You are here

  • Home
  • Archive
  • Volume 11, Issue 5
  • Detailed descriptions of physical activity patterns among individuals with diabetes and prediabetes: the Lolland-Falster Health Study

Article Text

Article menu
Epidemiology/Health services research
Detailed descriptions of physical activity patterns among individuals with diabetes and prediabetes: the Lolland-Falster Health Study
  1. Sofie Rath Mortensen1,2,
  2. Søren T Skou2,3,
  3. Jan Christian Brønd1,
  4. Mathias Ried-Larsen1,4,
  5. Therese Lockenwitz Petersen5,6,
  6. Lars Bo Jørgensen2,3,7,
  7. Randi Jepsen6,
  8. Lars Hermann Tang2,8,
  9. Neda Esmailzadeh Bruun-Rasmussen9,
  10. Anders Grøntved1
  1. 1The Research Unit for Exercise Epidemiology, Department of Sports Science and Clinical Biomechanics, University of Southern Denmark, Odense, Denmark
  2. 2The Research and Implementation Unit PROgrez, Department of Physiotherapy and Occupational Therapy, Næstved, Slagelse, Ringsted Hospitals, Slagelse, Denmark
  3. 3Research Unit for Musculoskeletal Function and Physiotherapy, Department of Sports Science and Clinical Biomechanics, University of Southern Denmark, Odense, Denmark
  4. 4Center for Physical Activity Research, Copenhagen University Hospital, Copenhagen, Denmark
  5. 5Steno Diabetes Center Sjælland, Holbæk, Denmark
  6. 6Lolland-Falster Health Study, Nykøbing Falster Sygehus, Nykøbing Falster, Denmark
  7. 7Department of Physiotherapy and Occupational Therapy, Zealand University Hospital, Roskilde, Denmark
  8. 8Department of Regional Health Research, University of Southern Denmark, Odense, Denmark
  9. 9Center for Epidemiological Research, Nykøbing Falster Hospital, Nykøbing Falster, Denmark
  1. Correspondence to Professor Anders Grøntved; agroentved{at}health.sdu.dk

Abstract

Introduction This study aimed to describe objectively measured physical activity patterns, including daily activity according to day type (weekdays and weekend days) and the four seasons, frequency, distribution, and timing of engagement in activity during the day in individuals with diabetes and prediabetes and compared with individuals with no diabetes.

Research design and methods This cross-sectional study included data from the Danish household-based, mixed rural-provincial population study, The Lolland-Falster Health Study from 2016 to 2020. Participants were categorized into diabetes, prediabetes, and no diabetes based on their glycated hemoglobin level and self-reported use of diabetes medication. Outcome was physical activity in terms of intensity (time spent in sedentary, light, moderate, vigorous, and moderate to vigorous physical activity (MVPA) intensities), adherence to recommendations, frequency and distribution of highly inactive days (<5 min MVPA/day), and timing of engagement in activity assessed with a lower-back worn accelerometer.

Results Among 3157 participants, 181 (5.7 %) had diabetes and 568 (18.0 %) had prediabetes. Of participants with diabetes, 63.2% did not adhere to the WHO recommendations of weekly MVPA, while numbers of participants with prediabetes and participants with no diabetes were 59.5% and 49.6%, respectively. Around a third of participants with diabetes were highly inactive daily (<5 min MVPA/day) and had >2 consecutive days of inactivity during a 7-days period. Mean time spent physically active at any intensity (light, moderate, and vigorous) during a day was lower among participants with diabetes compared with participants with no diabetes and particularly from 12:00 to 15:00 (mean difference of −6.3 min MVPA (95% CI −10.2 to −2.4)). Following adjustments, significant differences in physical activity persisted between diabetes versus no diabetes, but between participants with prediabetes versus no diabetes, results were non-significant after adjusting for body mass index.

Conclusions Inactivity was highly prevalent among individuals with diabetes and prediabetes, and distinct daily activity patterns surfaced when comparing these groups with those having no diabetes. This highlights a need to optimize current diabetes treatment and prevention to accommodate the large differences in activity engagement.

  • diabetes mellitus, type 2
  • exercise
  • epidemiology
  • prediabetic state

Data availability statement

Data are available upon reasonable request. All data used for this study were derived from the Lolland-Falster Health Study(LOFUS). Research groups can apply to the LOFUS steering group for accessto use LOFUS data. Each project must adhere to the rules and regulations onresearch ethics and data protection.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/bmjdrc-2023-003493

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC?

    • Engagement in regular physical activity is a cornerstone of type 2 diabetes management and prevention, but no studies have investigated differences in objectively measured physical activity between individuals with diabetes, prediabetes, and no diabetes (normal blood glucose levels based on glycated hemoglobin).

    WHAT THIS STUDY ADDS?

    • This study found that most individuals with diabetes and prediabetes were insufficiently physically active and a third with diabetes was highly inactive (<5 min MVPA/day).

    • Differences in daily and weekly patterns of physical activity were observed across diabetes status, but after adjusting for body mass index, the differences between individuals with prediabetes and no diabetes were equalized.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE, OR POLICY

    • This study highlights the need to optimize current diabetes treatment and prevention at the individual level and group level to better use resources and accommodate the large differences in engagement in physical activity among individuals with diabetes.

    Introduction

    Regular physical activity is a cornerstone of type 2 diabetes management and prevention.1 Inactive adults have a substantially higher risk of developing type 2 diabetes and a range of other chronic conditions compared with those adhering to the physical activity recommendations.2 3 While the general recommendation is to be physically active throughout the week, a specific recommendation for adults with type 2 diabetes is to spread the activity over at least 3 days per week and have no more than two consecutive days of inactivity.4 This recommendation refers to a whole-day approach, which is considered a way to achieve regular physical activity for individuals with diabetes, who are unable to engage in more structured exercise. Indeed, previous short-term experiments found that increasing volume of activities of daily living favorably affect postprandial glucose metabolism in prediabetes and type 2 diabetes.5 6 Although regular physical activity among individuals with diabetes is important, the whole-day physical activity pattern is largely unknown. Most population-based studies used self-reported instruments to assess physical activity and reported that a large proportion of individuals with diabetes did not adhere to physical activity recommendations.7 8 However, numbers in these reports may be inaccurate, for example, due to social desirability and recall bias. Studies using accelerometer-based device measurements are needed to scrutinize the daily physical activity pattern across diabetes status to inform future physical activity interventions that can help to use resources in diabetes treatment and prevention.9

    This study aimed to describe objectively measured physical activity patterns, including daily activity according to day type (weekdays and weekend days) and the four seasons, frequency, distribution, and timing of engagement in activity during the day among individuals with diabetes and individuals with prediabetes and compare these patterns with individuals with no diabetes. We also aimed to investigate whether there were any distinct differences in physical activity patterns across diabetes status, while considering other important diabetes-related determinants of activity such as obesity, comorbidities, and mental well-being.

