Patterns of Functional Impairment over Time Amongst Older Men—the Manitoba Follow-up Study

Philip D. St John, MD, FRCPC1,2, Audrey U. Swift, PhD3, Robert B. Tate, PhD2,3

1Section of Geriatrics, Max Rady College of Medicine, University of Manitoba, Winnipeg, MB
2Centre on Aging, University of Manitoba, Winnipeg, MB
3Dept of Community Health Sciences, Max Rady College of Medicine, University of Manitoba, Winnipeg, MB




To determine the incidence and prevalence patterns of activity of daily living (ADL) impairments in ageing men.


3,983 men were enrolled in the Manitoba Follow-up Study (MFUS) cohort study in 1948. From 1996 onwards, functional status was measured. We classified basic (BADL) and instrumental (IADL) into mutually exclusive categories as a time dependant factor after the second survey wave as: First survey response; no limitation; incident (first episode of disability); persistent (limitation which was seen on all questionnaires after the incident episode); resilient (noted in previous surveys but not present); and recurrent (noted in present survey, and limitations noted as present and absent in previous surveys).


There were 1,745 participants in 1996 at a mean age of 76 years. Incident BADL limitations increased substantially with age: from 1% at age 75 to 15% at age 95. Similarly, persistent limitations increased with age: from 0.4% at age 75 to 18% at age 95. However, BADL function was fluid, with many individuals grouped within the resilient and recurrent patterns. Similar age effects and variability were noted in IADLs.


New and persistent disabilities are highly associated with age. However, there is considerable change in functional status over time.

Key words: ageing, men, disability, functional status, cohort study, Manitoba Follow-up Study


Impaired functional status has been shown to predict death,(1,2) depression,(3) falls,(4) long-term care utilization,(5) and hospitalization(6,7) in community-living older adults. In hospitalized older adults, impaired functional status at admission predicts a long length of stay(8,9) and death,(10,11) and has been considered as an important outcome in randomized clinical trials of models of care for hospitalized older adults.(12) Moreover, older adults value highly being able to maintain a high level of functioning.(13) Understanding the epidemiology of disability in population of older adults is, therefore, important for planning health-care and social care services.

To date, a considerable body of population-based and clinic-based studies shows that functional impairment is closely related to increasing age,(14) female sex,(14) cognitive impairment,(15) depression,(16) and to an increased burden of chronic illness.(17) Most of this research has considered the point prevalence of disability. This is likely because there are relatively few prospective cohort studies with frequent measures of functional status over time. However, functional status may not be constant over time, but may fluctuate and vary over time. In analyses of the Precipitating Events Project (PEP) study, Han et al.(18) have found that functional status is fairly dynamic with considerable variability over time. Furthermore, there are highly heterogeneous trajectories of disability.(18) Thus, presenting a point prevalence of disability at any one time may be an incomplete measure, since functional status may be variable, and the course that led up to the current state of functional impairment may be variable. Some of the individuals with impairment may have been functionally intact shortly prior to the survey, and some may have been disabled for prolonged periods. Conversely, some of the functionally intact individuals may have been intact, or have been impaired on previous occasions.

Understanding patterns of impairment is important for several reasons. First, the epidemiological study of functional loss may need to consider the trajectory that a population has followed prior to a measurement at a particular point in time—and not simply the point prevalence at a particular time. Disability may be a dynamic consideration, unlike the occurrence of a chronic illness, which is usually considered present from diagnosis until death. Similarly, considering only the point estimate of disability also complicates the study of the incidence of disability, since episodes of disability may be recurrent. Second, clinicians may want to understand the course of disability in their patients in planning care and in treatment considerations. Third, health and social care planning requires an understanding not only of the prevalence of disability, but of the path to disability taken by an individual or a population. If disability is stable, then services such as home care will be relatively predictable, and the needs can be anticipated in advance and services altered slowly. However, if disability is dynamic, then home care services must be flexible and able to respond quickly to changes in functional status. Finally, understanding patterns of functional status is important for older adults and their families. Since maintaining functional status is highly valued by older persons, understanding how older adults recover from a period of disability is likely important.

