Howard A. Fink

Prior studies evaluating International Classification of Diseases (ICD) codes for identifying medication-related osteonecrosis of the jaw (MRONJ) have reported poor validity but were largely limited to postmenopausal women, small samples, or patients receiving bisphosphonate or denosumab for cancer-related complications rather than fracture prevention. To evaluate the performance of ICD-9 and ICD-10 codes for identifying MRONJ among adults treated with bisphosphonates or denosumab for fracture prevention. We included two retrospective U.S. cohorts of adults age ≥50 years: the national Veterans Health Administration (VHA) and HealthPartners (HP), a large nonprofit health system. All patients had dental benefits and filled ≥1 prescription for a bisphosphonate or denosumab for fracture prevention from 10/1/1999-12/31/2022, with follow-up through 12/31/2023. In VHA, a previously validated natural language processing algorithm (F-Measure:0.96) ascertained MRONJ events. In HP, all electronic medical records associated with candidate ICD-9/10 claims underwent clinician review. MRONJ was defined per the American Association of Oral and Maxillofacial Surgeons criteria. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value were calculated for each code in VHA; PPV was calculated in HP. Among 46,408 VHA Veterans and 4,625 HP participants (64 undergoing review), 47 and 10 MRONJ cases were identified, respectively. The only code with PPV>0.40 in both systems was ICD-9733.45 (osteonecrosis of the jaw), with PPVs of 0.43 (95% confidence interval [CI]: 0.1.0–0.82) in VHA and 0.50 (95% CI: 0.07–0.93) in HP. The highest-performing ICD-10 code in both cohorts was M87.10 (osteonecrosis due to drug, unspecified bone), with PPVs of 0.27 (95% CI: 0.06–0.61) and 0.67 (95% CI: 0.09–0.99) in VHA and HP, respectively. Sensitivity ranged from 0.00 to 0.08. Available ICD-9/10 codes demonstrate low sensitivity and PPV for MRONJ, limiting accurate ascertainment in administrative data and highlighting the need for improved hybrid ascertainment algorithms or dedicated MRONJ registries for epidemiologic study of this rare outcome.

Some older adults cannot participate in community activities alone and their caregivers may be physically unable to assist them. In these circumstances, there is an opportunity to decrease caregiver workload through caregiver-targeted, mobility technology. We developed the Caregiver Assist Transport Chair (CATC), a motorised chair that responds to caregiver effort. To examine factors influencing design, operability, and perceived impact of the CATC. We employed a mixed methods study with older Veterans and their caregivers at the Minneapolis Veterans Affairs Health Care System. Testing sessions included a functional assessment (Short Physical Performance Battery), participant trials of the CATC prototype, semi-structured interviews, and completion of two questionnaires (Community Mobility Priorities and Life-Space Assessment). Data were analysed using rapid qualitative methods and descriptive statistics. Five caregivers and three Veterans participated. All participants noted the potential of the CATC to allow them to resume participation in meaningful activities they have otherwise stopped or reduced. Room for improvement included (i) ease in traversing thresholds or uneven terrain, (ii) reduced weight for loading and unloading into a vehicle, and (iii) diminished choppy or jerky feeling described by caregiver participants when operating the CATC. All participants spoke of the potential for the CATC to expand the number and types of places they visit. Design and operability features noted for improvement are addressable in the next stage of design. Caregiver-targeted mobile technologies-like the CATC-have the potential to ease caregiver physical burden and allow for better navigation of community settings.

Atypical femoral fractures (AFFs) are an uncommon complication of bisphosphonate (BP) treatment for osteoporosis. The purpose of this report was to determine differences in mortality among men treated with BPs who experienced AFFs, hip fractures, or typical subtrochanteric/femoral shaft (ST/FS) fractures. We included men from the Veterans Affairs Corporate Data Warehouse who filled at least one BP prescription over a nearly 20-yr span (October 1, 1999, to December 31, 2022). Typical ST/FS and hip fractures after a first filled prescription for a BP were identified using ICD codes. Radiographic review was conducted for ST/FS fractures to determine which were AFFs using the American Society of Bone and Mineral Research 2014 criterion. Demographic and clinical characteristics and cumulative duration of BP exposure prior to the incident fracture were determined. There were 23 men with an AFF, 4591 with a hip fracture, and 225 with a (non-AFF) ST/FS fracture. At 12-mo post-fracture, 4% of men with an AFF had died, compared with 25% of men with a hip fracture and 15% of men with an ST/FS fracture. Over 5 yrs. of follow-up, mortality following fracture differed among the 3 different fracture groups (χ2 = 6.21, d.f. = 2, p = .045) with death occurring in 7 of 23 (30%) men with an AFF, 2380 of 4591 (52%) men with a hip fracture and 109 of 225 (48%) men with a ST/FS fracture. In Cox proportional hazard models, over 5 yrs, with and without adjustment for covariates, there were no significant differences in mortality following an AFF compared with a hip fracture or an AFF compared with a ST/FS fracture (p > .10 for all). It is possible that mortality is lower following AFF compared with a hip or ST/FS fracture; however, limited sample size among those with AFF may have precluded ascertainment of such a difference.