    Methods

    This study is reported in accordance with the ‘Strengthening the Reporting of Observational Studies in Epidemiology’ checklist.10

    Setting and data sources

    Data were derived from the Danish household-based population study: The Lolland-Falster Health Study (LOFUS) that collected data in a mixed rural-provincial area between February 2016 and February 2020. Inhabitants ≥18 years were randomly selected from the Danish Civil Registration System and invited to participate with their household members of all ages. The data collection in LOFUS encompassed questionnaires, a site visit including physical examinations, and biological samples. Detailed information about the LOFUS study protocol is described by Jepsen et al.11 In continuation of the physical examinations, a subsample was asked to wear accelerometers.12 Between February 2017 and November 2018, families were included if at least one adult and one child agreed to accelerometer assessment. From December 2018 to February 2020, all participants were eligible for inclusion.13

    Written informed consent from participants was obtained at the site visit.11

    Study population

    In total, 7208 adults (above 18 years) participating in LOFUS were eligible to wear accelerometers. The present study included LOFUS participants with valid accelerometer data and information about diabetes status.

    Exposure

    Diabetes status

    Data on glycated hemoglobin (HbA1c) from blood samples were used to classify participants’ diabetes status. Participants were categorized as ‘Having diabetes’ if one of the following criteria were met: (1) HbA1c ≥48 mmol/mol, or (2) HbA1c<48 mmol/mol and self-reported use of antidiabetic medication (insulin and other diabetes medication). ‘Having prediabetes’ was defined as: HbA1c between <48 mmol/mol and ≥39 mmol/mol according to the American Diabetes Association (ADA) and no self-reported use of antidiabetic medication.4 Participants were categorized as having ‘No diabetes’ if HbA1c were <39 mmol/mol and there was no self-reported use of antidiabetic medication.

    Outcome

    Physical activity

    Physical activity was measured using Axivity AX3 (Axivity, Newcastle, UK) accelerometers that were attached to the skin using adhesive plaster and placed on the lower back.12 Axivity AX3 accelerometers have previously shown to be valid when measuring physical activity in individuals with and without functional impairments.14 15 Participants were instructed to wear the accelerometers consecutively for 7 days, including during sleep and water activities. By evaluating acceleration and temperature data from the accelerometer, raw valid wear periods were identified. The intensity cut-points for adults were as follows: light: 100 counts, moderate: 3522, and vigorous: 6016.13 A minimum of 22 hours of wear time was the criterion for valid data for a day. We included participants with minimum three valid weekdays and one valid weekend day of measurement. Time spent in different physical activity intensities was determined by generating ActiGraph counts using 10 s epochs from the raw acceleration16, and time spent sedentary (sitting and lying) was determined using the method by Skotte et al.17 Data were processed as described in Petersen et al.13

    The following physical activity outcomes were included:

    • Sedentary behavior (SB): Hours spent daily on sedentary activity weighted by 5/7 for weekdays and 2/7 for weekend days.

    • Light physical activity (LPA): Minutes spent daily on light intensity activity.

    • Moderate physical activity (MPA): Minutes spent daily on moderate intensity activity.

    • Moderate to vigorous physical activity (MVPA): Minutes spent daily on moderate to vigorous intensity activity.

    • Vigorous physical activity (VPA): Minutes spent daily on vigorous intensity activity.

    Adherence to recommendations

    Adherence to the WHO recommendations of weekly physical activity and SB was assessed with MPA and VPA. The WHO recommends that adults engage in at least 150 min of MPA or 75 min of VPA weekly.18 Importantly, these guidelines imply that time spent in VPA is effectively ‘double-counted’ compared with MPA, reflecting the greater intensity of VPA. Therefore, when calculating adherence on a daily level (necessary due to the varying number of valid measurement days among participants), we derived the total daily MVPA as follows: MVPA=MPA + (VPA*2). MVPA < (150 min/7 days) was categorized as ‘Not following recommendations’ and MVPA ≥ (150 min/7 days) as ‘Following recommendations’.

    Adherence to the ADA recommendations of daily physical activity (engagement in ≥30 min MVPA/day) was calculated by summarizing daily MVPA and categorized into: (1) Sufficient physical activity: ≥30 min MVPA/day, (2) Some physical activity: ≥5 min MVPA/day and <30 min MVPA/day, and (3) Highly inactivity: <5 min MVPA/day.4 19 The cut-off between high inactivity and some physical activity was applied because it has been suggested to provide the minimum clinical important difference among inactive adults.20

    Covariates and variables to describe participant characteristics

    Information about age and sex of the participants came from the Civil Registration System, whereas other background information stemmed from the LOFUS Questionnaire.21

    Self-reported measures

    Marital status was dichotomized into: ‘Married or living with partner’ or ‘Living alone’. Highest level of education was categorized as follows: (1) Primary or lower secondary education, (2) Upper secondary or vocational education, and (3) Higher education. Occupational status was categorized into (1) Employed, (2) Unemployed, (3) Sick leave, (4) Retired, (5) Student and (6) Other.

    Self-reported information on a variety of long-term conditions was used to assess comorbidity in addition to diabetes. The definition of comorbidity was based on the 10 body system groups according to Willadsen et al.22 In addition, participants were asked to add if they had any other condition(s). All ‘other’ conditions were coded into the 10 groups by first author SRM and coauthor LBJ independently following the classification by Tang et al.23

    Psychological stress was obtained with Cohen’s 10-item Perceived Stress Scale.24 Stress was classified as: (1) Low perceived stress (scores from 0 to 13), (2) Moderate perceived stress (scores from 14 to 26), and (3) High perceived stress (scores from 27 to 40).

