HIV infection can be successfully treated using antiretroviral therapy (ART), which consists of a set of drugs that restore the immune system and prevent HIV replication. Since the introduction of ART, there has been a dramatic reduction in HIV-related mortality and morbidity. However, because the medications helped to transition HIV to a chronic disease, people living with HIV (PLWH) have experienced elevated rates of cardiovascular disease, dyslipidemia, diabetes, osteoporosis, and other noncommunicable diseases when compared with people without HIV (Eyawo et al., 2017). The elevated rates of comorbidities require attention to lifestyle factors that can affect health and quality of life for PLWH, such as physical activity (PA) and stationary behavior (SB; Tremblay et al., 2017; Vancampfort et al., 2018). Studies have demonstrated that physically active PLWH have lower rates of chronic diseases, decreased prevalence of fatigue and depression, better mental health, decreased risk of cardiovascular disease, and better quality of life (Monroe et al., 2017; Quiles, Garber, & Ciccolo, 2018). SB refers to any waking behavior done while lying, reclining, sitting, or standing, with no ambulation, irrespective of energy expenditure (it differs from sedentary behavior, which also includes energy expenditure measurement; Tremblay et al., 2017). Moreover, time spent in SB is independently associated with deleterious health outcomes (i.e., all-cause mortality, cardiovascular disease incidence and mortality, cancer incidence and mortality, and Type 2 diabetes incidence) regardless of PA in the general population (Biswas et al., 2015).
In general, PLWH spend most of their time engaged in SB (Vancampfort, Mugisha, De Hert, Probst, & Stubbs, 2017) and a considerable proportion of PLWH do not meet the recommendations for PA (that adults engage in at least 150 min of moderate-to-vigorous PA [MVPA] during the week; Piercy et al., 2018; Vancampfort et al., 2018). Considering that PLWH experience high symptom burden, and the importance of PA and SB on health-related outcomes, reliable and accurate data regarding PA and SB in this population are necessary (Eyawo et al., 2017; Piercy et al., 2018). The accurate measurement of PA and SB is also important to identify correlates and determinants and to inform policies for health promotion. Moreover, it serves to evaluate the efficacy of interventions to increase PA and to lower SB levels.
However, PA and SB are difficult to measure accurately, and in different measurement methods, increasing attention is being given to accelerometry. Accelerometry is one objective method to measure PA, and different parameters can be estimated using accelerometers, such as time spent in different intensities of PA, number and length of activity bouts, number of steps, amount of time spent in SB, breaks in SB, and others.
Given the characteristics of PLWH, who are known to have a high prevalence of comorbidities, fatigue, pain, and unemployment (Eyawo et al., 2017; Perazzo, Webel, Voss, & Prince-Paul, 2017), different parameters may need to be considered when collecting accelerometer data about PLWH. Also, due to PLWH spending most of their time in SB and having low levels of PA (Vancampfort et al., 2017, 2018), the parameters used for PLWH may be different than those used for the general population. Some of these issues have received special attention in other populations, such as cancer patients (Broderick, Ryan, O'Donnell, & Hussey, 2014). However, to the best of our knowledge, there have been no studies that explore these issues in PLWH. Thus, our study had two purposes: (a) to analyze differences regarding PA (i.e., light PA, MVPA, and steps count) and SB patterns between weekdays and weekend days in a cohort of PLWH, and (b) to determine whether demographic, physical, and behavioral variables moderated these differences.
A randomized clinical trial was conducted to test the effect of the SystemCHANGE intervention on PA and dietary intake in PLWH (NCT02553291). The institutional review board at University Hospitals, Cleveland Medical Center, approved the study. This was a secondary analysis of the SystemCHANGE parent study, analyzing PA and SB data obtained from accelerometers.
This study included PLWH (n = 99) who were recruited through an HIV research registry and flyers posted in HIV care organizations in northeast Ohio, United States. Participants were required to be on ART and have a suppressed HIV-1 viremia (<400 copies/mL) for at least 1 year prior to enrollment. With regard to PA, potential participants were excluded if they had a medical contraindication for exercise or already met weekly PA recommendations of 150 m of MVPA (Piercy et al., 2018). For a full description of inclusion and exclusion criteria for the parent study, refer to Webel et al. (2018).
