Geriatric Update Aug 11, 2025

The Alzheimer's Association established a new clinical practice guideline on Brain biomarker (BBM) tests and based on evidence, established 90% sensitivity (Sn) and 75% specificity (Sp) for triaging or 90% Sn and Sp for confirmatory testing. Across all 31 tests, pooled Sn ranged from 49.31% to 91.41, and Sp ranged from 61.54% to 96.72%. review of 1050 papers, 49 observational studies of 31 different BBM tests were selected for inclusion. Table S2 lists the 31 BBMs tested.

Those taking a multivitamin had higher global cognition scores than those who did not take it. Specific effects on global cognition, episodic memory, and executive function were noted. The improvement was most pronounced in those with a history of cardiovascular disease.

Dexmedetomidine IV given intra- or postoperatively has shown to reduce delirium in older adults by 1/3. A COCA call on medetomidine overdose in recreational drug users is becoming more common, addressed: what to look for in this sedating alpha 2 agonist, to use clonidine in withdrawal and provides free CME.

Drug classes associated with moderate certainty of reduced risk of dementia were anti-hypertensives, statins, sodium-glucose transport protein 2 (SGLT2) inhibitors, and glucagon-like peptide-1 receptor agonists (GLP-1 RAs), and of increased risk with anticholinergics, based on 68 meta-analyses, across 11 drug categories.

The endothelial glycocalyx is a sugar-rich layer lining the luminal surface of endothelial cells and in the brain’s blood vessels, which makes up the blood-brain barrier (BBB). They become patchy with age, leading to the age-related breakdown of the BBB. The levels of two key sugar-building enzymes, C1GALT1 and B3GNT3, were significantly lower in the brain’s blood vessels of old mice than in those of young mice and directly correlated with the reduced amount of the final mucin “sugar coat” that was present. Using adeno-associated viruses (AAVs) to overexpress these enzymes, restored core 1 mucin-type O-glycans of the glycocalyx to the brain endothelium and the integrity of the glycocalyx, improving the barrier’s function, reducing leaking, and reducing neuroinflammation and cognitive deficits in mice.

Cardiovascular risk factors, along with abdominal subcutaneous and visceral adipose tissue volume (obesity) had the greatest influence on lower grey matter volume on brain MRI, between 55–64 years of age in men, between 65–74 years of age in women. Based on 34,425 participants between the ages of 45 and 82 years, recruited from the UK Biobank database.

Semaglutide initiation was associated with reductions in weight and cardiovascular risk factors but increases in health care expenditures by $80 per member per month, excluding semaglutide costs over 24-months, in a Yale Health system study. The authors speculate higher health system engagement. It would be interesting to see if preventive services increased in that time frame and if over longer term savings would be seen, particularly in hospitalizations.

Semaglutide showed benefits in patients with peripheral arterial disease (PAD), improving maximum walking distance approximately 40 meters at week 52 over placebo (1·21 [IQR 0·95–1·55] vs 1·08 [0·86–1·36]; estimated treatment ratio 1·13 [95% CI 1·06–1·21]; p=0·0004). An improvement of 20 meters on a flat surface is considered clinically meaningful, and this result is double that on a 12% grade, likely to by the 4 kg weight loss in the GLP-1 group. 

GLP-1 receptor agonists may improve hidradenitis suppurativa (HS) disease severity and quality of life, 58.6% of participants saw improvement in their HS after initiating a GLP-1. DLQI scores (range 0-30) fell from 13.05 before treatment to 9.19 after treatment.

Statins have anti-inflammatory, neurogenesis, stress hormones, and relevant neurotransmitter effects, and some studies have shown benefit in depression. However, simvastatin did not exert additional antidepressive effects when added to escitalopram in patients with comorbid MDD and obesity, despite improving the cardiovascular risk profile, in a randomized clinical trial including 161 Germans with comorbid MDD and obesity, age 39 (18-65 yo).

Simvastatin was associated with the strongest correlation with hearing loss (OR = 1.56, P < .01) and tinnitus (OR = 1.50, P < .01) while fluvastatin was the least associated both hearing loss (OR = 1.15, P < .01) and tinnitus (OR = 1.02, P = .68). Atorvastatin, the most used statin fell in the middle (OR = 1.27, P < .01) and tinnitus (OR = 1.21, P < .01). Prevalence of hearing loss increases with age, >50% in those >65 and >80% in those >80 yo. I feel the association is due to age and I will not stop prescribing statins for fear of hearing loss.