To address some of these issues, we will conduct an analysis of burden of disability in a cohort of men as they age. The Manitoba Follow-up Study (MFUS)(19) is a cohort study of men who were found fit for aircrew training during the Second World War and who have been followed regularly since then. Beginning in 1996, functional status has also been assessed regularly. The specific objectives of these analyses are: 1) To determine the point prevalence of limitations in basic activities of daily living (BADL) and instrumental ADL (IADL) between the ages of 75 to 100; and 2) To describe the pattern of BADL and IADL impairment between the ages of 75 and 100.


The Manitoba Follow-up Study (MFUS) cohort consists of 3,983 men who qualified for air crew training in the Royal Canadian Air Force during the Second World War. When the cohort was sealed on 1 July 1948, their mean age was 31 years, with 90% between ages 20 and 39 years. All study members were free of clinical evidence of ischemic heart disease. The protocol of MFUS was to obtain routine medical examinations from these men at regular intervals over time. The initial research goal of the study was to examine the role that any non-specific abnormalities detected on routine electrocardiograms from apparently healthy men might play in the prediction of subsequent diagnoses of cardiovascular disease. The research focus was expanded in 1996 to explore the roles of physical, mental, and social functioning. By the spring of 2017, 124 of the participants were still alive and responding to the survey, and the mean age was 94 years old. At the time of these analyses, there were 22 participants who have been lost to follow-up (defined as having no contact during the last three years). Details of the study have been reported elsewhere.(19)

The focus of MFUS was expanded in 1996 when a Successful Aging Questionnaire (SAQ) was added.(20,21) The SAQ was conducted in 1996, 2002, 2004, and annually since then. The SAQ is sent only to those participants living in the community, and is not sent to those living in long-term care or to those who ask not to receive the SAQ. Core components of the SAQ in all years have been: living arrangements, marital status, items of social engagement, self-rated health (SRH), items of life satisfaction, a definition and self-assessment of successful aging, and the Short Form-36 (SF-36).(22) The SAQ also includes measures of ADLs and IADLs. These measures were based on the items used in the Aging in Manitoba Study (AIM),(23) which included items from the Older American Resource Survey (OARS),(24) as well as unique items added into the study survey.

Specifically, participants were asked:

“Now I have some questions about your ability to carry on different activities. I am interested in your capability, not whether or not you actually do them. Are you capable of … without any help from anyone else?”

The response categories were yes/no. We considered BADLs as: getting up out of chair and walking 3 meters, using the toilet, eating, dressing and putting shoes on, buttoning a sweater, washing or bathing or grooming, getting in and out of bed, going out doors in good weather, getting around the house, and going up and down the stairs. We considered IADLs to be such activities as doing light house work, making a cup of tea or coffee, preparing a hot meal, shopping, managing financial affairs (banking, paying bills), taking medications or treatments, and using the telephone. We considered one or more BADL impairment as “BADL impairment” and two or more IADL impairments as “IADL impairment.”

To determine the pattern of ADL impairment, we used an approach similar to the approach we used to categorize other time varying characteristics in the MFUS, specifically atrial fibrillation.(25) We consider a participant’s state at a particular survey year based upon their state at that time and also their state in previous survey waves. The seven categories we created are:

We then tabulated these categories by the year of the survey. Since the prevalence of disability at a particular age is more relevant than at a calendar year, we currently present the graphs of functional status by age at the time of the survey response. Since the age of the cohort is relatively homogeneous, the results by calendar year are similar to the results by age. In the graphs for the pattern of disability, we have excluded the participants’ first survey response, since the pattern of disability is only apparent after the baseline status is known. Graphs by year with the initial survey response are available upon request.