    Mental well-being was obtained with the WHO-5 Well-Being Index. The scoring of the WHO-5 was done by multiplying the raw score by 4 to obtain a percentage score ranging from 0 to 100.25 A higher score indicated a better mental well-being. Scores <50 were categorized as low mental well-being. Long-lasting chronic pain was reported as Yes or No, and use of selected medication was obtained from the questionnaire. We used information on use of insulin, other diabetes medication, cholesterol-lowering medication, and diuretics.

    Objectively measures

    Participants’ height and weight were obtained at the health examination to calculate body mass index (BMI) (kg/m2). BMI was categorized into: Underweight/normal weight (BMI<25), overweight (BMI≥25–<30), obese class I (BMI≥30–<35), obese class II (BMI≥35–<40), and obese class III (BMI≥40), as defined by the WHO.26 HbA1c were used to classify controlled glycemic level for adults: Controlled glycemic level (HbA1c<53 mmol/mol); Uncontrolled glycemic level (HbA1c≥53 mmol/mol).4

    Statistical analyses

    Prior to commencing the analyses, a statistical analysis plan was developed and stored as openly available (https://osf.io/34t2c/). Some deviations from the plan were made in some of the analyses to fit the models due to low participant numbers in the diabetes group.

    The dstat function in STATA27 was used to describe statistical distributions by diabetes status with standardization of age and sex. Descriptive characteristics of participants with diabetes, prediabetes, and no diabetes were summarized as numbers and proportions or means and standard errors. The distribution and comparison of daily SB, LPA, MPA, VPA, and MVPA in total, during weekdays, and weekend days by diabetes status were estimated with median and quantiles (25th and 75th centiles). Difference in MVPA percentiles between diabetes status by weekdays, weekend days, and season of the year were estimated with coefficients and 95% CI. Adherence to recommendations of physical activity was distributed and displayed with numbers and proportions. Differences across diabetes status were investigated using Wald test within regression models, which varied based on outcome distribution, to adjust for age and sex by testing the null hypothesis that all coefficients are jointly zero.

    The distribution of 0, 1, 2, 3, 4, 5, 6, and ≥7 inactive days and the prevalence of >2 consecutive inactive days were estimated by diabetes status with adjustment for age, sex, and number of days with valid accelerometer data.27 Zero-inflated Poisson regression models were used to predict number of days with inactivity during a 7-day period of measurement by diabetes status adjusted for age and sex and with number of valid days with accelerometer data as exposure time.

    Mixed linear regression models with adjustment for age and sex were used to estimate and display the daily activity profile (mean time spent physically active at any intensity over the waking hours (per 15 min)) of weekdays and weekend days by diabetes status. Savitzky-Golay smoothing filter with an order of 3 and length of 15 was used to generate a smoothed trend based on the point estimates for every 15 min obtained from the mixed model. The order of 3 was chosen to reflect the expected pattern in physical activity data, with the length of 15 determined iteratively to best describe the general trend in the data. Based on visual inspection of the plot, we conducted post hoc analyses of daily activity profiles with additional adjustments of occupational status, BMI, and stress to investigate if differences in daily activity profile could be explained by other major determinants of inactivity. Finally, multiple quantile regression models on daily physical activity intensities and diabetes status were performed with additional adjustments to investigate if any differences by diabetes status were explained by other factors. Therefore, Model 1 was adjusted for age and sex; Model 2 was adjusted for age, sex, and BMI; and Model 3 was adjusted for age, sex, BMI, comorbidities, stress, mental well-being, and chronic pain.

    All statistical analyses were performed using the software STATA BE V.17.0 and R statistical software (R Core Team, Vienna, Austria) V.4.2.2 (November 10, 2022), RStudio (RStudio, Boston, Massachusetts, USA) V.2022.07.2.

    Results

    Of the 3157 participants with valid accelerometer data and information on diabetes status, 181 (5.7 %) participants had diabetes, 568 (18.0 %) participants had prediabetes, and 2408 (76.3 %) participants had no diabetes (flow chart in online supplemental file 1).

    Supplemental material

    Median (25th and 75th centiles) age was 67.8 (60.7–73.8) years among participants with diabetes, 65.1 (54.5–72.2) years among participants with prediabetes, and 51.1 (40.1–65.4) years among participants with no diabetes. The proportion of men was higher among participants with diabetes (59.1 %) compared with participants with prediabetes (46.1 %) or no diabetes (44.7 %). Characteristics of participants with diabetes, prediabetes, and no diabetes, standardized on age and sex, are displayed in table 1. A larger proportion of participants with diabetes were on sick leave or retired, had higher BMI, and more comorbidities compared with participants with prediabetes or no diabetes. Further, participants with prediabetes had higher BMI and more comorbidities compared with participants with no diabetes.

    View this table:
    Table 1

    Characteristics of participants with diabetes, prediabetes, and no diabetes

    Among participants with diabetes, 63.2% did not adhere to the WHO recommendations of weekly MVPA, while 59.5% of participants with prediabetes and 49.6% of participants with no diabetes did not follow the recommendations. The proportion of participants with diabetes who was highly inactive daily (<5 min MVPA/day) was 33.0% (table 2 and online supplemental file 2). The percentage point difference in highly inactive participants with diabetes (reference) compared with prediabetes and no diabetes was −14.7% (95% CI −18.2 to −11.4) and −20.1% (95% CI −25.9 to −15.1), respectively (online supplement 2).

    View this table:
    Table 2

    Distribution of daily SB, LPA, MPA, VPA, and MVPA and adherence to recommendations among participants with diabetes, prediabetes, and no diabetes

    The 25th, 50th, and 75th centiles of MVPA were significantly higher among participants with no diabetes compared with participants with diabetes (difference p25: 6.1 min/day, 95% CI 4.9 to 7.3, difference p50: 11.9 min/day, 95% CI 9.9 to 14.0, and difference p75: 10.0 min/day, 95% CI 4.2 to 15.9) (online supplemental files 3 and 4). No variations in seasonal distribution of daily MVPA by diabetes status were present (online supplemental file 5).

    After adjustment for sex, there was no age-related differences in MVPA (min/day) between participants with diabetes and no diabetes except for a difference in the lowest centile (p25) of MVPA (4.2 min/day difference among participants ≥65 years and 7.9 min/day difference among participants <65 years, p=0.02 for interaction (for more information see online supplemental file 6)).

