Inside Our Latest Paper: Predictors of BMI Change Over 10 Years among Midlife and Older Adults


Our new paper, which is published in the journal Social Science & Medicine, gave us an opportunity to learn more about weight change in midlife and older adulthood (ages 50-85). This is related to our general interest in risk for cardiovascular disease (or CVD) as people age – weight gain is common during these periods and increases CVD risk, but weight loss actually places older adults at risk for dementia and injuries, such as hip fractures after a fall. And there is new evidence to suggest that weight fluctuations (weight gain or loss over fairly short periods) have long-term health risks. But very little is known about who is most likely to experience weight gain, loss, or fluctuations during midlife and older adulthood. Knowing who is most likely to experience these changes could help us direct prevention efforts toward these individuals, and reduce their risk for aging-related health problems.

What did we do?

To learn more about who is most likely to experience weight change during midlife and older adulthood, we used an existing longitudinal dataset housed at the New Jersey Institute for Successful Aging (at Rowan School of Osteopathic Medicine, one of our institution’s medical schools). This study, called ORANJ BOWL (for Ongoing Research on Aging in New Jersey: Bettering Opportunities for Wellness in Life), enrolled 5,688 adults between the ages of 50 and 74. These adults were assessed multiple times over the following 10 years using phone and electronic surveys. We examined whether gender, CVD risk status (hypertension and/or type 2 diabetes), depressive symptoms, and social support – assessed at the start of the study- predicted change in BMI over the following 10 years. So for this study, we used secondary data analysis, which means that we asked a new question using a dataset that was already available, rather than collecting new data.

What did we find?

Consistent with existing information, we found that individuals who had CVD risk conditions, high depressive symptoms, and low social support also had a higher BMIs at the start of the study. Overall BMI meaningfully changed over 10 years, and change differed by several of the factors we examined. For example, across genders, adults who began the study with both CVD risk and high depressive symptoms had higher and less stable BMIs over time. Men with low social support and women with high depressive symptoms had consistently high BMIs over time. Moreover, fluctuations in weight were greater among women than men, and among those who began the study with CVD risk conditions then those without. However, none of our predictors were associated with consistent decreases in BMI over 10 years. 

What does this mean and why is it important?

BMI change in midlife and older adulthood is associated with Alzheimer’s Disease, frailty, increased use of health care services, and early mortality. Understanding the factors that predict  BMI change early in these periods can inform the creation and implementation of programs to promote weight stability for midlife and older adults. Our findings indicate that these programs would be most useful among women who already have conditions such as hypertension or type 2 diabetes, though men may benefit specifically from increased social support.

What was it like to work on this project?

This was a wonderful opportunity for our group to use an existing data resource to answer questions we’re interested in, that relate to our work on gender and health risks in midlife. And even though the data were already collected, it was a very long process – IRB and data use applications (with revisions), several months on data analysis and writing the paper, and then it was in the publication review process for 11 months (with revisions). So it still took about a year and a half to complete the project. It’s great to be able to share the paper now and we’re grateful to Dr. Rachel Pruchno and the ORANJ BOWL team for supporting the project.

— Dr. Dani Arigo, CHASE Lab director

Working on this project provided an opportunity for an experience from beginning to end with a secondary analysis. It was exciting to see the similarities and differences between other projects we have worked on previously, especially using longitudinal data (over a ten year period). The findings provided additional insight to changes in BMI among the midlife to older adult population, and I am grateful to have had the chance to be a part of the process! 

— Laura Travers, 2nd-year clinical psychology Ph.D. student 

One of my greatest takeaways from working on this project has been a greater understanding and appreciation of weight and body composition changes in midlife to older adulthood. I found it fascinating that in other populations, weight gain is often undesirable for health and fitness outcomes, but for this population it can be health-protective. I am excited to share these findings!

— Kristen Pasko, 3rd-year clinical psychology Ph.D. student 

It can be a challenge to work with existing data. Sometimes we don’t know exactly what we’re going to get until we receive access to it, and that may require being flexible with research aims and creatively critical in how data is managed. It was exciting to be a part of that process! I appreciate the opportunity to work with the ORANJ BOWL dataset and look forward to spreading the word about our findings.

— Dr. Cole Ainsworth, CHASE Lab program manager

Next week, we’ll have an update on our ongoing work to understand short-term predictors of behavior change among women in midlife with elevated CVD risk – stay tuned!