The characteristics of men responding to each survey are shown in Table 1. In 1996, 2,043 questionnaires were mailed out, and 1,745 were returned with complete responses on functional status. As the cohort aged, participants died, and as of the spring of 2017, there were 124 surveys sent out. The annual completion rate has consistently been higher than 80%. We believe that these missing responses are not completely at random, but reflect non-response due to illness in most cases. The survey is sent only to those who are still living in the community. The number of deaths increased as a percentage of the remaining sample, as one would expect. Note that these are participants who were known to have died in the interval between the SAQ survey waves, at the time the SAQ was mailed to the participant. It does not include those who were identified as having died a later date; many of the deaths are noted when the SAQ is returned by family.

TABLE 1 Characteristics of the Manitoba Follow-up Study cohort from 1996 through 2017


Over the course of the study, the proportion of individuals with impairments in ADLs and IADLs increased steadily. As expected, IADL impairments were more common than basic ADL impairments. Note that the number of participants who have never had an episode of disability declines steadily over the years of the study, so that by the end of the period of observation, the majority of the sample had experienced a period of IADL impairment, and many had also experienced a period of ADL impairment.

At age 75, the point prevalence of disability was low: 77% had no disability; 14% had only IADL disability; 3% had only BADL impairment; and 6% had both IADL and BADL impairment. By age 95, however, 26% had no impairment; 38% had only IADL impairment; 4% had only BADL impairment, and 32% had both BADL and IADL impairment. The association between age and functional status is shown in Figure 1. There is a clear and strong association between the point prevalence of disability and age. In younger men (i.e., those 75 to 80), impairments are very rare; but by the age of 90, most participants had some IADL or BADL impairment. Also as expected, IADL impairments were more commonly seen than BADL impairments. There were very few men with impairments only in IADLs.



FIGURE 1 Point prevalence of functional impairment by age

When we consider the pattern of ADLs, a more nuanced picture emerges. At age 75, very few participants had persistent deficits, and the incidence of BADL was very low. However, incidence and prevalence increased steadily with advancing age. As well, the proportion of participants with recurrent and resilient patterns increased with age (Figure 2). A similar pattern was seen with IADLs (Figure 3). BADL impairment which persists through all study waves is less common than IADL impairments, and considerable numbers of participants exhibit resilient and recurrent patterns (Figures 2 and 3). Indeed, the combination of these two states is more common than persistent disability. Also of note, even in very old men, the prevalence of persistent BADL is fairly low.



FIGURE 2 BADL impairment categories by age



FIGURE 3 IADL impairment categories by age


We have examined the point prevalence and prevalence pattern of functional impairment in a prospective cohort of ageing men. We have found a very strong and linear association between age and limitations in both BADLs and IADLs, and that IADL limitation is more common than BADL limitation. However, the point prevalence of functional impairment may be an incomplete reflection of functional impairment among individuals and populations, since there are different patterns of limitation. At any time, those with resilient and recurrent patterns of functional impairment are common.

There are some strengths and limitations to our approach. First, this is a large, established cohort with few missing responses. However, the missing responses to the survey are unlikely to be at random, but due to illness during that survey wave. Thus, we may be underestimating the point prevalence of disability. The effect of missing variables in the estimation of the pattern of disability is less clear. Second, there are frequent sampling intervals. Most large epidemiological surveys have sampling intervals of three to five years. Fluctuations in function during this time period may not be apparent, so it may be difficult to assign a pattern of disability to an individual participant. With frequent sampling intervals, it is easier to detect rapid fluctuations in function. Third, MFUS considers measures of both IADLs and BADLs with measures similar to other studies. We do not, however, use the OARS scale itself, but items drawn from it. We also did not consider the use of assistive devices in the measurement of BADL, as the original OARS did. Thus, the results may differ somewhat from other studies which use the OARS.

