Results

Parts of this chapter have been reproduced from Clemes et al.105 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.

Internal pilot

Site recruitment for the internal pilot phase, which involved six sites, commenced in August 2017, with participant recruitment across these sites commencing in October 2017. Participants consented into the study and baseline measurements were undertaken between January and August 2018. A delay in recruitment was experienced in one site. Baseline measurements were undertaken in five sites between January and April 2018 and between July and August 2018 in one site.

Data regarding worksite and participant recruitment, randomisation, compliance to the primary outcome and retention rates at 6 months’ follow-up were examined to determine whether or not the trial should progress to the main trial phase. The study’s TSC agreed to base the progression criteria review on the baseline data collected from all six sites, which included site and participant recruitment numbers and activPAL compliance data, and to base the follow-up progression criteria (in terms of retention rates and activPAL compliance at follow-up) on the data collected from the five sites that had completed the 6-month follow-up measures by November 2018. This enabled the progression criteria to be reviewed by the end of 2018, as opposed to waiting until March 2019, when the follow-up measures were due to be completed in the final site, thereby minimising further delays to the trial.

Recruitment, compliance and retention outcomes

Table 4 summarises the recruitment, compliance and retention outcomes observed from the internal pilot sites. Outcomes were reviewed by the TSC on the 11 December 2018; on the basis of the data reviewed, the TSC recommended continuation of the trial.

Main trial

Participant recruitment

Participant recruitment across the remaining 19 sites commenced in January 2019, and baseline measurements for participants consented into the study were undertaken across these sites between February and July 2019.

As experienced in the internal pilot, a number of delays were encountered during the main trial phase. The delays were predominantly associated with challenges across sites in scheduling drivers for their measurement sessions, which required drivers to be released from their duties for 2 hours. The challenges were further exacerbated in the sites randomised to the SHIFT arm, which required drivers to attend the 6-hour structured education session, which was scheduled during work time. As a consequence, delays in the 6-month follow-up measures were experienced across the majority of sites.

Overall cluster and participant numbers

Figure 2 shows the flow of all participants through the study, combining data from the internal pilot and main trial phases. Overall, 386 participants across 25 clusters (i.e. sites) were recruited and consented into the study. The 25 sites were located across the Midlands region of the UK (1502 drivers were employed across these sites), and the sites operated within the transport, retail, hospitality, health-care, pharmaceutical, construction, oil and gas, and automotive industries. Of the 386 participants recruited, 382 participants were randomised into the two trial arms and four participants withdrew prior to randomisation. Thirteen sites (n = 199 participants) were randomised to the control arm and 12 sites (n = 183 participants) were randomised to the SHIFT arm. Between baseline and 6-month follow-up measures, two sites (i.e. one intervention site and one control site) dropped out of the trial. For both sites, this was because of site closures due to the collapse of the contracting companies.

FIGURE 2 A CONSORT diagram of participant flow through the study. (Figure not reproduced here; see the original article.)

Baseline characteristics

Characteristics of the clusters and baseline demographic characteristics of participants within each trial arm, and overall, are shown in Table 5. Table 6 displays the biometric measurements collected from the sample overall, and according to trial arm, at baseline. Table 7 displays the accelerometer-derived measures (i.e. physical activity, sitting time and sleep) collected at baseline. Descriptive comparisons between baseline characteristics of participants completing the trial and non-completers are shown in Appendix 1, Table 42. There were no noticeable differences between completers (i.e. participants who provided valid activPAL data at baseline and at 6 months) and non-completers in terms of cluster size, age, BMI, number of years as a HGV driver and number of steps per day at baseline.

Primary outcome analysis

A mixed-effect linear regression model revealed a statistically significant difference in mean daily step counts at 6 months’ follow-up, in favour of the SHIFT group [SHIFT group mean change: 32 (SD 2939) steps/day; control group mean change: –716 (SD 2109) steps/day], in the complete-case analysis (1008 steps/day, 95% CI 145 to 1871 steps/day; p = 0.022) (Table 8). The ICC for the model was 0.112. Mixed results were seen in the ITT and per-protocol analyses (see Table 8).

Sensitivity analyses

Sensitivity analyses showed similar results to the primary analysis, with significant differences observed between groups in terms of daily step counts measured at 6 months’ follow-up, when including participants with ≥ 2, ≥ 3 and ≥ 4 valid days of activPAL data (see Table 8).

