Article Text

Download PDFPDF

Walking on sunshine: effect of weather conditions on physical activity in older people
  1. Jochen Klenk1,2,
  2. Gisela Büchele1,
  3. Kilian Rapp1,2,
  4. Sebastian Franke1,
  5. Richard Peter1,
  6. the ActiFE Study Group
  1. 1Institute of Epidemiology and Medical Biometry, Ulm University, Ulm, Germany
  2. 2Department of Clinical Gerontology, Robert-Bosch-Hospital, Stuttgart, Germany
  1. Correspondence to Dr Jochen Klenk, Institute of Epidemiology, Ulm University, Helmholtzstrasse 22, Ulm 89081, Germany; jochen.klenk{at}


Background It is unclear which weather parameters effect the motion-sensor-based measurement of physical activity in terms of walking duration in older people.

Methods Between March 2009 and April 2010, the physical activity of 1324 German community-dwelling older people (≥65 years, 56.4% men) was recorded over 5 days using accelerometers. Multilevel linear regression analysis was used to estimate the effect of local daily weather parameters (daylight, maximum temperature, total global radiation, average precipitation, average wind speed and average humidity) on walking duration.

Results Mean daily walking duration was comparable for men and women, with 104.4±50.7 min and 102.9±47.8 min, respectively. A linear relationship with walking duration was seen for all considered weather parameters. The strongest effect was found for global radiation, which involved an increase in walking duration of 16.1 min in men and 19.2 min in women between an average winter day (with about 0.8 kWh/m2 radiation) and an average summer day (with about 6 kWh/m2 radiation); similar patterns were found for daily maximum temperature and daylight. Furthermore, physical activity decreased significantly with increasing wind speed, precipitation and humidity.

Conclusions Studies on physical activity in community dwelling older people should consider weather conditions.

  • Elderly
  • physical activity
  • weather

Statistics from


There has been an increase in the number of studies using objective measures from sensor devices to quantify physical activity. Studies published in the literature suggest that weather conditions can considerably influence study results.1 2 Yet, mostly the effect of season on physical activity has been analysed.2 Studies considering the effect of specific meteorological factors on physical activity are rare, and have reported only activity counts, which are not very intuitive to interpret compared with walking duration.3–6 Therefore, the aim of this study was to analyse the effect of various weather parameters on objectively assessed daily walking duration in older people.


The Activity and Function in the Elderly in Ulm (ActiFE Ulm) study is a population-based cohort study in older people aged ≥65 years in Southern Germany.7 It was approved by the ethical committee of Ulm University. Between March 2009 and April 2010, 1505 non-institutionalised individuals underwent a baseline assessment. Physical activity was recorded over 7 days using a thigh-worn uni-axial accelerometer (activPAL; PAL Technologies, Glasgow, Scotland), and classified into categories: (1) laying or sitting, (2) standing and (3) walking. Data from 1324 participants were used for analyses. Daily measures of maximum temperature (°C), total global radiation (kWh/m2), average precipitation (mm/h), average humidity (%), average wind speed (m/s) and daylight (h) (period between sunrise and sunset) served as exposure variables.

Average daily walking duration in minutes (min) and 95% confidence interval (CI) were estimated for all categorised weather parameters using multilevel analyses. To check whether there was confounding between exposure variables, mutually adjusted estimates were also calculated. Since maximum temperature, global radiation and day length were strongly correlated, they were not included within the same models to avoid collinearity.


The study population consisted of 747 men and 577 women (mean age 74.6 years) with at least one completed day (24 h) of physical activity assessment. Overall 7525 measurements of average daily walking duration were available for analysis. Mean daily walking duration was comparable for men and women, with 104.4 min (SD 50.7 min) and 102.9 min (SD 47.8 min), respectively.

Figure 1 shows a clear linear relationship with daily walking duration for all the weather parameters analysed. In corresponding linear regression analyses, the strongest effect was caused by global radiation, with an increase of 16.1 min in men and 19.2 min in women between an average winter day (with about 0.8 kWh/m2 radiation) and an average summer day (with about 6 kWh/m2 radiation). An increase of 10°C in maximum daily temperature extended the walking duration by more than 7 min in both genders. Between days with a short daylight period (9 h) and a long daylight period (16 h) the walking duration increased by 12.6 min in men and 13.3 min in women. Similar, yet inverted, effects were seen in wind speed, daily precipitation and humidity. After adjustment for other weather parameters, daylight was no longer significant. The significance of all other weather parameters was reduced, but they remained significant.

Figure 1

Relationships between average daily walking duration (mean±95% CI) and daily maximum temperature (A), daylight (B), total global radiation (C), average daily precipitation (D), average daily humidity (E), and average daily wind speed (F).


