Diabetes Care Volume 42, April 2019 507 Guowei Li,1,2,3 Jerilynn C. Prior,4 Frailty and Risk of Fractures in CLIN CARE/EDUCATION/NUTRITION/PSYCHOSOCIAL William D. Leslie,5 Lehana Thabane,2,3 Patients With Type 2 Diabetes Alexandra Papaioannou,2,6 Robert G. Josse,7 Stephanie M. Kaiser,8 9 Diabetes Care 2019;42:507–513 | https://doi.org/10.2337/dc18-1965 Christopher S. Kovacs, Tassos Anastassiades,10 Tanveer Towheed,10 K. Shawn Davison,11 Mitchell Levine,2,3,6 David Goltzman,12 and Jonathan D. Adachi,3,6 for the CaMos Research Group OBJECTIVE We aimed to explore whether frailty was associated with fracture risk and whether frailty could modify the propensity of type 2 diabetes toward increased risk of fractures. RESEARCH DESIGN AND METHODS Data were from a prospective cohort study. Our primary outcome was time to the first incident clinical fragility fracture; secondary outcomes included time to hip 1Center for Clinical Epidemiology and Method- fracture and to clinical spine fracture. Frailty status was measured by a Frailty Index ology, Guangdong Second Provincial General (FI) of deficit accumulation. The Cox model incorporating an interaction term Hospital, Guangzhou, China (frailty 3 diabetes) was used for analyses. 2Department of Health Research Methods, Evi- dence, and Impact, McMaster University, Ham- RESULTS ilton, Ontario, Canada 3St. Joseph’s Healthcare Hamilton, Hamilton, The analysis included 3,149 (70% women) participants; 138 (60% women) had Ontario, Canada diabetes. Higher bone mineral density and FI were observed in participants with 4Department of Medicine, University of British diabetes compared with control subjects. A significant relationship between the FI Columbia, Vancouver, British Columbia, Canada 5 and the risk of incident fragility fractures was found, with a hazard ratio (HR) of 1.02 Departments of Medicine and Radiology, Uni- versity of Manitoba, Winnipeg, Manitoba, (95% CI 1.01–1.03) and 1.19 (95% CI 1.10–1.33) for per-0.01 and per-0.10 FI increase, Canada respectively. The interaction was also statistically significant (P = 0.018). The HR for 6Department of Medicine, McMaster University, per-0.1 increase in the FI was 1.33 for participants with diabetes and 1.19 for those Hamilton, Ontario, Canada 7 without diabetes if combining the estimate for the FI itself with the estimate from Department of Medicine, University of Toronto, Toronto, Ontario, Canada the interaction term. No evidence of interaction between frailty and diabetes was 8Department of Medicine, Dalhousie University, found for risk of hip and clinical spine fractures. Halifax, Nova Scotia, Canada 9Faculty of Medicine, Memorial University of CONCLUSIONS Newfoundland, St. John’s, Newfoundland and Participants with type 2 diabetes were significantly frailer than individuals without Labrador, Canada 10Department of Medicine, Queen’s University, diabetes. Frailty increases the risk of fragility fracture and enhances the effect of Kingston, Ontario, Canada diabetes on fragility fractures. Particular attention should be paid to diabetes as a 11Saskatoon Osteoporosis and CaMos Centre, risk factor for fragility fractures in those who are frail. Saskatoon, Saskatchewan, Canada 12Department of Medicine, McGill University, Montreal,´ Quebec,´ Canada Fragility fractures are a skeletal complication associated with type 2 diabetes, Corresponding author: Guowei Li, lig28@ resulting in substantial morbidity, hospitalization, high health care costs, impaired mcmaster.ca quality of life, disability, and death (1). Type 2 diabetes itself is reported to be an Received 17 September 2018 and accepted 7 independent risk factor for fractures. For instance, the risk of a hip fracture in patients January 2019 with type 2 diabetes is ;70% higher than in individuals without diabetes (2). In type 2 This article contains Supplementary Data online diabetes, bone mineral density (BMD) is usually higher than in individuals without at http://care.diabetesjournals.org/lookup/suppl/ diabetes, and their BMI is often increased, both of which are typically protective doi:10.2337/dc18-1965/-/DC1. factors for most fractures (3,4). Some studies suggested that even in the presence of © 2019 by the American Diabetes Association. normal or increased BMD, other factors, including poor glycemic control, abnormal Readers may use this article as long as the work isproperlycited,theuseiseducationalandnotfor bone turnover, and bone loss, may explain the increased risk of fractures in type 2 profit, and the work is not altered. More infor- diabetes (5–8). Nevertheless, the “diabetes bone paradox” (high risk of fracture but mation is available at http://www.diabetesjournals normal or increased BMD) in type 2 diabetes remains to be further investigated (9). .org/content/license. 