Demographic and Clinical Data
Age, race, income, and employment status were collected via self-report on a private computer. Additionally, participants consented to medical chart abstraction of relevant data including the number of years living with HIV and using ART, current CD4+ T-cell count and HIV viral load, CD4+ T-cell nadir, and presence of comorbidities.
Handgrip strength is commonly used as a measure of overall muscle strength, and it was measured with a hand-held dynamometer (Jamar Technologies, Hatfield, PA). A research assistant followed a script and asked the subject about hand surgery or pain that could have influenced the results. The participants sat in a chair and rested an arm on a table with the elbow on top of a mouse pad and the arm forming a right angle. The technician adjusted the grip size of the dynamometer to the subject's hand size and explained how the device worked. Participants squeezed the dynamometer twice as practice trials. After that, two trials were attempted with each hand, and all results were recorded on a form to the nearest 2 pounds. The values presented were the sum of the best results of each hand.
Depression and Friendship Assessment
Two validated and widely used questionnaires measuring depression (Beck Depression Inventory; Beck, 1961) and friendship (The Friendship Scale; Hawthorne & Griffith, 2000) were used, then scored and interpreted based on the recommended scoring procedures.
The Beck Depression Inventory is composed of 21 items that assess depressive symptoms, such as depressed mood, hopelessness, suicidal ideation, sleep disturbance, and appetite change. Based on scoring, participants could be classified as minimal or no depression, mild depression, moderate depression, or severe depression. The inventory has showed high construct validity with the symptoms it measures, with a coefficient alpha rating of 0.92 for outpatients. It also had a 1-week test–retest reliability of r = 0.93 and an internal consistency of α = 0.91 (Beck, 1961).
The Friendship Scale is composed of 5 items and assesses social isolation or social support in the past 4 weeks. Based on scoring, participants could be classified as isolated/low social support, some isolation or some support, or socially connected. Validity tests provided evidence that the scale was sufficiently robust to be used with confidence (internal consistency of the scale was α = 0.76; Hawthorne & Griffith, 2000).
Physical Activity and Stationary Behavior
All participants wore an ActiGraph GT3X/+ accelerometer (ActiGraph, LLC, Fort Walton Beach, FL). Participants were instructed to wear the accelerometer during all waking hours for 7 consecutive days, except for while they were in water (e.g., showering or swimming). The Actigraph was placed on the participants' nondominant hip. Study staff instructed the participants on proper placement and counseled participants on the importance of wearing the Actigraph every day. We processed accelerometer data according to recommendations for adults using a sampling of 30 Hz, 60-s epochs, and the low-frequency filter was not applied (Migueles et al., 2017). We used ActiLife software to calculate the amount of time spent in light PA and MVPA per valid day, including total time spent in each intensity, as well as the time spent in bouts of 10 or more minutes using the Sasaki, John, and Freedson (2011) adult calculation for tri-axial accelerometers. SB was also calculated as the total time and as stationary bouts, which were calculated and defined as 10 or more minutes with 100 or fewer counts per minute. Wear time was determined by subtracting nonwear time from 24 hr. Nonwear time was defined as at least 60 consecutive minutes of zero counts.
When analyzing PA and SB time, we included both total and bouted times. Total time was the sum of minutes spent in either PA and SB, whereas bouted time was the sum of minutes spent in an aggregated period of at least 10 min. For example, considering MVPA time: if a subject spent 14 min in MVPA, then 5 min in SB, and after that 8 more minutes in MVPA, this would count 22 min for MVPA total time, but only 14 min for MVPA bouted time. Previous PA recommendations required that time spent in MVPA should be spent in bouts of at least 10 min to gain more health benefits from PA (Haskell et al., 2007; World Health Organization, 2011).
Data distribution was checked through the Kolmogorov–Smirnov test, which is a one-sample nonparametric goodness-of-fit test for determining whether a variable followed a given distribution in a population (Dodge, 2008). To test the difference between weekday (Monday–Friday) and weekend day (Saturday and Sunday) activities, we used paired t-test for normally distributed variables and Wilcoxon test for non-normally distributed variables. To examine potential moderation of different demographic, physical, and behavioral variables on the difference between weekdays and weekend days, general linear models for repeated measures were used, with the values for weekdays and weekend days as within-participants factor, and the selected variables as between-participants factor. Statistical significance was set at p ≤ .05 and data were analyzed using SPSS for Windows Version 24.0 (SPSS, Inc., Chicago, IL).