There are also limitations. First, the MFUS cohort consists of only men, and we cannot ascertain the prevalence or patterns of disability amongst ageing women. Second, this cohort was healthy enough to qualify for air crew,(26,27) and have shared life experiences which may have shaped late-life health and not be generalizable to other older populations—notably those outside Canada and women. Finally, accounting for the effect of death on patterns of disability is complicated. Since functional status predicts death very strongly, it is likely that those who die experience a period of disability prior to death. They are also removed from the pool of observed individuals, making interpretation of the pattern of disability prior to death complicated. This may also partly explain the low numbers of participants with persistent disability. If incident disability is associated with a high death rate, coupled with a short survival time in a state of disability, then there may be an incidence-prevalence (Berkson’s) bias(28) in the prevalence estimates of disability, particularly persistent disability. This is also true for the men who were not sent an SAQ, most of whom were in long-term care.

Our results are similar to other studies of functional impairment amongst older populations. The effect of age on functional status has been widely accepted, and is not surprising. Nevertheless, it remains important. By very late life, most individuals have functional impairment. The focus of health interventions to date has focused on preventing disability in late life, and disability rates seem to have declined modestly in recent decades.(29) This is clearly very important. However, even with dramatic reductions in late-life disability, there will still be a high proportion of older adults with disability who require health and social care. Any future reduction in disability may be offset by the growing proportion of very old adults in the population. Ensuring adequate attention to those with functional impairment is, therefore, important.

Our second finding of fluctuating functional status, with substantial numbers of individuals recovering from disability and/or experiencing recurring episodes of disability, is perhaps more important. It is important to note that, at most periods in the study, there were more participants with recurrent or resilient patterns than with persistent disability. To date, service planning has largely considered the point prevalence of disability and assumed that the pattern was fairly stable over time. As this may not be true, reconsidering service delivery may be required. Notably, services such as home care may need to adapt to rapid changes in functional status, and geared up or down depending on the course of functional status. Moreover, these fluctuations may be heterogeneous to an individual, so services may need to be tailored to an individual rather than to a population. Providing rehabilitation and geriatric assessment in hospital(12) and the community(30) may also be important, and may also need to be responsive and adaptive. Home care programs may need to deliver rehabilitation, as well as care provision, in order to maximize recovery from acute disability.(31,32)

Finally, the effect of the pattern of functional status on older adults, their families, and their caregivers may be important. Functional impairment is associated with depressive symptoms and other measures of well-being. The effect of functional recovery on well-being merits further attention. As well, the pattern of recurring disability on well-being also merits further study. Some hope of functional recovery may be encouraging. On the other hand, unpredictable episodes of functional decline may be quite distressing for older adults, and the uncertainty around the precipitants of these declines may be further distressing. Further study into the factors that predict and precipitate these declines, as well as the effect of these declines on well-being, is needed in order to more broadly understand late-life health and well-being.

Our findings also suggest other future research strategies for this and other longitudinal studies. First, the relatively frequent sampling time intervals allow for the study of the natural history of new onset functional impairment—both the risk factors and the outcomes. This may be particularly important in the study of disease states with a high mortality, or in the study of populations with a high baseline mortality. Second, determining the trajectories and/or transitions of functional status is possible. Analytic strategies such as latent class models, joint trajectory models, multilevel models, and multistate modelling can be used to investigate the predictors and consequences of the trajectories and transitions. This information could prove useful in planning services and understanding risk factors for those with changing functional status.


These analyses were funded by the Canadian Institutes of Health Research (CIHR) Project Grant Number PJT-152874. The Manitoba Follow-up Study also received Bridge Funding from the Dean’s Fund (Max Rady College of Medicine, University of Manitoba) and charitable donations from the participants and families of the Manitoba Follow-up Study. These sources of funding had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; in the preparation of the manuscript; or in the review or approval of the manuscript.


The authors declare that no conflicts of interest exist.