Secondary outcomes: statistical analyses

activPAL-assessed secondary outcomes

Steps per day, time spent sitting, standing and stepping, and time in light physical activity and MVPA across all monitored days

In complete-case analyses, at 6 months’ follow-up, a series of mixed-effect linear regression models revealed statistically significant differences in favour of the SHIFT group in time spent sitting, standing and stepping, and time in MVPA. At 6 months, daily sitting time was significantly shorter in the SHIFT arm (–24 minutes/day, 95% CI –43 to –6 minutes/day), whereas times spent standing (14 minutes/day, 95% CI 2 to 26 minutes/day) and stepping (11 minutes/day, 95% CI 2 to 21 minutes/day) and time in MVPA (6 minutes/day, 95% CI 0.3 to 11 minutes/day) were greater, than in the control arm. There were no statistically significant differences between groups at 6 months’ follow-up in time spent in light physical activity (Table 9). There were no statistically significant differences observed between groups in activPAL variables at 16–18 months’ follow-up (see Table 9).

Steps per day, time spent sitting, standing and stepping, and time in light physical activity and MVPA on workdays

There were no statistically significant differences observed between groups in any activPAL variables measured on workdays at 6 months’ follow-up or at 16–18 months’ follow-up (Table 10).

Steps per day, time spent sitting, standing and stepping, and time in light physical activity and MVPA on non-workdays

In complete-case analyses, at 6 months’ follow-up, mixed-effect linear regression models revealed statistically significant differences in favour of the SHIFT group in daily step counts, time spent sitting and stepping, and time in light physical activity and MVPA on non-workdays. At 6 months, on non-workdays, daily step counts were larger in the SHIFT group than in the control group (2012 steps/day, 95% CI 480 to 3545 steps/day). In the SHIFT group, non-workday sitting time was shorter (–40 minutes/day, 95% CI –65 to –14 minutes/day), whereas time spent stepping was greater (21, 95% CI 6 to 37 minutes/day), as was time in light physical activity (10 minutes/day, 95% CI 2 to 17 minutes/day) and time in MVPA (11 minutes/day, 95% CI 1 to 20 minutes/day), than in the control group. There were no statistically significant differences observed between groups in activPAL variables measured at 16–18 months’ follow-up (Table 11).

Anthropometry and markers of adiposity

There were no statistically significant differences observed between groups in anthropometric measures or markers of adiposity at 6 months’ follow-up, although differences in weight and BMI were marginal, with these differences being in favour of the SHIFT group (weight: –1.2 kg, 95% CI –2.6 kg to 0.1 kg; BMI: –0.35 kg/m², 95% CI –0.75 kg/m² to 0.05 kg/m²) (Table 12).

Biochemical assessments

There were no statistically significant differences observed between groups in any biochemical measures at 6 months’ follow-up (Table 13).

Dietary quality and fruit and vegetable intake

There were no statistically significant differences observed between groups in reported fruit and vegetable intake or overall dietary quality at 6 months’ follow-up or at 16–18 months’ follow-up (Table 14).

Secondary outcomes: descriptive analyses

Further activPAL variables

Across all monitored days and workdays, for the SHIFT group, there were no noticeable differences in time spent sitting in prolonged bouts (> 30 minutes), the number of transitions from sitting to standing and the proportions of time spent sitting, standing and stepping, and the proportion of sitting spent in prolonged bouts, between baseline and 6 months’ follow-up. Similar findings were observed for the control group, except for the time spent sitting in prolonged bouts, which tended to increase at 6 months’ follow-up (Table 15). On non-workdays, the control group exhibited increases in the time spent sitting (and the proportion of sitting) in prolonged bouts at 6 months’ follow-up, relative to baseline. The control group also exhibited an increase in the overall proportion of time spent sitting and a decrease in the proportion of time spent standing on non-workdays at 6 months’ follow-up. For the SHIFT group, no noticeable differences were observed for any variables on non-workdays between baseline and 6 months (see Table 15).

Across all monitored days, workdays and non-workdays, for both groups, there were no noticeable differences in time spent sitting in prolonged bouts (> 30 minutes), the number of transitions from sitting to standing, the proportions of time spent sitting, standing and stepping, and the proportion of sitting spent in prolonged bouts, between baseline and 16–18 months’ follow-up (Table 16).

