -0.283; sr2 = 0.06). Waist circumference accounted for 10.3% of the variance in the mean step count per hour.


Step Count per Hour (During Duty Hours)

The statistic model was significant at F = 3.051 (P = 0.019), and the model equation correlated to R = 0.265 with the criterion variable (R2 = 0.07; R2 adjusted = 0.047; f2 = 0.075; power = 0.944). The variance in the step count per hour during work time was predicted to a significant extent by the variable age (ß =0.215; b = 

-6.351; SE = 2.517; P = 0.013; r = -0.153). Age accounted for 7% of the variance in the step count per hour during duty hours.


Step Count per Hour (During Leisure Time)

The statistic model was significant at F = 3.431 (P = 0.01), and the model equation correlated to R = 0.279 with the criterion variable (R2 = 0.078; R2 adjusted = 0.055; f2 = 0.085; power = 0.96). The variance in the step count per hour during leisure time was predicted to a significant extent by the variable waist circumference (ß = -0.469; b = -6.972; SE = 2.487; P = 0.006; r = -0.284; sr2 = 0.033). Waist circumference accounted for 7.8% of the variance in the step count per hour during leisure time.


Step Count per Day (Mean)

The statistic model was significant at F = 2.653 (P = 0.007), and the model equation correlated to R = 0.362 with the criterion variable (R2 = 0.131; R2 adjusted = 0.082; f2 = 0.151; power = 0.999). The variance in the mean step count was predicted to a significant extent by the variables waist circumference (group) (ß = -0.417; b = -6020.038; SE = 1827.427; P = 0.001; r = -0.298; sr2 = 0.06) and BMI (group) (ß = 0.242; b = 5609.392; SE = 2687.715; P = 0.038; r = -0.128; sr2 = 0.024). Waist circumference (group) and BMI (group) accounted for 13.1% (41.7%: 24.2%) of the variance in the mean step count.


Step Count per Day (During Leisure Time)

The statistic model was significant at F = 4.298 (P < 0.001), and the model equation correlated to R = 0.443 with the criterion variable (R2 = 0.197; R2 adjusted = 0.151; f2 = 0.245; power = 0.99). The variance in the step count during leisure time was predicted to a significant extent by the variables waist circumference (group) (ß = -0.352; b = -4196.461; SE = 1451.402; P = 0.004; r = -0.296; sr2 = 0.009), BMI (group) (ß= 0.221; b = 4221.637; SE = 2134.67; P = 0.05; r = -0.127; sr2 = 0.02), and gender (ß = 0.271; b = 11369.217; SE = 3394.708; P = 0.001; r = 0.326; sr2 = 0.057). Waist circumference (group), BMI (group), and gender accounted for 19.7% (35.2% : 22.1% : 27.1% : 19.7%) of the variance in the step count during leisure time.


Comment

This study shows that some individual determinants correlate with physical activity. Analogous to the study conducted by Woolf et al, it has been shown that a large waist circumference correlates negatively with physical activity.5 Our study has also shown that the leisure time PAL was particularly dependent on the waist circumference.

With regard to age, subjects under 25 years of age and over 50 years of age in particular showed a comparably high level of physical activity. The subjects in the age groups in between were less active, particularly during their leisure time. Subjects who had a history of injury to the locomotor system or suffered from a musculoskeletal disorder were less active than those who had only internal disorders or were healthy. Physical activity is thus limited by musculoskeletal disorders, which was specifically demonstrated for spondyloarthropathy.18 As a consequence of these factors, attention should be paid to the fact that the BMI and waist circumference in- crease in relation to the level of physical inactivity, particularly during leisure time, and constitute risk factors for the development of orthopedic and internal disorders. For example, a high body weight is a significant risk factor for the incidence of gonarthrosis.19,20

Regarding the influence of gender on physical activity, our survey is unlike the findings of Cheah et al10 in that we found women are more active than men, particularly during their leisure time. This also correlated with our finding that the female subjects had smaller waist circumferences and lower BMIs than the male subjects. All recruited women had fixed-term employment contracts. Irrespective of gender, it was revealed that people with fixed-term employment contracts were more likely to have normal parameters (BMI and waist circumference) than employees with contracts of indefinite duration.11

Other working group studies have shown that social status is an especially important predictor and that a low status correlates with poor health behavior.21 In our study, employment contracts of indefinite duration are associ- ated with higher earnings and status within the cohort. Age is a key factor here. Some permanent employees had a lower status than some temporary employees who held executive positions. In an earlier study, we showed that noncommissioned officers, who are comparable to intermediate service employees, had a significantly unhealthier lifestyle than officers.11 The current study has also shown that officers are less physically active during duty hours than other compared groups, but they compensate for this deficiency by increased leisure activity. Beenackers et al observed a greater health awareness in higher social classes.22 Our study has confirmed that in addition to affiliation to a specific job position group; age, waist circumference, BMI, and gender in particular constitute predictors of physical activity.

This study has some limitations. Environmental factors, which also affect physical activity, such as the area in which the subjects live, public transport, and the route they take to work or shopping facilities, have not been analyzed.23 The fact that the subjects were recruited from a single occupational group (Soldiers), the limited number of participants, and the resulting imbalance in their age and gender structure must also be acknowledged as limitations.

In summary, the present study identified waist circumference, BMI, age, and gender as essential predictors of physical activity. A large waist circumference, a high BMI, and the male gender correlate with low PALs, particularly during leisure time. The subjects under 25 years of age and between 50 and 60 years of age were more active than subjects between those age groups.

Subjects with musculoskeletal disorders were less active than healthy subjects or those with internal disorders. Health promotion should focus on raising the physical activity level, particularly during leisure time, with the aim of exerting a positive influence on the associated risk factors. n


References: ref@mci-forum.com

This article is reprinted with the permission of the US Army Medical Department Journal.

Date: 07/21/2016

Source: MCIF 3/16