508 Frailty and Fracture Risk in Type 2 Diabetes Diabetes Care Volume 42, April 2019 Frailty is a dynamic aging condition of collected from the questionnaires in- an incident hip fracture and to clinical increased vulnerability affecting psycho- cluded demographic information, self- spine fracture during follow-up. All in- logical, physical, and social functioning reported health conditions, history of cident fractures were documented on (10,11). Emerging evidence has shown family and personal fractures, medica- the annual mailed questionnaires by that increased frailty is significantly re- tion use, dietary intake, lifestyle infor- participants’ self-reports. An individual lated to higher risks for adverse health mation, and quality of life. The physical who reported an incident fracture was outcomes; thus, capturing the degree of examination included height and weight contacted to seek consent to obtain the frailty could quantify an individual’s risks measures and DXA at the lumbar spine medical report and/or hospital discharge of adverse outcomes and may predict and hip. BMD was assessed by DXA at the for verification. Medical or radiographic their responses to therapeutic interven- lumbar spine and femoral neck regions, validation was available for 78% of all tions (12–14). The concept of frailty and and BMD T-scores were calculated based reported incident fractures in CaMos the risk of adverse outcome relies on the on the Canadian standards (16). In ad- (18). fact that the frailer an individual is, the dition, a short questionnaire was mailed greater the likelihood that the person will at annual follow-up to collect data on Independent Variables experience adverse health outcomes in participants’ fractures, hospitalization, Frailty status was measured by the the future (12). It is therefore possible and use of bone health medications. CaMos-based algorithm for Frailty Index that measuring frailty status may assist in Fasting blood samples were collected at (FI) of deficit accumulation in this the understanding of the diabetes bone baseline at one center (Quebec´ City), at study. Details on the construction and paradox. Specifically, our hypotheses in- year 5 at three centers (Calgary, Hamilton, scoring of the FI have been provided cluded 1) patients with type 2 diabetes and Quebec´ City), and at year 10 at eight elsewhere (19). In brief, the original FI may be frailer than individuals without centers (Vancouver, Calgary, Saskatoon, included 30 health-related deficits, with diabetes despite their greater BMD, and Kingston, Toronto, Quebec´ City, St. John’s, each deficit scoring from 0 to 1. Dichot- 2) frailty may modify the propensity of and Halifax). omous deficits were coded as 1 (present) those with type 2 diabetes toward an in- Participants self-reported whether or 0 (absent). Deficits with multilevel creased risk of fractures. In this study, we they had diabetes that was “insulin de- responses were polychotomized to re- used the data from the Canadian Multi- pendent” (to imply type 1 diabetes) or ceive equal points to map the interval 0 to centre Osteoporosis Study (CaMos) to as- “insulin-independent” (to imply type 2 1 (e.g., excellent = 0; very good = 0.25; sess the relationship between frailty and diabetes) at baseline, year 5, and year 10. good = 0.5; fair = 0.75; poor = 1) or the risk of incident fractures in patients During this study, insulin began to be combined into logical groups (e.g., same/ with type 2 diabetes. used in those with type 2 diabetes; thus, somewhat better/better = 0, somewhat more than 98% of participants likely had worse/worse = 1). The FI was then cal- type 2 diabetes (subsequently called “di- culated by summing the scores of all RESEARCH DESIGN AND METHODS abetes”). In this study, the participants deficits divided by the total number of Participants and Settings were included for analyses if they had deficits (n = 30). The FI ranged from 0 to 1, Details on CaMos have been published 1) a blood sample for fasting plasma with higher values indicating greater elsewhere (15) and can be found at www glucose measurement, 2) self-report frailty. However, the original FI in- .camos.org. Briefly, CaMos is a prospec- data on whether they had diabetes, cluded a deficit of diabetes that was tive cohort study of two-thirds women and 3) a follow-up of 1 year or more. coded as 1 (present) or 0 (absent). To and one-third men that aims to esti- Therefore, the year for participants’ co- align with the current study, we modified mate the incidence and prevalence hort entry may be different from the year the FI slightly by removing the deficit of fractures and declining bone mass in which the participants were enrolled of diabetes from the construction. The and the effect of osteoporosis in Cana- into the CaMos to ensure the availability FI therefore consisted of the remaining dians. The study enrolled 9,423 noninsti- of data on blood collections. Participants 29 deficits, ranging from 0 to 1, and with tutionalized participants aged $25 years who had no data on fasting plasma higher scores indicating greater frailty. in nine study centers across Canada (Van- glucose measures or had a follow-up of Supplementary Table 1 provides the def- couver, British Columbia; Calgary, Alberta; less than 1 year were excluded.