The presented results show strong associations between weather conditions and physical activity in older people. We believe that this is the first study to quantify the outcome measure in terms of objectively assessed walking duration. The strongest relationships were found for global radiation and maximum temperature. Daylight was only slightly associated with activity, and after adjustment for other weather parameters it did not remain significant.

The results are important for further studies. If study design and data analysis do not account for the influence of weather, findings could be considerably misinterpreted due to bias (eg, exposed measured on cold days, and unexposed measured on warm days) or noise (reduced power).

Previous studies have used season to account for weather conditions. However, in the present study, season was not associated with walking duration (data not shown), and daylight, which is causally dependent on season, showed only a weak association but a high variance compared with the other weather parameters analysed. This is in line with findings in older Japanese people,3 but in contrast with the results of a Scottish study in older functionally impaired people.4

All other considered weather parameters were more explicitly associated with walking duration in the present study. The strong association of physical activity with global radiation as a surrogate for duration of bright sunshine confirms findings reported by Sumukadas et al, but contradicts those reported by Togo et al.3 4 For temperature, comparable results were found in three previous studies.4–6

Precipitation led to a reduction in physical activity, as consistently reported in the literature.3–6 In addition to other reasons, fear of falling may be additionally responsible for these findings, especially in older people, since rain and snow cause slippery conditions outside. Walking duration also decreased with humidity in the present study. As humidity to some degree has a negative effect on wellbeing and health, a similar inverse relation with walking duration may exist.8 Interestingly, mean wind speed was also related to a decrease in physical activity, and seemed to be the only independent weather parameter within the variables that were considered in this study. The mechanism is unclear, and only the Canadian study5 has found a similar relationship.

The major strength of this study is the objectively measured data of weather parameters and physical activity on a day-by-day basis in a large population-based sample. The main limitations are: the participants were measured on seven consecutive days only. Assessment of physical activity at several points throughout the year could have improved the results. Extreme weather events were not observed during the study period of more than 1 year. With higher values it is possible that the associations would not remain linear. Although accelerometry is probably one of the best methods to quantify physical activity in observational studies,9 sensitivity of step detection decreased with decreasing walking speed.10 This may have biased the results, as hot summer days and cold winter days could lead to a reduction in gait speed. Finally, knowledge of being under observation can increase physical activity11; however, this should only affect the magnitude of the effects, since it is unlikely that it depends on weather conditions.

In conclusion, the present results strongly suggest that studies on physical activity in community-dwelling older people should consider weather conditions. Daylight and season do not appear to be the best parameters. Maximum temperature and global radiation were the most important in terms of effect size.

What is already known on this subject

  • There is evidence that physical activity in older people is affected by season. Few studies have analysed the effect of specific weather parameters on objectively measured physical activity. Those publications have reported step or activity counts, but not walking duration. It is unclear which weather parameters are most important, how they are linked to each other, and their magnitude in terms of walking duration.

What this study adds

  • We found considerable effects of maximum temperature, global radiation, precipitation, humidity and wind speed on walking duration in community-dwelling older people. Therefore, studies on physical activity should account for weather conditions in study design and data analysis.


Other members of the Activity and Function in the Elderly in Ulm (ActiFE) study group are: Thorsten Nikolaus MD, Michael Denkinger MD (Bethesda Geriatric Clinic, Ulm, Germany), Gabriele Nagel MD, Gudrun Weinmayr PhD, Florian Herbolsheimer MSc (Institute of Epidemiology, Ulm University, Germany), Albert C Ludolph MD, Christine A F von Arnim MD (Department of Neurology, Ulm University Medical Centre, Germany), Karin Scharffetter-Kochanek MD, Hartmut Geiger MD (Department of Dermatology and Allergology, Ulm University Medical Centre, Germany), Jürgen M Steinacker MD (Department of Sports and Rehabilitation Medicine, Ulm University Medical Centre, Germany), Bernhard O Böhm MD (Department of Internal Medicine I, Ulm University Medical Centre, Germany), Julia Kirchheiner MD (Institute of Clinical Pharmacology, Ulm University Medical Centre, Germany), Wolfgang Koenig MD, Christian Schumann MD (Department of Internal Medicine II, Ulm University Medical Centre, Germany), Matthias Riepe MD (Department of Psychiatry and Psychotherapy II, Ulm University, Germany), Lenhard Rudolph MD (Max-Planck Group for Stem Cell Research, Ulm University, Germany).



  • Funding The study was funded by a grant from the Ministry of Science, Research and Arts, Baden-Wuerttemberg, Germany, as part of the Geriatric Competence Centre, Ulm University.

  • Competing interests None declared.

  • Ethics approval This study was conducted with the approval of the ethical committee of Ulm University.

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

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.