The Osteoporotic Fractures in Men study (MrOS) is among the largest and longest running, prospective cohort studies of older men. MrOS determined how fracture risk was related to bone mass, bone geometry, lifestyle, anthropometric and neuromuscular measures, comorbidity, biomarkers, and fall propensity. The cohort consisted of 5994 community-dwelling, ambulatory US men aged 65 yr or older recruited at 6 US academic clinical centers in 2000-2002 and followed through November 2024. After enrollment, men were contacted by mail/phone every 4 mo to ascertain information on incident falls, fractures, and deaths/loss to follow-up. All fractures were confirmed by imaging reports. Over the 25-yr study, 95% of active surviving participants had complete follow-up. The MrOS study had 5 major clinic visits; while an ancillary sleep study had 2 additional clinic visits. MrOS has provided a comprehensive analysis of factors associated with areal (a) and volumetric (v) BMD and clinical risk factors for aBMD loss among men. Further analyses identified risk factors for hip, all non-spine, rib, wrist, and vertebral fractures. MrOS was the largest single cohort to estimate associations of aBMD with incident fractures in men and results indicated stronger associations in men than in women. MrOS also analyzed other structural features from DXA and both central and peripheral quantitative computed tomography in relationship to fractures in men. Serum, urine, and DNA were collected at clinic visits, and extensive analyses have been performed with respect to sex steroid and calciotropic hormones, bone turnover, and other novel measures of bone health. Several analyses evaluated the performance of formal tools in estimating the absolute risk of fractures in older men; however, findings indicated that better fracture prediction tools are needed. MrOS encourages outside investigators to make use of the publicly available data and to access the biospecimen bank at https://mrosonline.ucsf.edu.

Purpose To examine whether lower medication adherence in individuals without documented dementia or cognitive impairment (DCI) is associated with subsequent clinically recognized DCI. Methods Retrospective cohort analysis that included veterans aged ≥ 65 years, without documented DCI at baseline, without past or current prescription for DCI medications, with ≥ 1 annual Veterans Affairs (VA) primary care visit, and with an ongoing VA prescription for any of the one or more of the following during a 3-year assessment period: lisinopril, metoprolol, omeprazole, or simvastatin. Adherence was estimated for each medication using proportion of days covered (PDC) and, secondarily, using medication possession ratio (MPR), with < 0.8 classified as low adherence. Incident DCI was determined via ICD-9/ICD-10 diagnosis codes. Results Analyses included 794,569 unique veterans, separated in 4 medication cohorts (mean ages ranged from 75.6 [7.2] to 76.5 [7.2]), % male range was 98.2–98.6%). Between 16–21% of participants were categorized with low medication adherence by PDC. Over a 10-year follow-up period, incident DCI ranged from 17–18% in the low adherence group and 14–15% in the high adherence group. After multivariate adjustment, hazard for incident DCI in users of each of the four medications was significantly increased, about 20% greater among individuals with low adherence compared to those with high adherence. Conclusion Among veterans without clinically recognized DCI, lower baseline medication adherence was associated with an increased risk for future clinically recognized DCI. Further studies should seek to disentangle whether low adherence is a risk factor for future DCI or an indicator of existing but not yet recognized DCI and whether this can inform clinical practice decisions.

Background: Cognitive impairment is associated with higher total healthcare costs (THC) after accounting for demographic factors and comorbidity burden. Objective: Determine the proportion of incremental THC associated with cognitive impairment that is accounted for by functional impairments and frailty. Design: Four prospective cohort studies of community-dwelling older adults are linked with each other and Medicare claims. Participants: Eight thousand one hundred sixty-five community-dwelling, fee-for-service Medicare beneficiaries (mean age 79 years, 53% women, 79% non-Hispanic White). Main Measures: Cognitive impairment was defined by either abnormal cognitive tests or self-or-proxy report of a clinician diagnosis of dementia. Comorbidity was the count of claims-based chronic conditions. Functional impairments were self-reported difficulty performing four activities of daily living. Frailty was defined by the Cardiovascular Health Study phenotype or a claims-based deficit accumulation index (CFI). Annualized THC were ascertained for 36 months following an index exam. Incremental THC of cognitive impairment were mean THC in women (men) with cognitive impairment minus the mean THC in women (men) without cognitive impairment. Key Results: Incremental THC of cognitive impairment after adjustment for age, race, geographic region, and comorbidities was $6883 (95% CI, 3461–10,305) in women and $7276 (3298–11,254) in men. A substantial proportion of incremental THC of cognitive impairment were attributable to functional impairment, phenotypic frailty, or CFI individually (range 31.1–40.2% in women and 27.6–40.1% in men), while the combination of these three domains accounted for the majority of incremental THC (63.4% in women and 58.2% in men). Conclusions: For both women and men, after accounting for demographics and comorbidity, approximately 60% of incremental THC associated with cognitive impairment was attributable to functional impairments and frailty. Investigation is warranted to determine if addressing these geriatric syndromes will mitigate their impact on THC among community-dwelling older adults with cognitive impairment. Graphical