Participants (n = 99) were age 53.2 years on average and the majority were men (67.7%), African American or Black (84.8%), and unemployed (90.9%). The participants had been living with HIV an average of 17 years, and the most prevalent comorbidities were hypertension (48.5%) and depression (46.5%). Most of the participants were classified as experiencing minimal depression (71.7%) and as socially connected (50.5%). Differences in time living with HIV, handgrip strength, and body mass index (BMI) were found between men and women (p < .05). Male participants had lived with HIV longer (18.4 vs. 14.3 years), were stronger (141.9 vs. 103.6 lbs), and had a lower BMI (26.6 vs. 34.7 kg/m2) compared with women. Other demographic, clinical, physical, and behavioral characteristics are shown in Table 1.
When comparing PA and SB patterns between weekdays and weekend days (Table 2), we found different patterns between total and bouted times. When analyzing the total time spent in each intensity, PLWH spent 7.8 fewer minutes on MVPA during weekend days. In addition, when analyzing the bouted time, PLWH accumulated 24.2 more minutes in bouted SB and had more bouted activity time on weekend days (bouts of ≥ 10 min), usually spent on SB or light activity. Furthermore, the participants accumulated an average of 1,011.9 fewer steps on weekend days.
Additionally, we analyzed the moderation effects of selected variables (i.e., age, sex, BMI, employment status, depression rates, social isolation, and handgrip strength) on the difference between PA and SB on weekdays versus weekend days. Among these variables, age and BMI influenced bouted MVPA values (p < .05, Figure 1). Regarding age, PLWH who were 50 years old or younger spent less time in bouted MVPA on weekend days (week = 14.3 vs. weekend = 9 min). Regarding BMI, people classified as thin or eutrophy (BMI < 25.0 kg/m2) spent more time in bouted MVPA on weekend days (week = 7.7 vs. weekend = 12.1 min), and people classified as overweight or obese (BMI ≥ 25.0 kg/m2) spent less time in bouted MVPA on weekend days (week = 8.5 vs. weekend = 4.3 min). No difference was found for total MVPA time (p > .05). Also, none of the variables influenced light activity and SB values (p > .05).
The main findings of our study were that PLWH were less active on weekend days, spending 7.8 less minutes in MVPA and 1,011.9 fewer steps on weekend days, and also accumulating 24.2 more minutes of bouted SB. Additionally, age and BMI exerted influence on bouted MVPA values. These results demonstrated the need to consider weekend days during acquisition of PA and SB data of PLWH using accelerometry to get more accurate data, a recommendation not usually followed in other studies or populations (Migueles et al., 2017).
Recently, accelerometry has been the primary method to evaluate PA and SB patterns in different populations for several reasons (Migueles et al., 2017). First, accelerometry is not as expensive nor as difficult to replicate in real-life settings as gold-standard or reference methods (i.e., direct observation, indirect calorimetry, and doubly labeled water). Second, it provides more accurate information than self-report measures (e.g., questionnaires). Moreover, hip-worn accelerometers present good acceptability and wear compliance by participants (O'Brien et al., 2017). However, although the literature has advanced in standardizing the use and application of accelerometers, some aspects of data acquisition still need to be improved to obtain accurate and reliable data. In this regard, the wear time is one of the main decisions that the researcher must make to obtain valid and representative data of a subject's PA and SB patterns. The parameter of 4 valid days with 10 hr per day of wear time is widely accepted for a variety of populations (Migueles et al., 2017), but these recommendations usually do not discriminate between weekdays and weekend days.
There is a growing body of literature involving youth that recommends the inclusion of weekend days when evaluating the PA and SB patterns because there are generally significant differences in activity level due to schooling (Nader, Bradley, Houts, McRitchie, & O'Brien, 2008). In adults, we could expect the same due to employment. For example, Gretebeck and Montoye (1992) analyzed a working population of men and found no day-to-day difference regarding PA patterns when only weekdays were compared, but the day-to-day difference was significant when weekend days were also analyzed. However, even when employment status was not considered, adults tended to show differences in PA and SB patterns between weekdays and weekend days (Evenson, Wen, Metzger, & Herring, 2015). In our study, most of the participants were unemployed (90.9%), but we still found differences regarding weekday and weekend day activity. This can likely be attributed to daily duties that are usually done on weekdays (i.e., take children to school, clean the house, go to the physician, get medication) that require participants to be more active on these days. This was true when we looked at bouted MVPA by age group (Figure 1), where participants who were 50 years of age or younger (e.g., participants that usually perform these duties) showed lower MVPA levels on weekends. Furthermore, the participants in our study presented with characteristics of stationary leisure time (as seen by the overall low MVPA levels) and tended to engage in activities with these characteristics during free time.
