I would like to thank the editor of International Psychogeriatrics for the opportunity to comment on this article (Unützer et al., Reference Unützer, Patrick, Diehr, Simon, Grembowski and Katon2000). The work leading to this paper was undertaken early in my career and has become the foundation for a decade of research focused on improving the care of older adults with depression.

This paper was initially inspired by the work of one of my mentors, Dr. Kenneth Wells, who had published a seminal article that established depression as a major source of disability on a par with other common medical disorders (Wells et al., Reference Wells1989). Other sources of inspiration came from my mentor, Dr. Wayne Katon, who introduced me to the world of primary care psychiatry and helped me appreciate that while most depressed older adults receive treatment from primary care providers, few of them seek care from a mental health specialist such as a geriatric psychiatrist (Unützer et al., Reference Unützer, Katon, Sullivan and Miranda1999b). Research by other influential investigators early in my career had established high rates of depression in older primary care patients (Schulberg et al., Reference Schulberg, Mulsant, Schulz, Rollman, Houck and Reynolds1998) and poor or ineffective treatment of depression in such patients (Callahan, Reference Callahan2001).

In this context, I set out to examine the impact of depression on older primary care patients. Specifically, I hoped to help put the burden of depression in perspective for primary care providers and policy-makers by comparing it with the burden associated with chronic medical disorders common in late life.

I was fortunate to identify a well-characterized representative data set of community-living older adults who had been prospectively followed with repeated measurements of depression, chronic medical disorders, and health-related quality of life (Patrick et al., Reference Patrick1999) and to have the outstanding support of Dr. Patrick, the principal investigator of the parent study and several experienced colleagues with expertise in clinical epidemiology, health services research, and biostatistics. In addition to documenting the effects of depression on health-related quality of life, we were able to show that in this large sample of community-living older adults, depression was associated with significant increases in total health care costs (Unützer et al., Reference Unützer1997) but that few older adults received effective care for depression (Unützer et al., Reference Unützer1999a).

Why has this article become frequently cited? I believe that our paper helped document the substantial public health burden associated with late-life depression and that it called attention to this problem not only from geriatric psychiatrists and other mental health experts but from primary care and other health care providers who were more familiar with other chronic medical disorders in older adults. By comparing the impact of depression with that of other common medical conditions using a representative sample of community-living older adults, we helped to demonstrate that mental health is an important part of overall health in late life.

In the ten years since the publication of this article, our knowledge of late-life depression in primary care has grown substantially. We have important new evidence about the extent and long-term effects of late-life depression in primary care (Cui et al., Reference Cui, Lyness, Tang, Tu and Conwell2008) and other population-based studies have documented the substantial impact of depression on functioning (Pérès et al., Reference Pérès, Jagger and Mathews2007) and active life expectancy (Reynolds et al., Reference Reynolds, Haley and Kozlenko2008) in older adults. Research from several countries has begun to demonstrate important interactions between depression and other chronic medical disorders (Katon, Reference Katon2003; Moussavi et al., Reference Moussavi, Chatterji, Verdes, Tandon, Patel and Ustun2007; Pérès et al., Reference Pérès, Jagger and Mathews2007).

Our original work on the impact of and quality of care for late-life depression also laid the groundwork for what has become a decade of research on ways to improve care for depressed older adults. In recent years we have seen the development of several evidence-based programs in which primary care providers and mental health specialists collaborate effectively to improve care for late-life depression (Unützer et al., Reference Unützer2002; Bruce et al., Reference Bruce2004). These programs have been found to be effective not only across a range of health care settings but also in patients with and without comorbid medical disorders (Harpole et al., Reference Harpole2005). They have also been found to be more cost-effective than usual care for depressed older adults (Katon et al., Reference Katon2005) with potential savings in total health care costs over the long run (Unützer et al., Reference Unützer2008a).

While we have made important progress in understanding the impact and treatment of late-life depression in primary care, substantial challenges persist. Important structural, policy, and financial barriers must be overcome to allow the widespread implementation of evidence-based collaborative care programs for depression. And while well-implemented collaborative care programs can more than double the likelihood that a depressed older person experiences a substantial reduction in their depressive symptoms, (Unützer et al., Reference Unützer2002) many participants in trials of collaborative care for late-life depression have persistent depression symptoms despite fairly intensive pharmacologic and nonpharmacologic treatment for depression.

One of the strongest predictors of poor treatment response is the presence of functionally disabling chronic physical pain (Thielke et al., Reference Thielke, Fan, Sullivan and Unützer2007). The combination of emotional pain associated with depression and physical pain often associated with osteoarthritis in older adults seems to be particularly challenging. Interestingly, this finding takes us back to our original paper in which we found that at a population level, arthritis was associated with the greatest losses in quality adjusted life years. The effective treatment of older adults who have combinations of emotional and physical pain creates important new clinical challenges for those of us who are trained as mental health specialists. While most of us would agree with the World Health Organization's definition of health as “a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity,” few of us are equipped to address the full range of physical, emotional, and social challenges faced by our older patients.

Integrated models of care in which professionals with different areas of expertise effectively collaborate to care for older adults with depression and other comorbid medical and social challenges are likely to be among the most promising avenues for improving health among older depressed patients. Recent work by Katon and colleagues (2005) demonstrates the feasibility of collaborative care programs that address not only common mental disorders such as depression but also important chronic physical problems such as diabetes or heart disease. We have also recently developed and pilot-tested a promising new program of care management for older primary care patients with depression and functionally disabling physical pain from arthritis. In this program, a nurse care manager working closely with the patient's primary care physician and a consulting geriatric psychiatrist was able to achieve substantial reductions in both depression and physical pain (Unützer et al., Reference Unützer2008b). More work is needed on such combined intervention strategies.

Returning to the original article (Unützer et al., Reference Unützer, Patrick, Diehr, Simon, Grembowski and Katon2000), I would like to share one final “lesson” from this work. The data for this study came from a large randomized controlled trial (RCT) of health prevention for community-living Medicare recipients (Patrick et al. Reference Patrick1999). While well intended and well executed, this large RCT of health prevention strategies for older adults turned out to be a “negative trial.” Our paper was a secondary data analysis that was conducted several years after the conclusion of the original study. I point this out to encourage young investigators to seek out and make use of existing data sources for their initial work. If well done, such “secondary analyses” of existing data can make important contributions to science, become frequently cited, and help launch the career of a young investigator, especially if he or she is as fortunate as I was at the time to receive the guidance and technical support from an experienced team of researchers who helped coauthor this paper with me.