    Mean time spent physically active at any intensity during a weekday and a weekend day was lower among participants with diabetes compared with participants with prediabetes and no diabetes (figure 1). Participants with diabetes were significantly less physically active in the early afternoon (from 12:00 to 15:00) compared with participants with no diabetes (−6.3 min, 95% CI −10.2 to −2.4, p=0.001). Additional adjustments for BMI, stress, and occupational status showed similar daily activity profiles across diabetes status (online supplemental files 7–9).

    Figure 1
    Figure 1

    Daily activity profiles by diabetes status are presented as mean time (minutes per 15 min interval) spent physically active at any intensity (light physical activity (LPA) was the threshold for detection of physical activity at any intensity) during a weekday and a weekend day adjusted for age and sex. The points represent the raw age-adjusted and sex-adjusted estimates, and the fitted line represents a smoothing trend based on the point estimates. The smoothed trend was generated using a sgolay filtering with an order of 3 and length 15.

    Among participants with diabetes, 33.2% had more than two consecutive days with high inactivity (<5 min/day of MVPA) during a 7-day period (table 3) which is at a rate that is 2.30 (95% CI 1.80 to 2.94) and 1.36 (95% CI 1.12 to 1.66) times higher compared with participants with prediabetes and participants with no diabetes, respectively, after adjustment for age and sex. Predicted number of days with high inactivity during a 7-day period were higher among participants with diabetes (2.2 days, 95% CI 1.98 to 2.37) compared with participants with prediabetes (1.75 days, 95% CI 1.63 to 1.87) and no diabetes (1.47 days, 95% CI 1.40 to 1.54) (online supplemental file 10).

    View this table:
    Table 3

    The prevalence for accumulating 0, 1, 2, 3, 4, 5, 6, and 7 days of high inactivity (<5 min MVPA/day) among participants with diabetes, prediabetes, and no diabetes

    Participants with diabetes had significantly lower median LPA, MPA, MVPA, and higher median SB after adjustments for BMI and other major determinants compared with participants with no diabetes. Additionally, participants with prediabetes had significantly lower median MPA, MVPA, and higher SB compared with participants with no diabetes when adjusting for age and sex. After adjusting for BMI, these differences were no longer significant (table 4). Further, participants with prediabetes had significantly higher median LPA and lower SB after adjustments compared with participants with diabetes (online supplemental file 11).

    View this table:
    Table 4

    Quantile regression models on daily LPA, MPA, VPA, MVPA, and SB by diabetes status with additional adjustment for other major determinants of physical activity

    Discussion

    We found that a large proportion of participants with diabetes and prediabetes were insufficiently physically active. Also, results revealed that over a third of participants with diabetes and prediabetes met the WHO recommendations for weekly physical activity. Participants with diabetes engaged significantly less in physical activity during weekdays and weekend days and had a higher frequency of highly inactive days compared with participants with prediabetes or no diabetes. These differences were evident even after adjustment for other major determinants of physical activity such as BMI and prevalent comorbidities. Participants with prediabetes were also less physically active compared with age-matched and sex-matched participants with no diabetes. However, after adjusting for BMI, these differences were no longer significant.

    Comparison with other studies

    Most prior studies used self-reported instruments and only estimated adherence to the recommendations of physical activity in individuals with diabetes.7 8 28 29 We identified few studies that used device-based measurements among individuals with diabetes or prediabetes to report physical activity intensities and adherence to WHO recommendations. A Danish study by Domazet et al30 found that 62% of individuals with a median age of 61.8 years and newly diagnosed type 2 diabetes met the recommendations. A Swedish population-based study by Hult et al31 found that 43% of 70-year-old adults with diabetes adhered to the WHO recommendations. In our study, participants with diabetes had a median age of 67.8 years and only 36.8% adhered to the WHO recommendations. Furthermore, our study participants were resident in a socioeconomically disadvantaged area of Denmark,32 which could explain the differences in results between the studies as physical activity is typically lower in individuals with low socioeconomic status.33 34 Swindell et al35 found that mean MVPA among overweight and prediabetic women and men was 26.2 min/day and 31.6 min/day, respectively. Participants were 15 years younger and volunteers in a lifestyle intervention which could explain why we found a lower median daily MVPA (14.3 min/day) among participants with prediabetes in our study.35

    Steeves et al,36 using the National Health and Nutrition Examination Survey (NHANES) data from 2003 to 2006, also reported lower physical activity levels among those with diabetes and a noticeable drop in the afternoon, similar to our findings. They found comparable activity levels between individuals with prediabetes and normoglycemic individuals, as well. However, unlike their study, which focused solely on adults over the age of 60 years, our study included a broader range of participants.36 Importantly, comparisons of results between studies using accelerometer-measured physical activity is challenged by a lack of consensus about the method used in the data reduction process.37 38

    Highly inactive days

    Our study provides new insights into physical activity patterns distributed over a week, which have not been addressed in prior studies.4 We found that 33.0% of participants with diabetes were highly inactive (<5 min MVPA/day), and 33.2% had more than two consecutive highly inactive days during a 7-day period, while numbers were 12.8% and 15.1% among participants with no diabetes. Achieving and maintaining a physically active lifestyle can be a challenge for individuals with diabetes, and some may not even view daily activity as a crucial aspect of managing their diabetes.39 Many physical activity intervention programs for adults with type 2 diabetes have been developed, but a limited number of these interventions focus on implementation and maintenance.40 Given the high proportion of participants not adhering to physical activity recommendations in our study, especially among those with diabetes, efforts on an individual level and societal level are needed to promote physical activity and improve health.41 42 Among others, this may be accomplished by offering exercise communities and support and increasing accessibility and flexibility for participation such as digital solutions or group-based interventions.39 43 Our study also revealed that around 20%–40% of participants with diabetes and prediabetes were sufficiently active daily and weekly according to the ADA and WHO recommendations. The observed differences in engagement in physical activity within individuals with diabetes and prediabetes suggest a need of rethinking how diabetes treatment and prevention is delivered to the individual. Personalized medicine has been highlighted as a clinical approach aiming to improve patient health and experience and reducing costs.44 Individuals with diabetes and prediabetes who are sufficiently physically active may not need support from, for example, health professionals, peers, or family to become physically active, however, resources could be spent on supporting these individuals to maintain their physical activity level through easily accessible long-term physical activity interventions with a focus on social support or physical activity monitoring with activity trackers or apps.39 43 45 Furthermore, it is important to note that inactive individuals, especially those with diabetes, may require not only additional, but also more intensive, and long-term support. This could include continuous guidance from healthcare professionals to increase their physical activity levels, sustain these changes over time, and ultimately achieve health-related benefits. Therefore, physical activity interventions for individuals with diabetes should be concentrated on those who will benefit from it, and spare resources for those who will not. Physical activity screening tools might be needed to be able to reach out to those individuals with diabetes who may need extra support to change their physical activity behavior. Wearable accelerometer-based devices could be used as a screening tool to introduce a personalized medicine approach to identify and stratify individuals with diabetes and prediabetes into subgroups based on their habitual physical activity levels and patterns.46 Furthermore, it is also important to include the individual’s preferences and motivation in terms of increasing and maintaining physical activity in their daily life.47 48 These approaches would enable clinicians to treat patients with diabetes and prediabetes individually based on their needs to have a physically active lifestyle.