Regarding PA and SB patterns, studies involving only PLWH found that this population spent most of their time engaged in stationary activities (532.8 min/day, or approximately 9 hr/day; Vancampfort et al., 2017). A considerable proportion of PLWH do not meet the recommendations for PA (at least 150 min of MVPA during the week; Vancampfort et al., 2018). Moreover, the time that PLWH spend in MVPA is lower than other people living with chronic diseases (Vancampfort et al., 2018). In our sample, we found similar results, where participants spent more time in SB and presented very low MVPA levels. Also, similar to other studies involving adults, we found differences in MVPA levels between weekdays and weekend days (Evenson et al., 2015; Gretebeck & Montoye, 1992).
Another difference in MVPA values was found based on BMI of subjects. Although participants classified as “thinness and eutrophy” had the same bouted MVPA levels during weekdays when compared with participants classified as “overweight and obesity,” this pattern was different during weekend days. On weekends, participants with excess weight tended to have lower bouted MVPA levels, possibly spending most of their weekends in screen time. Other studies have demonstrated a relationship between high screen time and participants classified as overweight or obese (Duncan, Vandelanotte, Caperchione, Hanley, & Mummery, 2012). Moreover, studies have shown that TV viewing was the leisure activity that occupied most of adult time; although a slight decrease in total sedentary time had been observed over the years, TV viewing had not declined (Scholes & Mindell, 2013).
One methodologic decision regarding our data was to also include the time spent in bouted activity (bouts of ≥ 10 min). Not only the total time spent in MVPA and SB but also the number and durations of bouts and breaks of these behaviors have received attention (Tremblay et al., 2017). Studies have found that longer sedentary bout duration was associated with a higher risk for all-cause mortality (Diaz et al., 2017), and that interrupting bouts of sedentary behavior with PA might confer positive health benefits, independent of total sedentary time (Chastin, Egerton, Leask, & Stamatakis, 2015). Of interest, we observed that although total MVPA was lower during the weekdays, bouted MVPA was similar between weekdays and weekend days. In contrast, total SB was similar between weekdays and weekend days, but bouted SB was higher during the weekdays. These findings show that both the total time and the pattern of MVPA and SB can represent specific outcomes. PLWH probably spent more time in MVPA during weekend days, but this time is not bouted. However, the higher SB observed at the weekend days was especially bouted, which represented a more harmful manifestation of this behavior. Thus, interventions should consider not only reducing total SB but also reducing bout duration with a higher number of SB breaks for PLWH, especially during weekend days.
Among the limitations of our study, we included participants from a single location (northeast Ohio, USA), and our cohort was characterized by a high percentage of unemployed participants with low MVPA levels. However, ours was the first study to investigate behavioral differences between weekdays and weekend days in PLWH. Our results highlight the importance of considering weekdays and weekend days when measuring PA and SB in PLWH to get accurate and reliable data, as well as the need to develop effective strategies to promote overall active behaviors, especially during weekends. Observed differences between weekdays and weekend days activity may seem small, but evidence from studies with different populations found that even small differences in week/weekend activity, or small differences over time, can have a significant impact in health outcomes (Chastin, Mandrichenko, Helbostadt, & Skelton, 2014; Drenowatz et al., 2016); also, even small differences in interday activity could overestimate weekly activity levels (Kocherginsky, Huisingh-Scheetz, Dale, Lauderdale, & Waite, 2017). Finally, we highlighted the need to replicate our method in PLWH from other locations to ensure that our observed behaviors are similar to patients from other countries/locations.
Physical activity and SB patterns of PLWH differ between weekdays and weekend days. When using accelerometry to measure activity patterns of PLWH, we suggest that validation parameters consider both weekdays and weekend days. Also, because of being less active on weekend days, interventions to promote PA could focus on these days.
