The road to ADL disability for older persons generally follows one of two paths. The first is characterized by catastrophic events that strike previously healthy persons and quickly leave them temporarily or permanently dependent. Such dependency often occurs after a traumatic event (such as hip fracture), an acute medical illness (such as a stroke), or a surgical procedure with serious complications. For physicians, the key to preventing this route to disability is the conscientious practice of preventive medicine that addresses risk factors for these calamities and minimizes the complications (sometimes iatrogenic) of acute medical and surgical illness.

The second route to disability is a more incremental, and sometimes insidious, path along which older persons progressively lose function as a result of disease and deconditioning. These insults are aggravated by the loss of physiologic reserves that occurs with age [3]. The stages of progression along this path are not well defined and may vary from person to person and among ADLs.

In a state of good health, we perform many of these basic activities (except perhaps for selecting clothing and preparing meals) so routinely that they barely enter our consciousness. Perhaps the first level of compromise is a heightened awareness that a task needs to be done and that some planning may be necessary. The ensuing sequence of decline may then be difficulty (self-reported or observed) performing the task; the need for assistive devices; and, finally, dependency to the extent that another person's help is necessary to get the job done.

In this issue, Gill and colleagues [4] have made an important contribution to understanding the sequence of the chronic pathway of functional decline. They have isolated a middle step, independence with difficulty, that occurs between independence and dependence and that has prognostic significance for functional status decline, health care utilization, nursing home placement, and survival. The link between difficulty in performing functional tasks and subsequent adverse health outcomes is not new. In fact, some functional status scales and national surveys have used degree of difficulty to frame the response items [1, 5]. Moreover, the link between difficulty with performing ADL function and subsequent adverse health outcomes, including death, has been previously reported [6]. Thus, the study's principal findings come as no great surprise. Nevertheless, Gill and colleagues' elegant cohort study provides convincing evidence that this intermediate state can be easily identified and can provide additional clinical information. They also clearly state that "difficulty" should not replace "dependence" as the sole definition of impaired function. Not only does the dependent state carry a worse prognosis (as shown by Gill and colleagues), but the economic and resource implications of dependency are much more profound. Once personal assistance is needed for ADLs, the range of options for living situations becomes much more limited.

As implied from the above theoretical model of chronic functional decline, there may be yet more information about functional status that helps identify groups that are independent but have a higher risk for functional decline. Among the most promising methods are performance-based measures, which have also shown prognostic value [7-10] for adverse outcomes and death. It has been speculated that measurable observations (for example, slowing of gait velocity) can detect "preclinical" impairment even before the patient is aware of decline [11]. Performance-based measures may be the earliest harbingers of decline in some functional tasks in some older persons. Conversely, for other functional tasks, the patient may be able to sense an impending problem earlier. The additional value of combining performance-based measures and such self-report measures as independence with difficulty will probably be the topic for future research on functional status.

So how does this new insight into the prognostic importance of self-reported difficulty affect the care of older persons? First, Gill and colleagues identified a set of follow-up questions that should be asked when older persons deny that they have functional dependency. This new intermediate state is common; in this study, 21% of patients were affected. However, most clinicians will groan at the thought of adding to the large amount of information that must be collected from older patients. A reasonable and efficient approach would be to probe into the patient's functional status by using previsit questionnaires that are completed by patients and can be reviewed by the clinician [12].

Once identified, independence with difficulty may be remediable. Interventions implemented earlier in the cascade that leads from "independence without a thought" to dependency may slow or temporarily reverse the progression of decline. Such interventions may range from improved management of diseases (for example, congestive heart failure or chronic obstructive pulmonary disease) to earlier (that is, before the onset of disability) rehabilitative efforts through endurance and strength training. The prognostic information gained by knowing about difficulty with ADLs should also be discussed with patients and, when appropriate, families. Knowing that they are at increased risk may help motivate patients to more actively participate in exercise programs or increase their adherence to prescribed therapy. Nevertheless, such optimistic scenarios must be considered speculative. It remains to be determined which, if any, interventions will be effective in reversing difficulty and preventing subsequent disability.

In a state of good health, we subconsciously take our ability to complete ADLs for granted. When performing these tasks begins to be a chore, it is a strong signal that decline has already occurred.