Falls: prevention of fall accidents (ancillary study)

Prevention of fall accidents in elderly people. From risk profile to intervention strategy (ancillary study)

Contact: Natasja van Schoor


Falls often occur in the elderly. About 30% of community-dwelling persons over the age of 65 fall at least once each year and about 15% fall two or more times a year (Tinetti, Speechly & Ginter, 1988; Tromp et al., 1998). The consequences of falls could be serious, for example a fall could result in a hip fracture (1%), other fractures (3%), or a head trauma (3%). However, little is known about other consequences. Injurious falls could place heavy demands on health care systems. To preserve the health and independence of the elderly, it is important to prevent falls.

Since 1995, LASA has investigated predictors of falls and fractures. For the prevention of falls it is important to investigate which persons are at a higher risk of falling as well as to develop an intervention program. Tromp et al. (2000) developed a fall-risk screening test with easy measurable variables. Two easy measurable modifiable variables, namely immobility (measured by performance tests) as well as low physical activity (measured by a questionnaire), seemed to be important risk factors for falls (Tromp et al., 1998). However, for the development of an intervention program more accurate measurements of mobility and activity are needed. That’s why in the present study, which was performed in a subsample of the LASA population, besides the easy measurable variables also more accurate measurements were obtained to measure balance, muscle strength and physical activity. Furthermore, the circumstances and consequences of falls were investigated.


  1. To investigate the circumstances and consequences of falls in the elderly. (Stel et al., 2004)
  2. To assess to what extent balance, muscle strength and physical activity are related to falls in the elderly. (Stel et al., 2003)
  3. To validate questionnaires and tests of physical activity and mobility in the elderly. (Stel et al., 2004)
  4. To investigate which combinations of predictors for falls (risk profiles) frequently lead to falls in the elderly and to investigate how an intervention strategy can be developed.



Respondents were recruited from 1509 elderly men and women, aged 65 years and older at January 01 in 1996, who participated in a follow-up study on falls in LASA (see side study “Falls and fractures” in 1995). Subjects who completed the medical interview of LASA in 1998/1999 and fell at least once from January 1998 till January 1999 (n=328) were eligible for this additional study as well as 196 respondents who did not fall during this year (randomly selected). Of these 524 eligible subjects, 85 did not participate because 14 were deceased before approach, 37 refused, 4 were cognitive not capable, 24 were physically not capable, and 6 could not be contacted.

Study design

Trained interviewers visited the participants two times at home. During the first visit a questionnaire about the circumstances and consequences of falls was assessed, and postural sway, muscle strength (m. quadriceps femoris and handgrip strength), physical activity (questionnaire) and three physical performance tests were measured. In addition, the interviewer instructed the respondents how to use the seven-day diary and pedometer the next seven days. After a week the interviewer visited the participant again to collect the pedometer and the physical activity forms (data of the seven-day diary and pedometer). The interviewer asked the respondent whether the amount of physical activity of the seven days of registration was representative for the usual amount of activity. Furthermore, during one year of follow-up falls and fractures were ascertained with a ‘fall calendar’ (see documentation ‘falls’).


Some variables were also obtained in a main or medical interview of LASA (documentation elsewhere), whereas other variables were new variables for LASA (documentation below) (see table 1). The names of the files of the data sets are shown in table 1.

Table 1. Variables and data sets in study

Obtained in
LASA before
Circumstances and consequences of falls yes LASADo60
Performance tests (walking test, chair and tandem stand) yes LASADo61
Physical activity yes LASADo62
Muscle strength quadriceps femoris (MicroFET) no LASADo64
Grip strength yes LASADo63
Postural sway (AccuSway) no LASADo66
Response on physical activity diary and pedometer no LASADo71
Actual diary and pedometer data for seven days no LASADo70
Status per fall calendar sheet yes LASADo72
Fall follow-up and composed fall variables yes LASADo73

Circumstances and consequences of falls (questionnaire)

The circumstances and consequences of falls were investigated using a questionnaire. The interviewer asked whether the respondent experienced falls or fractures during the last year. The questions focused on the circumstances and consequences of the last fall. Furthermore, questions on the consequences of all falls were included for subjects who fell more than once after the age of 65 years. The questions about consequences of falls included: medical help, investigation of causes of the fall, treatment of consequences of the fall, changes in physical activity, social activities, and ADL as a consequence of the fall, and psychological and emotional changes as a consequence of falling. The questionnaire was completed between 0-18 months after the fall. These results have been described and published (Stel et al., 2004)

(Note: The circumstances and consequences of falls have also been assessed during the medical interview of LASA-D (1998/1999). In this interview the circumstances and consequences of a maximum of two falls in the previous year were recorded.)

Muscle strength of quadriceps femoris (MicroFET)

The isometric muscle strength of the m. quadriceps femoris was measured using the MicroFET hand held dynamometer (Hoggan Health Industries, U.S.A.) according to a method described by Hsieh and Philips (1990). The MicroFET seems to be a reliable method for measuring maximum muscle strength (Hsieh and Philips, 1990). The measurements were done at least four times. If after four times the maximum score was not reached measurements continued (with a maximum of seven measurements). The measurements were performed on the right and left side. The maximum muscle strength (in Nm), which can be the right or left leg, can be used (Beld van den et al., 1999). It is also possible to sum up the maximum score of the right and left leg.