The authors report no real or perceived vested interests related to this article that could be construed as a conflict of interest.
This project was funded by grants from the American Heart Association (PI: A. R. Webel, Grant #14CRP20380259) and a developmental grant from the University Hospitals/Case Western Reserve University Center for AIDS Research (PI: Jonathon Karn, National Institutes of Health Grant # P30 AI036219). It was also supported by National Institutes of Health Grants (PI: Pamela Davis, Grant # UL1 RR024989). Also, V.H.F.d. Oliveira (first author) was granted a Brazilian scholarship for a research internship at A. R. Webel’s laboratory in Cleveland, USA (bolsista da Capes/Programa de Doutorado Sanduíche no Exterior/Processo no [88881.132132/2016-01]).
Beck A. T. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4(6), 561. doi:10.1001/archpsyc.1961.01710120031004
Biswas A., Oh P. I., Faulkner G. E., Bajaj R. R., Silver M. A., Mitchell M. S., Alter D. A. (2015). Sedentary time and its association with risk for disease incidence, mortality, and hospitalization in adults a systematic review and meta-analysis. Annals of Internal Medicine, 162(2), 123–132. doi:10.7326/M14-1651
Broderick J. M., Ryan J., O'Donnell D. M., Hussey J. (2014). A guide to assessing physical activity using accelerometry in cancer patients. Supportive Care in Cancer, 22(4), 1121–1130. doi:10.1007/s00520-013-2102-2
Chastin S. F., Egerton T., Leask C., Stamatakis E. (2015). Meta-analysis of the relationship between breaks in sedentary behavior and cardiometabolic health. Obesity
, 23(9), 1800–1810. doi:10.1002/oby.21180
Chastin S. F. M., Mandrichenko O., Helbostadt J. L., Skelton D. A. (2014). Associations between objectively-measured sedentary behaviour and physical activity with bone mineral density in adults and older adults, the NHANES study. Bone, 64, 254–262. doi:10.1016/j.bone.2014.04.009
Diaz K. M., Howard V. J., Hutto B., Colabianchi N., Vena J. E., Safford M. M., Hooker S. P. (2017). Patterns of sedentary behavior and mortality in U.S. middle-aged and older adults. Annals of Internal Medicine, 167(7), 465. doi:10.7326/M17-0212
Dodge Y. (2008). Kolmogorov-Smirnov test. In The concise encyclopedia of statistics (pp. 283–287). New York, NY: Springer.
Drenowatz C., Gribben N., Wirth M. D., Hand G. A., Shook R. P., Burgess S., Blair S. N. (2016). The association of physical activity during weekdays and weekend with body composition in young adults. Journal of Obesity
, 2016, 8236439. doi:10.1155/2016/8236439
Duncan M. J., Vandelanotte C., Caperchione C., Hanley C., Mummery W. K. (2012). Temporal trends in and relationships between screen time, physical activity, overweight and obesity
. BMC Public Health, 12(1), 1060. doi:10.1186/1471-2458-12-1060
Evenson K. R., Wen F., Metzger J. S., Herring A. H. (2015). Physical activity and sedentary behavior patterns using accelerometry from a national sample of United States adults. International Journal of Behavioral Nutrition and Physical Activity, 12(1), 20. doi:10.1186/s12966-015-0183-7
Eyawo O., Franco-Villalobos C., Hull M. W., Nohpal A., Samji H., Sereda P., Hogg R. S. (2017). Changes in mortality rates and causes of death in a population-based cohort of persons living with and without HIV from 1996 to 2012. BMC Infectious Diseases, 17(1), 174. doi:10.1186/s12879-017-2254-7
Gretebeck R. J., Montoye H. J. (1992). Variability of some objective measures of physical activity. Medicine and Science in Sports and Exercise, 24(10), 1167–1172. doi:10.1249/00005768-199210000-00016
Haskell W. L., Lee I. M., Pate R. R., Powell K. E., Blair S. N., Franklin B. A., Bauman A. (2007). Physical activity and public health: Updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Medicine and Science in Sports and Exercise, 39(8), 1423–1434. doi:10.1249/mss.0b013e3180616b27
Hawthorne G., Griffith P. (2000). The Friendship Scale: Development and properties. Melbourne, Australia: Centre for Health Program Evaluation.