Sleep duration and quality, subjective situational sleepiness and chronotype

Across all monitored days, between baseline and 6 months’ follow-up, both groups exhibited a decrease in their sleep window duration (defined as the time between ‘lights out’ and out of bed time) and a decrease in their overall sleep duration. These changes appeared to be driven by large reductions in sleep window duration and sleep duration on workdays at 6 months’ follow-up. In contrast, on non-workdays, increases in sleep window duration and sleep duration were observed for both groups at 6 months. There were no noticeable changes in sleep efficiency across any types of day between baseline and 6 months’ follow-up for either group (Table 17). There were no changes in ratings of situational sleepiness or chronotype score between baseline and either follow-up period for both groups (Table 18).

Blood pressure and psychophysiological reactivity

Small reductions in resting systolic and diastolic blood pressure were observed for both groups between the baseline and 6-month follow-up measures, but no noticeable changes were observed in resting heart rate for either group (Table 19). Mean blood pressure and heart rate measures increased during the mirror tracing task; however, the differences between resting values and values recorded during the task tended to be smaller for both groups during the 6-month follow-up measures (see Table 19). There were no differences in perceived stress ratings during this task between baseline and follow-up for either group. The control group tended to have fewer errors while undertaking this task at 6-months follow-up, whereas no evidence of a change in performance was observed for the SHIFT group (see Table 19).

Cognitive function

There were no noticeable differences in reaction times, measured using the Stroop test, between baseline and 6 months’ follow-up for both groups. No noticeable differences were observed between groups at baseline and 6 months (Table 20).

Functional fitness

No noticeable changes in grip strength were observed between baseline and 6 months’ follow-up in the control group, whereas modest improvements in grip strength for both hands were observed following completion of the intervention in the SHIFT group (Table 21).

Mental well-being

There were no noticeable differences in self-reported scores for symptoms of anxiety, depression or social isolation between baseline and 6 months’ follow-up for both groups, and similar findings were also observed for symptoms of anxiety and depression at 16–18 months’ follow-up. There was a tendency in both groups for perceived social isolation scores to increase marginally at 16–18 months’ follow-up (Table 22). No noticeable differences were observed between groups at any assessment point.

Musculoskeletal symptoms

Table 23 provides a summary of the prevalence of musculoskeletal discomfort reported in the past month for each body site, along with discomfort scores by body region, reported over the three time points. There was a tendency for the prevalence of musculoskeletal discomfort across the majority of body sites to decrease at the two follow-up assessments in both groups, with similar changes in prevalence occurring between groups. Similarly, there were no noticeable differences in discomfort scores (i.e. upper extremity, lower extremity and overall) between baseline and 6 months’ follow-up, and between baseline and 16–18 months’ follow-up, for both groups. No noticeable differences in discomfort scores were observed between groups at any assessment point.

Work-related psychosocial variables

Table 24 provides a descriptive summary of a range of work-related psychosocial variables assessed across the three time points. There were no noticeable differences in any of the outcome measures (i.e. work engagement, occupational fatigue, perceived job satisfaction and performance, sickness absence, presenteeism, perceived work ability and perceived job demands) between baseline and 6 months’ follow-up, and between baseline and 16–18 months’ follow-up, for both groups. No noticeable differences in any measure were observed between groups at any assessment point.

Driving-related safety behaviour

There were no noticeable differences in self-reported driving-related safety behaviour between baseline and 6 months’ follow-up, and between baseline and 16–18 months’ follow-up, for both groups (Table 25). No noticeable differences were observed between groups at any assessment point.

Health-related quality of life

There were no noticeable differences in perceived markers of HRQoL between baseline and 6 months’ follow-up, and between baseline and 16–18 months’ follow-up, for both groups (Table 26). No noticeable differences were observed between groups at any assessment point.

Lifestyle health-related behaviours and risk measures

There were no noticeable differences in reported alcohol intake between baseline and 6 months’ follow-up, and between baseline and the final follow-up, for both groups (Table 27). No noticeable differences were observed between groups at any assessment point.

There were no differences in QRISK3 scores between baseline and 6 months’ follow-up for both groups (see Table 27). Likewise, there were no noticeable differences in QRISK3 scores between the control group and the SHIFT group at any assessment point. When examining the proportion of participants with an estimated CVD risk of ≥ 10% over the next 10 years, 23.6% of control participants fell into this category at baseline, with this proportion increasing to 26.4% at 6 months. In contrast, 24.3% of participants in the SHIFT group exhibited a ≥ 10% risk of CVD over the next 10 years at baseline, but this fell slightly to 23.4% at 6 months.