    Daily physical activity patterns

    Differences in the daily activity profile across diabetes status were also revealed in our study. Participants with diabetes were particularly less physically active during the period from 12:00 to 15:00 compared with participants with no diabetes after adjustments for diabetes-related determinants of activity. Although speculative, a possible explanation of the lower levels of activity among participants with diabetes in the early afternoon could be higher with more prolonged postprandial glucose excursions compared with individuals with prediabetes and individuals with no diabetes.49 Postprandial hyperglycemia and hyperinsulinemia may cause increased fatigue following a meal, which might dampen motivation for activity in the postprandial period.50 51 The results indicate that many individuals with diabetes may have a more inactive daily pattern compared with individuals with prediabetes and individuals with no diabetes. In addition, since total volume of physical activity has been reported being equally strongly associated with cardiometabolic health as MVPA,35 52 a whole-day approach should be considered when increasing physical activity in individuals with diabetes. Focusing on increasing LPA in a whole-day perspective rather than exercise-based MVPA in a short timeframe among inactive individuals with diabetes could offer a seemingly equally effective approach, particularly if they suffer from other determinants such as obesity and comorbidities that prevent them from engaging in exercise-type activities with higher intensities.6 7 Considering the low levels of physical activity among participants with diabetes in the present study, replacing a significant part of the day spent being sedentary with LPA may be more feasible for some to overcome possible barriers of engagement in higher intensity activities. This approach is also in accordance with the WHO recommendations highlighting that doing some physical activity is better than none, because engagement in some physical activity will still be beneficial for the individual’s health.18

    Strengths and limitations of the study

    The present study has several strengths. To the best of our knowledge, this is the first population-based study to detail both the time spent in varying physical activity intensity domains and the weekly distribution of inactive days, specifically contrasting these characteristics across individuals with different diabetes statuses.11 Further, HbA1c measures from the blood samples were used to categorize diabetes status, which also enabled us to include individuals with prediabetes, however, we were not able to distinguish between type 1 and 2 diabetes in this study. Another strength is the use of accelerometry to assess 24-hour physical activity behavior under free-living conditions with a median of six valid days among all participants. Also, we controlled for age-related and sex-related differences in activity patterns across diabetes status in all analyses and performed additional adjustments in our regression models to investigate if factors such as comorbidities, stress, chronic pain, and obesity, that are more prevalent in diabetes, could explain differences in activity across diabetes status.53–55

    This study also has several potential limitations. Lolland-Falster is a socioeconomically disadvantaged area of Denmark,32 and given that low socioeconomic status is associated with higher incidence of type 2 diabetes,56 we would expect the proportion of participants with diabetes to be larger compared with the general population. However, out of 3157 participants 5.7% had diabetes which corresponds with latest available data on diabetes prevalence in Denmark.7 Participation in the LOFUS Study and the accelerometer assessment was voluntary which may have introduced selection bias. The participation rate in LOFUS was highest among the middle-aged population, women, Danish citizens and those from a high socioeconomic status.57 Therefore, the patterns and differences in activity observed may not be representative of other populations. Also, the sample size of participants with diabetes in this study was modest, which might also affect the generalizability of the study. There may be other diabetes-related determinants of physical activity that were not captured by those included that could explain differences in activity patterns across diabetes status. Furthermore, accelerometry measurements used in the study are not able to accurately capture non-ambulatory activities such as resistance training. Lastly, because of the cross-sectional study design, we cannot draw any conclusions on direction or causal nature of the associations.

    Conclusion

    We found that a large proportion of individuals with diabetes and individuals with prediabetes were insufficiently physically active. Most individuals with diabetes engaged less in physical activity during the day in terms of overall daily levels of activity, frequency, and distribution of highly inactive days and timing of engagement in activity compared with individuals with no diabetes. Among individuals with prediabetes, we found that they were less physically active compared with their age-matched and sex-matched counterparts with no diabetes. This difference, however, diminished when adjusting for BMI. Also, we found that more than a third of individuals with either diabetes or prediabetes were engaging in sufficient levels of physical activity. This emphasizes the necessity to tailor diabetes treatment and prevention strategies to the wide-ranging physical activity habits seen within these populations, ensuring resources are used in the most effective manner.

    Data availability statement

    Data are available upon reasonable request. All data used for this study were derived from the Lolland-Falster Health Study(LOFUS). Research groups can apply to the LOFUS steering group for accessto use LOFUS data. Each project must adhere to the rules and regulations onresearch ethics and data protection.

    Ethics statements

    Patient consent for publication

    Ethics approval

    This study involves human participants and was approved by Region Zealand’s Ethical Committee on Health Research (SJ-421) and the Danish Protection Agency (REG-024-2015) and registered in Clinical Trials (NCT02482896). Data storage and management for the present study were approved by the Danish Data Protection Agency through the University of Southern Denmark (Journal nr.: 11.396). Participants gave informed consent to participate in the study before taking part.

    Acknowledgments

    The LOFUS Study, Nykøbing Falster Hospital, Denmark, is a collaboration between Region Zealand, Nykøbing Falster Hospital, and Lolland and Guldborgsund Municipalities. The authors thank LOFUS for making the LOFUS research data available. However, LOFUS bears no responsibility for the analysis, or the interpretation within this study.