Kocherginsky M., Huisingh-Scheetz M., Dale W., Lauderdale D. S., Waite L. (2017). Measuring physical activity with hip accelerometry among U.S. older adults: How many days are enough? PLoS One, 12(1), e0170082. doi:10.1371/journal.pone.0170082
Migueles J. H., Cadenas-Sanchez C., Ekelund U., Nyström C. D., Mora-Gonzalez J., Löf M., Ortega F. B. (2017). Accelerometer
data collection and processing criteria to assess physical activity and other outcomes: A systematic review and practical considerations. Sports Medicine, 47(9), 1821–1845. doi:10.1007/s40279-017-0716-0
Monroe A., Zhang L., Jacobson L., Plankey M., Brown T., Miller E., Sacktor N. (2017). The association between physical activity and cognition in men with and without HIV infection. HIV Medicine, 18(8), 555–563. doi:10.1111/hiv.12490
Nader P. R., Bradley R. H., Houts R. M., McRitchie S. L., O'Brien M. (2008). Moderate-to-vigorous physical activity from ages 9 to 15 years. Journal of the American Medical Association, 300(3), 295–305. doi:10.1001/jama.300.3.295
O'Brien W. J., Shultz S. P., Firestone R. T., George L., Breier B. H., Kruger R. (2017). Exploring the challenges in obtaining physical activity data from women using hip-worn accelerometers. European Journal of Sport Science, 17(7), 922–930. doi:10.1080/17461391.2017.1323952
Perazzo J. D., Webel A. R., Voss J. G., Prince-Paul M. (2017). Fatigue symptom management in people living with human immunodeficiency virus. Journal of Hospice & Palliative Nursing, 19(2), 122–127. doi:10.1097/NJH.0000000000000329
Piercy K. L., Troiano R. P., Ballard R. M., Carlson S. A., Fulton J. E., Galuska D. A., Olson R. D. (2018). The physical activity guidelines for Americans. Journal of the American Medical Association, 320(19), 2020. doi:10.1001/jama.2018.14854
Quiles N., Garber C., Ciccolo J. (2018). Resting autonomic function in active and insufficiently active people living with HIV. International Journal of Sports Medicine, 39(01), 73–78. doi:10.1055/s-0043-118033
Sasaki J. E., John D., Freedson P. S. (2011). Validation and comparison of ActiGraph activity monitors. Journal of Science and Medicine in Sport, 14(5), 411–416. doi:10.1016/j.jsams.2011.04.003
Scholes S., Mindell J. (2013). Chapter 2—Physical activity in adults. InHealth Survey for England 2012 (Vol. 1, pp. 1–49). Retrieved from https://files.digital.nhs.uk/publicationimport/pub13xxx/pub13218/hse2012-ch2-phys-act-adults.pdf
Tremblay M. S., Aubert S., Barnes J. D., Saunders T. J., Carson V., Latimer-Cheung A. E., Chinapaw M. J. M. (2017). Sedentary Behavior Research Network (SBRN)—Terminology consensus project process and outcome. International Journal of Behavioral Nutrition and Physical Activity, 14(1), 75. doi:10.1186/s12966-017-0525-8
Vancampfort D., Mugisha J., De Hert M., Probst M., Firth J., Gorczynski P., Stubbs B. (2018). Global physical activity levels among people living with HIV: A systematic review and meta-analysis. Disability and Rehabilitation, 40(4), 388–397. doi:10.1080/09638288.2016.1260645
Vancampfort D., Mugisha J., De Hert M., Probst M., Stubbs B. (2017). Sedentary behavior in people living with HIV: A systematic review and meta-analysis. Journal of Physical Activity and Health, 14(7), 571–577. doi:10.1123/jpah.2016-0507
Webel A. R., Moore S. M., Longenecker C. T., Currie J., Horvat Davey C., Perazzo J., Josephson R. A. (2018). Randomized controlled trial of the SystemCHANGE intervention on behaviors related to cardiovascular risk in HIV+ adults. Journal of Acquired Immune Deficiency Syndromes, 78(1), 23–33. doi:10.1097/QAI.0000000000001635
World Health Organization. (2011). Global recommendations on physical activity for health. Retrieved from https://www.who.int/dietphysicalactivity/factsheet_recommendations/en/