Balance (AccuSway)

Postural sway was measured using the portable AccuSway System (AMTI, U.S.A.). This force platform records the anterior-posterior as well as the medial-lateral centre of pressure displacement. The deviation of the centre of pressure was recorded. The participants were instructed to stand quietly on the force platform without shoes, and to look straight ahead, with their feet comfortably spaced, and their arms at sides. The first trial was performed with eyes open, followed by a trial with eyes closed. Thereafter the trials were repeated three times. Each measurement took 30 seconds. For statistical analyses the mean deviation of the centre of pressure of the three measurements can be used.

Seven-day diary on physical activity

For a period of seven consecutive days, every evening the respondents answered questions of a diary about their physical activity of the past day. The questions covered the duration (0, 1-15, 16-30, 31-60, 61-120, > 120 minutes) of walking inside and outside, biking, gardening, light and heavy household work, and sports. Furthermore, a question about the number of climbed stairs (up and down is one stair) was added (no, 1-5, 6-10, >10 stairs). For statistical analyses the duration of the categories 0, 1-15, 16-30, 31-60, and 61-120 minutes can be corresponded to the mean of the duration of these categories 0, 8, 23, 45, 90, and 120 minutes, respectively, whereas the category >120 minutes corresponded to 121 minutes. The total time spent on physical activity can be calculated by summing up the mean of each activity per day (minutes) across activities. The intensity-weighted total physical activity can be calculated by summing up the mean of categories of each activity (minutes), multiplying this with the specific MET score of each activity, and summing up the MET scores across activities (see documentation physical activity questionnaire).


A pedometer is an instrument that can be attached around the waist and counts the number of steps by assuming that each up/down movement of the trunk represents one step. In this study respondents wore an electronic pedometer (DIGI walker Yamax, Optimal Health Products, U.S.A.) during the same seven days the diary was completed. Respondents were instructed to open the pedometer every evening, to write the number of steps in a notebook, and to push the reset button (zero steps). At night, the pedometer was detached from the waist. The next morning the respondents attached the pedometer to the waist belt again. Subjects who had difficulty opening the pedometer or who did not understand the instructions were asked to wear the pedometer for three consecutive days. During the second visit the interviewer opened the pedometer and wrote the number of steps in the notebook. For statistical analyses the mean number of steps per day can be used.

One year fall follow-up

Between 1999 and 2000, the participants from this substudy have completed a one-year fall-follow-up using fall calendars. The procedure was similar to the procedure of the fall follow-up between 1995/96 and 1998/99. The same falls definitions are available (i.e. time to first fall and time to recurrent falling). For further information see the documentation on falls.


  1. Beld van den, A. W., Huhtaniemi, I. T., Pettersson, S. L., Pols, H. A. P., Grobbee, D. E., Jong de, F. H., Lamberts, S. W. J. (1999). Luteinzing hormone and different genetic variants, as indicators of frailty in healthy elderly men. The Journal of Clinical Endocrinology & Metabolism, 84, 1334-1339.
  2. Hsieh, C. & Phillips, R. (1990). Reliability of manual muscle testing with a computerised dynamometer. Journal of Manipulative and Physiological Therapeutics, 13, 72-82.
  3. Nevitt, M. C., Cummings, S. R., & Hudes, E. S., (1991). Risk factors for injuries falls: a prospective study. Journal of Gerontology, 46, M164-M 170.
  4. Tinetti, M. E., Speechly, M., & Ginter, S. F. (1988). Risk factors for falls among elderly persons living in the community. The New England Journal of Medicine, 319, 1701-1707.
  5. Tromp, A. M., Smit, J. H., Deeg, D. J. H., Bouter, L. M., Lips, P. (1998). Predictors for falls and fractures in the Longitudinal Aging Study Amsterdam. Journal of Bone and Mineral Research, 13, 1932-1939.
  6. Tromp, A. M., Pluijm, S. M. F., Smit, J. H., Deeg., D. J. H., Bouter, L. M., Lips, P. (2000). Fall-risk screening test: a prospective study on predictors for falls in community-dwelling elderly. J Clin Epidemiol 2001; 54: 837-44.
  7. Stel VS, Smit JH, Pluijm SM et al. Consequences of falling in older men and women and risk factors for health service use and functional decline. Age Ageing 2004; 33(1): 58-65.
  8. Stel VS, Smit JH, Pluijm SM et al. Balance and mobility performance as treatable risk factors for recurrent falling in older persons. J Clin Epidemiol 2003; 56(7): 659-68.
  9. Stel VS, Smit JH, Pluijm SM et al. Comparison of the LASA Physical Activity Questionnaire with a 7-day diary and pedometer. J Clin Epidemiol 2004; 57(3): 252-8.

Date of last update: August 2019 by Natasja van Schoor