Statins improved survival in patients with advanced nasopharyngeal cancer (NPC) who underwent standard concurrent chemoradiotherapy (CCRT) in a propensity score matched cohort of 202 statin users. All-cause mortality in the statin group vs. non-statin group was aHR 0.48 (95% CI, 0.34-0.68; P < .0001).

Rosuvastatin showed the strongest effect aHR of 0.21 (95% CI, 0.19-0.50), atorvastatin aHR of 0.39 (95% CI, 0.23-0.66). Higher cumulative defined daily doses correlated with better survival outcomes.

Supratherapeutic international normalized ratio (INR) while taking warfarin does not significantly increase the risk for brain bleeding after blunt head trauma in 2686 older adults, age 86 in warfarin and 80 in control group. Among patients not on blood thinners, about 6% had brain bleeds versus approximately 7% of patients on warfarin. For INR levels above and below 3.0, ICH rates were similar. Although the number of bleeds was very small, the results remained insignificant in subgroup analysis. This adds to the literature showing no increased risk with anticoagulation, but dementia increases bleeding risk (HR, 2.23; 95% CI, 1.08–4.61).

Presence of cardiovascular (CV) risk factors may cause the heart to “age” faster than a person’s chronological age, and social determinants of health may exacerbate this heart age discordance. For both men and women, a high proportion of people of Hispanic or Black race, low educational attainment or low income had a heart age to chronological age discordance of 10 years or more.

By extending the area head CT scanned by at least 6 cm below the lower end of the trachea to include the upper heart and aortic arch, the detection of cardiac clots increases sixfold. This adjustment could improve the identification of stroke causes and optimize treatment in 465 Canadian participants.

Alteplase given 4.5 to 24 hours after acute ischemic stroke onset improved functional independence at 90 days in 40% receiving alteplase vs 26% with standard care, in 372 patients, age 72, adjusted risk ratio, 1.52 [95% CI, 1.14-2.02]; P = .004). In subgroup analysis there was no difference in men and after 9 hours, but the confidence intervals were large and the numbers were small.

In participants with high-risk  obstructive sleep apnea (OSA): drops in blood oxygen levels >87.1% min/hour or large surges in heart rate, >9.4 beats/min, wearing continuous positive airway pressure (CPAP) showed reduced major adverse cardiovascular and cerebrovascular events (MACCE) (3rd tertile for each), only in those without excessive sleepiness (iHR .59, 95% CI .41–.84; high-risk n = 1509), or without increased blood pressure (iHR .54, 95% CI .36–.81; high-risk n = 1244). CPAP benefits were observed in high-risk OSA, in whom oxygen levels stayed low and pulse was high, alongside harm in low-risk OSA. It seems that if CPAP resolves sleepiness and hypertension, it will help with MACCE, because the adherence (≥4 h/night) was 37%-42%. A previous article, it was one of 3 included in the previous study, also showed benefit in patients with greater increase in heart rate while asleep.

The Patient Health Questionnaire (PHQ) is a widely used depression screen. This study examined the sensitivity (Sn) and specificity (Sp) and negative (NPV) and positive predictive values (PPV) of the various versions of PHQ-1 to 9. The PHQ-4 showed a greater increases in PPV than other intervals, from 0.58 of PHQ-3 to 0.78 of PHQ-4, with little difference between PHQ-2 of 0.56. NNPs were all >0.97. The 4 questions are: 1. Little interest or pleasure in doing things, 2. Feeling down, depressed or hopeless, 6. Feeling bad about yourself - or that you’re a failure or have let yourself or your family down, 7. Trouble concentrating on things, such as reading the newspaper or watching television. Table 3 in the free article lists the Sn, Sp, NPV and PPV of all PHQ tests.

Each 10 g/day increase in alcohol intake was associated with a 3% increase in pancreatic cancer risk. In women, compared to weak intake (0.1–5 g/day), alcohol intake of 15–30 g/day was associated with a 12% increased risk, while in men alcohol intake of 30–60 and more than 60 g/day were associated with a 15% and 36% increased risk, respectively.