1 Marengoni A, von Strauss E, Rizzuto D, et al.. The impact of chronic multimorbidity and disability on functional decline and survival in elderly persons. A community-based, longitudinal study. J Intern Med. 2009;265(2):288–95.. Epub 2009/02/05..
cross-ref  pubmed  

2 St John PD, Tyas SL, Menec V, et al. Multimorbidity, disability, and mortality in community-dwelling older adults. Can Fam Phys. 2014;60(5):e272–80. Epub 2014/05/16.

3 Lenze EJ, Rogers JC, Martire LM, et al. The association of late-life depression and anxiety with physical disability: a review of the literature and prospectus for future research. Am J Geriatr Psychiatry. 2001;9(2):113–35. Epub 2001/04/24.
cross-ref  pubmed  

4 Welmer AK, Rizzuto D, Laukka EJ, et al. Cognitive and physical function in relation to the risk of injurious falls in older adults: a population-based study. J Gerontol A Biol Sci Med Sci. 2017;72(5):669–75. Epub 2016/07/28.

5 Gaugler JE, Duval S, Anderson KA, et al. Predicting nursing home admission in the U.S: a meta-analysis. BMC Geriatr. 2007;7:Article No.13. Epub 2007/06/21.
cross-ref  pubmed  pmc  

6 Callahan KE, Lovato L, Miller ME, et al. Self-reported physical function as a predictor of hospitalization in the Lifestyle Interventions and Independence for Elders Study. J Am Geriatr Soc. 2018;66(10):1927–33. Epub 2018/10/04.
cross-ref  pubmed  pmc  

7 Na L, Pan Q, Xie D, et al. Activity limitation stages are associated with risk of hospitalization among Medicare beneficiaries. PM & R. 2017;9(5):433–43. Epub 2016/10/23.

8 Turcotte LA, Perlman CM, Fries BE, et al. Clinical predictors of protracted length of stay in Ontario Complex Continuing Care hospitals. BMC Health Serv Res. 2019;19(1):218. Epub 2019/04/07.
cross-ref  pubmed  pmc  

9 Arnold J, Exton-Smith AN. The geriatric department and the community value of hospital treatment in the elderly. The Lancet. 1962;280(7255):551–53.

10 Covinsky KE, Justice AC, Rosenthal GE, et al. Measuring prognosis and case mix in hospitalized elders. The importance of functional status. J Gen Intern Med. 1997;12(4):203–08. Epub 1997/04/01.
pubmed  pmc  

11 Inouye SK, Peduzzi PN, Robison JT, et al. Importance of functional measures in predicting mortality among older hospitalized patients. JAMA. 1998;279(15):1187–93. Epub 1998/04/29.
cross-ref  pubmed  

12 Ellis G, Whitehead MA, Robinson D, et al. Comprehensive geriatric assessment for older adults admitted to hospital: meta-analysis of randomised controlled trials. BMJ. 2011;343:d6553. Epub 2011/10/29.
cross-ref  pubmed  pmc  

13 Fried TR, McGraw S, Agostini JV, et al. Views of older persons with multiple morbidities on competing outcomes and clinical decision-making. J Am Geriatr Soc. 2008 Oct;56(10):1839–44. Epub 2008/09/06.
cross-ref  pubmed  pmc  

14 Courtney-Long EA, Carroll DD, Zhang QC, et al. Prevalence of Disability and Disability Type Among Adults--United States, 2013. Morbid Mortal Wkly Rep. 2015;64(29):777–83. Epub 2015/08/01.

15 St John PD, Tyas SL, Montgomery PR. Cognition, even in the normal range, predicts disability: cross-sectional and prospective analyses of a population-based sample. Int J Geriatr Psychiatry. 2015;30(10):1008–16. Epub 2015/02/03.
cross-ref  pubmed  

16 Murphy RA, Hagaman AK, Reinders I, et al. Depressive trajectories and risk of disability and mortality in older adults: longitudinal findings from the Health, Aging, and Body Composition Study. J Gerontol Series A Biol Sci Med Sci. 2016;71(2):228–35. Epub 2015/08/15.