Table 27 also provides a summary of the smoking prevalence reported by participants at each time point. There was a tendency for a higher smoking prevalence to be seen in the control group, than in the SHIFT group, across all assessment points. Between baseline and 6 months, two participants in the control group and one participant in the SHIFT group reported changing from a past smoker to a current smoker (i.e. re-starting smoking), and four participants in the SHIFT group reported stopping smoking. No control participants reported stopping smoking between baseline and 6 months. The average number of cigarettes smoked per day by current smokers in the SHIFT group (baseline, n = 16; 6 months, n = 13) was 14 (SD 5) cigarettes per day at both baseline and 6 months’ follow-up. The average number of cigarettes smoked per day by current smokers in the control group (baseline, n = 27; 6 months, n = 29) was 14 (SD 6) cigarettes per day at baseline and 13 (SD 7) cigarettes per day at 6 months.

Between baseline and 16–18 months’ follow-up, one participant in the SHIFT group reported re-starting smoking at and three participants (control group, n = 1; SHIFT group, n = 3) reported stopping smoking. The average number of cigarettes smoked per day by current smokers in the SHIFT group (baseline, n = 14; 16–18 months, n = 13) was 15 (SD 5) cigarettes per day at both baseline and 16–18 months. The average number of cigarettes smoked per day by current smokers in the control group (baseline, n = 21; 16–18 months, n = 20) was 14 (SD 6) cigarettes per day at both baseline and 16–18 months.

COVID-19: impact of a temporary change in driving hour regulations on SHIFT participants

Participants

Of the 220 participants who were still enrolled in the study in May 2020, 91 (41.4%; control group, n = 48; SHIFT group, n = 43; 99% male) participants completed an additional online questionnaire that captured data on the effect of the pandemic on their working hours, mental well-being and health-related behaviours. The questionnaire was completed during the UK’s first national lockdown. At the time of completing the questionnaire, 20 (22%) participants [control group, n = 15 (31%); SHIFT group, n = 5 (12%)] were on furlough, and 44 (48%) participants [control group, n = 32 (67%); SHIFT group, n = 12 (28%)] reported being on furlough at some point during the pandemic.

There were no statistically significant differences in questionnaire responses between intervention or control participants, or between participants who were/had been furloughed and participants not furloughed. As a result, the responses received for the COVID-19 questionnaire are presented for the group as a whole. The only measure where a difference was reported was ‘sleep duration in the past 14 days’. Participants who had been or who were still on furlough reported a longer sleep duration [median 7.0 (IQR 5.5–8.5) hours/night] than participants who were not furloughed [median 6.5 (IQR 5.5–7.5) hours/night].

Baseline characteristics and measures of the 91 participants completing the COVID-19 questionnaire did not differ significantly from the remainder of the participants within the SHIFT study, and this suggests that the subsample of 91 participants are largely representative of the wider sample. Specifically, there were no differences at baseline between participants completing the COVID-19 questionnaire and the wider sample in terms of age, duration working as a HGV driver, duration working for DHL Supply Chain, hours worked per week, BMI, per cent body fat, waist circumference, self-reported symptoms of anxiety and depression, musculoskeletal complaints, physical activity levels, and sleep duration and efficiency.

Working hours and activity levels before and during the pandemic

Participants reported no changes to their working, driving, in-cab waiting or rest hours during the pandemic. Similarly, participants reported no changes in the time spent sitting, standing and walking/moving around on a workday during the pandemic (Table 28).

Anxiety, depression and fatigue before and during the pandemic

In comparison with baseline, there was a tendency for participants who completed the additional COVID-19 questionnaire to report lower levels of symptoms of anxiety during the pandemic (Table 29). There were no differences observed in symptoms of depression, levels of acute and chronic fatigue or intershift recovery during the pandemic, compared with baseline, for this subsample (see Table 29).

The impact of exposure to green space on anxiety, depression and fatigue before and during the pandemic

Within the online questionnaire, 72% (n = 65) of participants reported that they regularly spent time in nature (e.g. spending time in their garden/allotment, in parks and woodland, at the coast and in open green spaces) prior to the onset of the pandemic. Data collected at baseline revealed that participants who reported spending time in nature also reported significantly lower amounts of chronic fatigue associated with their work than participants who reported that they did not spend time in nature (n = 25) [median chronic fatigue score: 23.3 (IQR 10.0–40.0) vs. 43.3 (IQR 26.7–66.7); p = 0.008]. There were no other differences between groups in terms of other markers of fatigue and symptoms of anxiety and depression at baseline.