Almost twice as many patients with symptomatic hand osteoarthritis (OA) who took two tablets daily containing a combination of patented plant extracts of Curcuma longa and Boswellia serrata, achieved a patient acceptable symptom status (PASS) (Odd ratio (OR) 1.9 [95% CI 1.0 to 3.8]). Improvement from baseline to month 3 on the 0-100 mm visual pain scale was −24.7 mm (95% Confidence Interval (CI) [−30.7 to −18.7]) vs. −16.2 mm (95% CI [−22.5 to −9.9]) in patients assigned to placebo.

Additional enteral protein 100 g protein/L vs. 63 g protein/L, did not increase days free of the index hospital and alive at day 90 compared with usual care, nor any secondary outcomes, in a cluster randomized, crossover, open-label trial of 3397 patients from 8 intensive care units.

Colorful vegetables high in flavanols are encouraged and participants in the highest quintile of food diversity were found to have a 14% lower risk of all cause mortality than participants in the lowest quintile (HR: 0.86; 95% CI: 0.78–0.95), but does not meet the HR of 2 for increased risk, or 0.5 for reduced risk that would be actionable for observational studies. In my experience, a small waistline and exercise are more important than diet. However, eating healthy adds to the benefit and is tasty.

Cocoa flavanol supplementation of 500 mg flavanols/day over 3.6 y, 410 participants taking cocoa extract and 456 taking placebo had 10% fewer confirmed total cardiovascular events (HR: 0.90; 95% CI: 0.78, 1.02; P = 0.11). For secondary endpoints, HRs were 0.73 (95% CI: 0.54, 0.98). The level of flavanols is affected by how the cocoa beans are processed, with traditional chocolate processing methods often reducing flavanol content, according to ConsumerLab.com, 1 tablespoon (5 grams) of some cocoa powders can contain between 37 and 130 milligrams of flavanols, some can provide up to 630 mg per tablespoon. 

Drinking instant coffee could possibly make a person’s risk of dry age-related macular degeneration seven times more likely, OR 6.92 (1.79, 35.15). No association was found with ground or decaf coffee, in the genome-wide association studies (GWAS) and the latest AMD-related GWAS summary data from the Finngen consortium R11 database analysis.

The evidence analysis tool (EAT) EAT-Lancet diet emphasizes fruits, vegetables, whole grains, nuts, legumes and fish, with 50% of calories from non-starchy vegetables, which can reduce chronic disease risk and support environmental health. Using models the impact of the EAT-Lancet diet index, metabolomic signatures and gene-environment interactions, showed reduced incidence on rheumatoid arthritis (RA), hazard ratios for the incidence of RA were 0.93 (95% CI: 0.90, 0.96) and 0.8 (95% CI: 0.70, 0.93) for corresponding metabolomic signatures.

Each 1-hour delay in the midpoint of meal intake was associated with a 0.952 increase in BMI at baseline (P = .0002), a 2.2% increase in long-term body weight (P = .042), and a 0.046 kg/wk slower weight-loss rate (P = .013). This held true for those with the highest genetic susceptibility to obesity, whereas no significant effect was observed in those with lower genetic susceptibility. I love to eat and have found it easier to keep my weight down by eating earlier in the evening and just enough to not be hungry anymore, and be asleep when I get hungry again.

Artificial intelligence large language models showed variable accuracy: 6 of 68 questions were incorrect in model 1 DeepSeek-R1 (8.82%), 11 of 68 (16.18%) in model 2 o3-mini, 23 of 68 (33.82%) in model 3 Claude-3.5 Sonnet, 25 of 68 (36.76%) in model 4 Gemini-2.0-Flash, 18 of 68 (26.47%) in model 5 GPT-4o, and 26 of 68 (38.24%) in model 6 Llama-3.3-70B. The authors lay out 3 priorities for medical artificial intelligence: (1) development of benchmarks that distinguish clinical reasoning from pattern matching, (2) greater transparency about current reasoning limitations in clinical contexts, and (3) research into models that prioritize reasoning over pattern recognition. Until these systems maintain performance with novel scenarios, clinical applications should be limited to nonautonomous supportive roles with human oversight.