    References

    1. ↵
      1. Colberg SR,
      2. Sigal RJ,
      3. Yardley JE, et al
      . Physical activity/exercise and diabetes: A position statement of the American diabetes Association. Diabetes Care 2016;39:2065–79. doi:10.2337/dc16-1728
      OpenUrlFREE Full Text
    2. ↵
      1. Grøntved A,
      2. Rimm EB,
      3. Willett WC, et al
      . A prospective study of weight training and risk of type 2 diabetes mellitus in men. Arch Intern Med 2012;172:1306–12. doi:10.1001/archinternmed.2012.3138
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Aune D,
      2. Norat T,
      3. Leitzmann M, et al
      . Physical activity and the risk of type 2 diabetes: a systematic review and dose–response meta-analysis. Eur J Epidemiol 2015;30:529–42. doi:10.1007/s10654-015-0056-z
      OpenUrlCrossRefPubMed
    4. ↵
      1. American Diabetes Association
      . Standards of medical care in Diabetes—2022 abridged for primary care providers. Clin Diabetes 2022;40:10–38. doi:10.2337/cd22-as01
      OpenUrlCrossRefPubMed
    5. ↵
      1. Benatti FB,
      2. Ried-Larsen M
      . The effects of breaking up prolonged sitting time: A review of experimental studies. Med Sci Sports Exerc 2015;47:2053–61. doi:10.1249/MSS.0000000000000654
      OpenUrlCrossRefPubMed
    6. ↵
      1. Dempsey PC,
      2. Owen N,
      3. Yates TE, et al
      . Sitting less and moving more: improved Glycaemic control for type 2 diabetes prevention and management. Curr Diab Rep 2016;16:114. doi:10.1007/s11892-016-0797-4
    7. ↵
      1. Mortensen SR,
      2. Kristensen PL,
      3. Grøntved A, et al
      . Determinants of physical activity among 6856 individuals with diabetes: a nationwide cross-sectional study. BMJ Open Diab Res Care 2022;10:e002935. doi:10.1136/bmjdrc-2022-002935
    8. ↵
      1. Morrato EH,
      2. Hill JO,
      3. Wyatt HR, et al
      . Physical activity in U.S adults with diabetes and at risk for developing diabetes 2003. Diabetes Care 2007;30:203–9. doi:10.2337/dc06-1128
      OpenUrlAbstract/FREE Full Text
    9. ↵
      1. Steene-Johannessen J,
      2. Anderssen SA,
      3. van der Ploeg HP, et al
      . Are self-report measures able to define individuals as physically active or inactive Med Sci Sports Exerc 2016;48:235–44. doi:10.1249/MSS.0000000000000760
      OpenUrlCrossRefPubMed
    10. ↵
      1. Vandenbroucke JP,
      2. von Elm E,
      3. Altman DG, et al
      . Strengthening the reporting of observational studies in epidemiology (STROBE): explanation and elaboration. Int J Surg 2014;12:1500–24. doi:10.1016/j.ijsu.2014.07.014
      OpenUrlCrossRefPubMed
    11. ↵
      1. Jepsen R,
      2. Egholm CL,
      3. Brodersen J, et al
      . Lolland-Falster health study: study protocol for a household-based prospective cohort study. Scand J Public Health 2020;48:382–90. doi:10.1177/1403494818799613
      OpenUrlPubMed
    12. ↵
      1. Petersen TL,
      2. Brønd JC,
      3. Benfeldt E, et al
      . n.d. Integrity and performance of four tape solutions for mounting Accelerometry devices: Lolland-Falster health study. J Meas Phys Behav;5:224–32. doi:10.1123/jmpb.2022-0024
    13. ↵
      1. Petersen TL,
      2. Brønd JC,
      3. Kristensen PL, et al
      . Resemblance in accelerometer-assessed physical activity in families with children: the Lolland-Falster health study. Int J Behav Nutr Phys Act 2020;17:161. doi:10.1186/s12966-020-01067-7
    14. ↵
      1. Hedayatrad L,
      2. Stewart T,
      3. Duncan S
      . Concurrent validity of Actigraph Gt3X+ and Axivity Ax3 accelerometers for estimating physical activity and sedentary behavior. J Meas Phys Behav 2021;4:1–8. doi:10.1123/jmpb.2019-0075
      OpenUrl
    15. ↵
      1. Clarke CL,
      2. Taylor J,
      3. Crighton LJ, et al
      . Validation of the Ax3 Triaxial accelerometer in older functionally impaired people. Aging Clin Exp Res 2017;29:451–7. doi:10.1007/s40520-016-0604-8
      OpenUrlPubMed
    16. ↵
      1. Brønd JC,
      2. Andersen LB,
      3. Arvidsson D
      . Generating Actigraph counts from raw acceleration recorded by an alternative monitor. Med Sci Sports Exerc 2017;49:2351–60. doi:10.1249/MSS.0000000000001344
      OpenUrlCrossRefPubMed
    17. ↵
      1. Skotte J,
      2. Korshøj M,
      3. Kristiansen J, et al
      . Detection of physical activity types using Triaxial accelerometers. J Phys Act Health 2014;11:76–84. doi:10.1123/jpah.2011-0347
      OpenUrlCrossRefPubMed
    18. ↵
      1. World Health Organization
      . WHO guidelines on physical activity and sedentary behaviour. 2020.
    19. ↵
    20. ↵
      1. Rowlands A,
      2. Davies M,
      3. Dempsey P, et al
      . Wrist-worn accelerometers: recommending ~1.0 mg as the minimum clinically important difference (MCID) in daily average acceleration for inactive adults. Br J Sports Med 2021;55:814–5. doi:10.1136/bjsports-2020-102293
      OpenUrlFREE Full Text
    21. ↵
      1. Egholm CL,
      2. Packness A,
      3. Stokholm J, et al
      . Questionnaire development for the Lolland-Falster health study, Denmark: an Iterative and incremental process. BMC Med Res Methodol 2020;20:52. doi:10.1186/s12874-020-00931-1
    22. ↵
      1. Willadsen TG,
      2. Siersma V,
      3. Nicolaisdóttir DR, et al
      . Multimorbidity and mortality: A 15-year longitudinal Registry-based nationwide Danish population study. J Comorb 2018;8:2235042X18804063. doi:10.1177/2235042X18804063
    23. ↵
      1. Tang LH,
      2. Thygesen LC,
      3. Willadsen TG, et al
      . The association between clusters of chronic conditions and psychological well-being in younger and older people-A cross-sectional, population-based study from the Lolland-Falster health study, Denmark. J Comorb 2020;10:2235042X20981185. doi:10.1177/2235042X20981185
    24. ↵
      1. Cohen S,
      2. Kamarck T,
      3. Mermelstein R
      . A global measure of perceived stress. J Health Soc Behav 1983;24:385–96.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Topp CW,
      2. Østergaard SD,
      3. Søndergaard S, et al
      . The WHO-5 well-being index: a systematic review of the literature. Psychother Psychosom 2015;84:167–76. doi:10.1159/000376585
      OpenUrlCrossRefPubMed
    26. ↵
      1. World Health Organization
      . Obesity: preventing and managing the global epidemic: report of a WHO consultation. Geneva: World Health Organization, 2000.
    27. ↵
      1. Jann B
      . DSTAT: Stata module to compute summary statistics and distribution functions including standard errors and optional covariate balancing, Statistical Software Components. Boston College Department of Economics, 2022.
    28. ↵
      1. Martinez-Harvell G,
      2. Goluboff F,
      3. Rodriguez P, et al
      . Predictors of adherence to physical activity guidelines in patients with diabetes mellitus in the US in 2017: an exploratory analysis. Prim Care Diabetes 2020;14:645–53. doi:10.1016/j.pcd.2020.05.001
      OpenUrlPubMed
    29. ↵
      1. Salman A,
      2. Ukwaja KN,
      3. Alkhatib A
      . Factors associated with meeting current recommendation for physical activity in Scottish adults with diabetes. Int J Environ Res Public Health 2019;16:20.:3857. doi:10.3390/ijerph16203857
      OpenUrl
    30. ↵
      1. Domazet SL,
      2. Tarp J,
      3. Thomsen RW, et al
      . Accelerometer-derived physical activity and sedentary behaviors in individuals with newly diagnosed type 2 diabetes: A cross-sectional study from the Danish nationwide Dd2 cohort. Front Sports Act Living 2022;4:1089579. doi:10.3389/fspor.2022.1089579
    31. ↵
      1. Hult A,
      2. Johansson J,
      3. Nordström P, et al