During the pandemic, 78% (n = 70) of participants reported spending time in nature. Examining data from participants working at the time of completing the online questionnaire revealed that participants who reported spending time in nature (n = 51) also reported significantly lower amounts of chronic and acute fatigue associated with their work than participants who reported that they did not spend time in nature (n = 17) [median chronic fatigue score: 20.0 (IQR 3.3–41.7) vs. 51.7 (IQR 35.8–65.0), p = 0.002; median acute fatigue score: 40.0 (IQR 16.7–56.7) vs. 55.0 (IQR 47.5–79.2), p = 0.009]. The differences between groups for intershift recovery were marginal, but in favour of the group spending time in nature [median intershift recovery score: 63.3 (IQR 50.0–83.3) vs. 50.0 (IQR 36.7–66.7); p = 0.06]. There were no differences between groups in symptoms of anxiety and depression reported during the pandemic.

The impact of the pandemic on drivers’ lifestyle health-related behaviours

Twenty-three (25%) participants reported engaging in a new form of physical activity since the COVID-19 outbreak. New activities reported by participants included cycling (n = 10), walking (n = 5), gardening (n = 3), running (n = 2), weights at home (n = 2), boxing (punchbag training) (n = 1), exercises at home (n = 1), home workouts with a personal trainer (n = 1) and DIY (‘do it yourself’) (n = 1) (note that some participants reported more than one new activity). Seven of 47 (15%) participants who had not been furloughed reported engaging in a new activity, compared 16 of 44 (36%) participants who had been furloughed during the pandemic.

Twenty per cent of participants reported that their consumption of snacks (e.g. cakes, biscuits, crisps, chocolate and sweets) had decreased during the pandemic, whereas 19% reported an increase. Twenty-three per cent of participants reported that their consumption of fresh fruit and vegetables had increased during the pandemic, whereas 7% reported a decrease. Thirty-four per cent of participants reported not being able to access healthy food while at work during the pandemic, whereas 45% reported accessing healthy food, all stating that they brought their own from home.

Five per cent of participants reported a change to their smoking status during the pandemic. One participant reported starting smoking, two participants reported smoking less and two participants reported stopping smoking. Five per cent of participants reported a decrease in their alcohol intake during the pandemic, whereas 26% reported an increase.

Twenty-seven per cent of participants reported that their sleep duration had increased during the pandemic, whereas 13% reported a decrease. Of the participants reporting an increase in sleep duration, 23 of 25 participants had been furloughed. Participants currently furloughed at the time of completing the questionnaire (n = 20) reported a median sleep duration over the past 14 days of 8 (IQR 6–10) hours per day. Participants currently working (n = 70) reported a median sleep duration over the past 14 days of 7 (IQR 6–8) hours per day.

The impact of involvement in the SHIFT study on health behaviours during the pandemic, and lifestyle changes experienced

Participants were asked within the online questionnaire whether or not they felt that participating in the SHIFT study had given them the right knowledge to maintain a healthy lifestyle during the COVID-19 restrictions. A total of 63% of both intervention and control participants answered ‘yes’ to this question. Overall, 63% of both intervention and control participants answered ‘yes’. A range of qualitative quotes were provided by respondents on how the study had helped them maintain a healthy lifestyle during the pandemic. The responses received were similar between intervention and control participants, and largely centred around an increased understanding of the importance of activity and a better diet. The quotes are shown in Appendix 2, Box 1.

When asked within the questionnaire whether or not they had experienced any changes in lifestyle and/or work that had either a positive or negative impact on health, 40 (44%) participants answered ‘yes’. Of them, 20 (22%) participants reported that these changes had a positive impact and 20 (22%) reported that these changes had a negative impact. A range of qualitative quotes were provided from respondents regarding changes in lifestyle experienced (see Appendix 2, Box 2). Participants reporting positive changes tended to refer to having time off work as a result of being furloughed and, therefore, having more time to be physically active. Participants reporting negative changes tended to refer to a lack of access to facilities (e.g. gyms, swimming pools) to enable them to be active, increased snacking behaviours due to being at home and reductions in the overall quality of their diets due to limited food choices.

Adverse events

No serious adverse events were reported during the trial.