17 St John PD, Tyas SL, Menec V, et al. Multimorbidity predicts functional decline in community-dwelling older adults: Prospective cohort study. Can Fam Phys. 2019;65(2):e56–e63. Epub 2019/02/16.

18 Han L, Allore H, Murphy T, et al. Dynamics of functional aging based on latent-class trajectories of activities of daily living. Ann Epidemiol. 2013;23(2):87–92. Epub 2013/01/12.
cross-ref  pubmed  pmc  

19 Tate RB, Cuddy TE, Mathewson FA. Cohort Profile: The Manitoba Follow-up Study (MFUS). Int J Epidemiol. 2015;44(5):1528–36. Epub 2014/07/30.

20 Tate RB, Lah L, Cuddy TE. Definition of successful aging by elderly Canadian males: the Manitoba Follow-up Study. Gerontologist. 2003;43(5):735–44. Epub 2003/10/23.
cross-ref  pubmed  

21 Tate RB, Swift AU, Bayomi DJ. Older men’s lay definitions of successful aging over time: the Manitoba Follow-up Study. Int J Aging Hum Dev. 2013;76(4):297–322. Epub 2013/07/17.
cross-ref  pubmed  

22 Ware JE, Jr., Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care. 1992;30(6):473–83. Epub 1992/06/11.
cross-ref  pubmed  

23 Chipperfield JG, Havens B, Doig WD. Method and description of the Aging in Manitoba Project: a 20-year longitudinal study. Can J Aging. 1997;16(4):606–25.

24 Fillenbaum GG, Smyer MA. The development, validity, and reliability of the OARS Multidimensional Functional Assessment Questionnaire. J Gerontol. 1981;36(4):428–34.
cross-ref  pubmed  

25 McIntyre WF, John PDS, Torabi M, et al. Lifetime pattern of atrial fibrillation and the risks of stroke and death in a population-based cohort of men (from The Manitoba Follow-Up Study). Am J Cardiol. 2018;122(10):1688–93. Epub 2018/09/16.
cross-ref  pubmed  

26 Mathewson FA. Medical aspects of aircrew selection. Can Med Assoc J. 1942;47(4):318–22.
pubmed  pmc  

27 Mathewson FA, Wilson DR, Cameron WJ, et al. Medical classification of RCAF groundcrew on the basis of functional requirements. Can Med Assoc J. 1943;49(6):512–5.
pubmed  pmc  

28 Westreich D. Berkson’s bias, selection bias, and missing data. Epidemiology. 2012;23(1):159–64. Epub 2011/11/15.

29 Grasset L, Jacqmin-Gadda H, Proust-Lima C, et al. Temporal trends in the level and decline of cognition and disability in an elderly population: the PAQUID Study. Am J Epidemiol. 2018;187(10):2168–76. Epub 2018/06/13.
cross-ref  pubmed  

30 Beswick AD, Rees K, Dieppe P, et al. Complex interventions to improve physical function and maintain independent living in elderly people: a systematic review and meta-analysis. Lancet. 2008;371(9614):725–35. Epub 2008/03/04.
cross-ref  pubmed  pmc  

31 Armstrong JJ, Sims-Gould J, Stolee P. Allocation of rehabilitation services for older adults in the Ontario home care system. Physiother Can. 2016;68(4):346–54. Epub 2016/12/03.
cross-ref  pubmed  pmc  

32 Armstrong JJ, Zhu M, Hirdes JP, et al. Rehabilitation therapies for older clients of the Ontario home care system: regional variation and client-level predictors of service provision. Disabil Rehabil. 2015;37(7):625–31. Epub 2014/07/02.

Correspondence to: Philip St John, MD, FRCPC, Section of Geriatrics, Max Rady College of Medicine, University of Manitoba, GE 547 Health Sciences Centre, 820 Sherbrook St., Winnipeg, MB R2A 1R9,

(Return to Top)

Canadian Geriatrics Journal, Vol. 24, No. 2, June 2021