      . Objectively measured physical activity in older adults with and without diabetes. Clin Diabetes 2019;37:142–9. doi:10.2337/cd18-0041
      OpenUrlAbstract/FREE Full Text
    32. ↵
      1. Statistics Denmark
      . Income in the whole country, Lolland, and Guldborgsund Statistics Denmark. 2021. Available: https://www.statistikbanken.dk/statbank5a/SelectVarVal/saveselections.asp
    33. ↵
      1. Jenum AK,
      2. Lorentzen CAN,
      3. Ommundsen Y
      . “Targeting physical activity in a low socioeconomic status population: observations from the Norwegian 'Romsas in motion' study”. British Journal of Sports Medicine 2009;43:64–9. doi:10.1136/bjsm.2008.053637
      OpenUrlAbstract/FREE Full Text
    34. ↵
      1. Trost SG,
      2. Owen N,
      3. Bauman AE, et al
      . Correlates of adults' participation in physical activity: review and update. Medicine & Science in Sports & Exercise 2002;34:1996–2001. doi:10.1097/00005768-200212000-00020
      OpenUrl
    35. ↵
      1. Swindell N,
      2. Mackintosh K,
      3. McNarry M, et al
      . Objectively measured physical activity and sedentary time are associated with Cardiometabolic risk factors in adults with Prediabetes: the PREVIEW study. Diabetes Care 2018;41:562–9. doi:10.2337/dc17-1057
      OpenUrlAbstract/FREE Full Text
    36. ↵
      1. Steeves JA,
      2. Murphy RA,
      3. Crainiceanu CM, et al
      . Daily patterns of physical activity by type 2 diabetes definition: comparing diabetes, Prediabetes, and participants with normal glucose levels in NHANES 2003-2006. Prev Med Rep 2015;2:152–7. doi:10.1016/j.pmedr.2015.02.007
      OpenUrl
    37. ↵
      1. Migueles JH,
      2. Aadland E,
      3. Andersen LB, et al
      . GRANADA consensus on Analytical approaches to assess associations with accelerometer-determined physical Behaviours (physical activity, sedentary behaviour and sleep) in Epidemiological studies. Br J Sports Med 2022;56:376–84. doi:10.1136/bjsports-2020-103604
      OpenUrlAbstract/FREE Full Text
    38. ↵
      1. Migueles JH,
      2. Cadenas-Sanchez C,
      3. Ekelund U, et al
      . Accelerometer data collection and processing criteria to assess physical activity and other outcomes: A systematic review and practical considerations. Sports Med 2017;47:1821–45. doi:10.1007/s40279-017-0716-0
      OpenUrlPubMed
    39. ↵
      1. Thorsen IK,
      2. Kayser L,
      3. Teglgaard Lyk-Jensen H, et al
      . I tried forcing myself to do it, but then it becomes a boring chore": understanding (Dis)Engagement in physical activity among individuals with type 2 diabetes using a practice theory approach. Qual Health Res 2022;32:520–30. doi:10.1177/10497323211064598
      OpenUrl
    40. ↵
      1. Matthews L,
      2. Kirk A,
      3. Macmillan F, et al
      . Can physical activity interventions for adults with type 2 diabetes be translated into practice settings? A systematic review using the RE-AIM framework. Transl Behav Med 2014;4:60–78. doi:10.1007/s13142-013-0235-y
      OpenUrlCrossRefPubMed
    41. ↵
      1. Sluik D,
      2. Buijsse B,
      3. Muckelbauer R, et al
      . Physical activity and mortality in individuals with diabetes mellitus: A prospective study and meta-analysis. Arch Intern Med 2012;172:1285–95. doi:10.1001/archinternmed.2012.3130
      OpenUrlCrossRefPubMedWeb of Science
    42. ↵
      1. Ried-Larsen M,
      2. Rasmussen MG,
      3. Blond K, et al
      . Association of Cycling with all-cause and cardiovascular disease mortality among persons with diabetes: the European prospective investigation into cancer and nutrition (EPIC) study. JAMA Intern Med 2021;181:1196–205. doi:10.1001/jamainternmed.2021.3836
      OpenUrl
    43. ↵
      1. Whelan ME,
      2. Orme MW,
      3. Kingsnorth AP, et al
      . Examining the use of glucose and physical activity self-monitoring Technologies in individuals at moderate to high risk of developing type 2 diabetes: randomized trial. JMIR Mhealth Uhealth 2019;7:e14195. doi:10.2196/14195
    44. ↵
      1. Shaffer DW,
      2. Kigin CM,
      3. Kaput JJ, et al
      . What is Digital medicine Stud Health Technol Inform 2002;80:195–204.
      OpenUrlPubMed
    45. ↵
      1. Mortensen SR,
      2. Pedersen ME,
      3. Skou ST, et al
      . Online physical exercise and group sessions to increase and maintain physical activity in individuals with type 2 diabetes: a single-arm feasibility study [International journal of environmental research and public health. 2023;20:2893]. Int J Environ Res Public Health 2023;20:2893. doi:10.3390/ijerph20042893
    46. ↵
      1. Jeong IC,
      2. Bychkov D,
      3. Searson PC
      . Wearable devices for precision medicine and health state monitoring. IEEE Trans Biomed Eng 2019;66:1242–58. doi:10.1109/TBME.2018.2871638
      OpenUrl
    47. ↵
      1. Deci EL,
      2. Ryan RM
      . “The "what" and "Why" of goal pursuits: human needs and the self-determination of behavior”. Psychological Inquiry 2000;11:227–68. doi:10.1207/S15327965PLI1104_01
      OpenUrlCrossRefWeb of Science
    48. ↵
      1. Ng JYY,
      2. Ntoumanis N,
      3. Thøgersen-Ntoumani C, et al
      . Self-determination theory applied to health contexts: A meta-analysis. Perspect Psychol Sci 2012;7:325–40. doi:10.1177/1745691612447309
      OpenUrlCrossRefPubMed
    49. ↵
      1. American Diabetes Association
      . Postprandial blood glucose. Diabetes Care 2001;24:775–8. doi:10.2337/diacare.24.4.775
      OpenUrlFREE Full Text
    50. ↵
      1. Park H,
      2. Park C,
      3. Quinn L, et al
      . Glucose control and fatigue in type 2 diabetes: the mediating roles of diabetes symptoms and distress. J Adv Nurs 2015;71:1650–60. doi:10.1111/jan.12632
      OpenUrl
    51. ↵
      1. Ruiz-Tejada A,
      2. Neisewander J,
      3. Katsanos CS
      . Regulation of voluntary physical activity behavior: A review of evidence involving dopaminergic pathways in the brain. Brain Sci 2022;12:333. doi:10.3390/brainsci12030333
    52. ↵
      1. Healy GN,
      2. Wijndaele K,
      3. Dunstan DW, et al
      . Objectively measured sedentary time, physical activity, and metabolic risk: the Australian diabetes, obesity and lifestyle study (Ausdiab). Diabetes Care 2008;31:369–71. doi:10.2337/dc07-1795
      OpenUrlAbstract/FREE Full Text
    53. ↵
      1. Kahn SE,
      2. Hull RL,
      3. Utzschneider KM
      . Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature 2006;444:840–6. doi:10.1038/nature05482
      OpenUrlCrossRefPubMedWeb of Science
    54. ↵
      1. Jing X,
      2. Chen J,
      3. Dong Y, et al
      . Related factors of quality of life of type 2 diabetes patients: a systematic review and meta-analysis. Health Qual Life Outcomes 2018;16:189. doi:10.1186/s12955-018-1021-9
    55. ↵
      1. Liaghat B,
      2. Folkestad L,
      3. Skou ST, et al
      . Prevalence and consequences of musculoskeletal pain in the upper and lower extremities: A cross-sectional analysis of patients with type 1 and type 2 diabetes in Denmark. Prim Care Diabetes 2023;17:267–72. doi:10.1016/j.pcd.2023.02.003
      OpenUrl
    56. ↵
      1. Agardh E,
      2. Allebeck P,
      3. Hallqvist J, et al
      . Type 2 diabetes incidence and socio-economic position: a systematic review and meta-analysis. Int J Epidemiol 2011;40:804–18. doi:10.1093/ije/dyr029
      OpenUrlCrossRefPubMedWeb of Science
    57. ↵
      1. Jepsen R,
      2. Wingstrand A,
      3. Abild SL, et al
      . Socio-economic determinants of participation in the Lolland-Falster health study. J Public Health (Berl) 2020;28:657–64. doi:10.1007/s10389-019-01095-z
      OpenUrl