Individuals with dementia have a heightened hip fracture and fall risk but whether markers of brain injury are associated with hip fracture and falls is unknown. We tested the hypothesis that higher circulating brain injury markers were associated with increased risk of hip fracture and fall hospitalizations. Brain injury markers were measured in 2141 participants (mean age 77.9 yr; 60% women). Brain injury markers included neurofilament light chain (NfL), a marker of axonal injury; glial fibrillary acidic protein (GFAP), a marker of astrocytic injury; total Tau, whose many functions include neuron microtubule stabilization; and ubiquitin carboxy-terminal hydrolase L1 (UCH-L1), a major protein of neurons. Incident hip fractures and hospitalizations for falls were identified through participant report and confirmed with medical records or medicare claims. Hazard ratios were computed for a doubling exposure (log2 transformed brain injury marker) using multivariable-adjusted Cox models. After a median follow-up of 11 yr, 304 incident hip fractures and 284 incident fall hospitalizations occurred. Doubling of GFAP and NfL were associated with a 22% (p = .048) and 42% (p < .001) higher risk of hip fracture, respectively. Additional adjustment for cognitive function, gait speed, grip strength, inflammatory markers, and depressive symptoms had no effect on results. Models that adjusted for all 4 brain markers showed that only NfL was independent of the other markers. Neurofilament light chain was also associated with a 47% increase risk of hospitalization for falls. There was no association of total Tau or UCH-L1 with hip fracture or falls. GFAP was also unrelated to fall hospitalizations. Neurofilament light chain was independently associated with an incident risk of hip fracture and fall hospitalizations. These results suggest that subclinical degrees of brain injury may contribute to falls and hip fracture. Future research is needed to test whether the association between NfL and hip fracture is independent of falls.

Background and Objectives: While cross-sectional associations between any pain and sleep problems have been established, longitudinal studies examining the temporal relationship between back pain and multidimensional sleep health remain limited. We evaluated whether the association between back pain and sleep problems was bidirectional in older men aged 65 years and above. Research Design and Methods: Data came from the Osteoporotic Fractures in Men Study with a sample of 1,055 older men who completed 2 clinical sleep visits. A composite sleep problems score was created using self-report and actigraphy data reflecting irregularity, dissatisfaction, lack of daytime alertness, suboptimal timing, inefficiency, and suboptimal duration. Participants were queried by mail about back pain every 4 months, and we calculated the prevalence of any, frequent, severe, and activity-limiting back pain around their 2 sleep visits. Cross-lagged panel models estimated bidirectional associations between sleep problems and subsequent back pain, and vice versa, over 6 years. Results: Multivariable-adjusted results showed that having any back pain, frequent back pain, severe back pain, and activity-limiting back pain predicted 12%–25% greater sleep problems 6 years later (Exp(β) = 1.12; 95% confidence interval [CI] = 1.03–1.21 to Exp(β) = 1.25; 95% CI = 1.05–1.48), but sleep problems did not predict subsequent back pain. Discussion and Implications: This study highlights the long-term temporal directionality of the association between back pain and sleep problems in older men. Back pain preceded more sleep problems, but an inverse association was not observed. Our findings suggest that interventions targeting back pain may help decrease sleep problems in older men and warrant further investigation into potential mechanisms.

Background: Cognitive impairment and dementia are associated with higher healthcare costs; whether these increased costs are attributable to a greater comorbidity burden is unknown. We sought to determine associations of cognitive impairment and dementia with subsequent total and sector-specific healthcare costs after accounting for comorbidities and to compare costs by method of case ascertainment. Methods: Index examinations (2002–2011) of 4 prospective cohort studies linked with Medicare claims. 8 165 community-dwelling Medicare fee-for-service beneficiaries (4 318 women; 3 847 men). Cognitive impairment identified by self-or-proxy report of dementia and/or abnormal cognitive testing. Claims-based dementia and comorbidities derived from claims using Chronic Condition Warehouse algorithms. Annualized healthcare costs (2023 dollars) were ascertained for 36 months following index examinations. Results: 521 women (12.1%) and 418 men (10.9%) met the criteria for cognitive impairment; 388 women (9%) and 234 men (6.1%) met the criteria for claims-based dementia. After accounting for age, race, geographic region, and comorbidities, mean incremental costs of cognitive impairment versus no cognitive impairment in women (men) were $6 883 ($7 276) for total healthcare costs, $4 160 ($4 047) for inpatient costs, $1 206 ($1 587) for skilled nursing facility (SNF) costs, and $689 ($668) for home healthcare (HHC) costs. Mean adjusted incremental total and inpatient costs associated with claims-based dementia were smaller in magnitude and not statistically significant. Mean adjusted incremental costs of claims-based dementia versus no claims-based dementia in women (men) were $759 ($1 251) for SNF costs and $582 ($535) for HHC costs. Conclusions: Cognitive impairment is independently associated with substantial incremental total and sector-specific healthcare expenditures not fully captured by claims-based dementia or comorbidity burden.