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Contributors SRM, AG, STS, JCB, and MR-L were involved in the study concept and design. RJ, TLP, and NEB-R participated in the acquisition of data. SRM, AG and JCB conducted the statistical analyses. All authors participated in interpretation of data. SRM and AG drafted the manuscript. All authors participated in critical revision of the manuscript and approved the final version of the manuscript. The corresponding author (AG) is the guarantor of this work, has access to the data, and takes full responsibility for the integrity of the data and the accuracy of the data analyses, and controlled the decision to publish.

    • Funding SRM is currently funded by a research grant from Næstved, Slagelse, Ringsted Hospitals Research Fund (project no. A986), a research grant from The Health Science Research Fund of Region Zealand (project no. A1136), and a research grant from Steno Diabetes Center Zealand. AG is funded by the European Research Council (SCREENS, grant agreement No 716657), the Novo Nordisk Foundation (grant number NNF20SH0062965), and TrygFonden (grant number 130081 and 115606). STS is currently funded by a program grant from Region Zealand (Exercise First) and two grants from the European Union’s Horizon 2020 research and innovation program, one from the European Research Council (MOBILIZE, grant agreement No 801790) and the other under grant agreement No 945377 (ESCAPE). MRL is funded by Trygfonden. LHT is currently funded by a grant from the Danish Regions and The Danish Confederation through the Development and Research Fund for financial support (project no. 2703), a grant from Region Zealand (Exercise First) and Næstved, Slagelse, Ringsted Hospitals research fond, Denmark (project no. A1277). The funders had no role in the study design, data collection, decision to publish, or preparation of the manuscript.

    • Competing interests None declared.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.