Publications
Reverse-chronological list of peer-reviewed publications, preprints, and conference presentations. My name is for visibility.
Google Scholar · ORCID iD 0000-0001-5503-5305 · OpenAlex
660Citations
11h-index
13i10-index
26Works indexed
Bibliometrics sourced from OpenAlex, which counts citations only from works indexed in its corpus. These figures typically run lower than Google Scholar, which additionally captures citations from theses, conference proceedings, and non-indexed journals.
Peer-Reviewed Publications
2026
Erden, M., , Kho, R. M., & Oktay, K. H. (2026). Measuring what trials miss: decision modelling for value-based reproductive health. The Lancet Obstetrics, Gynaecology, & Women's Health, 2, e581–e582.
- Why we did it. Randomised trials are the backbone of reproductive medicine, but they tend to follow patients only briefly and stop at proximate endpoints such as live birth per cycle, even though many reproductive decisions carry consequences (ovarian ageing, cardiometabolic risk, treatment burden, quality of life) that unfold over years or decades. We argue that this short-horizon evidence leaves the outcomes that matter most to patients largely unmeasured.
- What we argue. Decision-analytic modelling and cost-effectiveness analysis can extend trials by linking their short-term results to long-term maternal health, quality of life, and costs, and we work through three cases where this changes the picture: frozen embryo transfer and later pre-eclampsia and cardiovascular risk, fertility preservation and the timing of premature menopause, and endometrioma surgery versus an in-vitro-fertilisation-first approach. A companion panel maps four model structures, the decision tree, Markov cohort model, dynamic compartmental model, and individual-based microsimulation, to the reproductive problems each is best suited to.
- Why it matters. If reproductive medicine is to align with value-based care, decision modelling should become a routine complement to clinical trials rather than an afterthought. That lets the field move beyond asking which intervention improves a single endpoint like live birth to asking which strategy best improves a patient's whole life course.
Jasani, S., , Stevenson, J., & Krawiec, C. (2026). A comparison of long-term maternal mortality associated with pathologic placental separation: Highlighting possible trends and mechanisms. PLOS ONE, 21(4), e0338586.
- Why we did it. Placental abruption (the placenta detaching before birth) and placental retention (the placenta failing to deliver within the normal window) are both known to cause dangerous bleeding and other acute harm around delivery, but their consequences for a mother's survival years later were less clear. These two conditions are usually studied on their own, so we grouped them together as forms of abnormal placental separation to ask whether either one raises long-term maternal death risk.
- What we found. In a retrospective cohort of 638,911 vaginal deliveries drawn from the TriNetX electronic health record network, mortality was 6.4 per 1,000 in normal deliveries versus 9.8 for abruption and 12.0 for retention, and after adjusting for age, race, and social factors both stayed elevated (retention HR 1.95, abruption HR 1.59). The two conditions followed different timelines: when we excluded deaths within 42 days of delivery, the retention risk held steady (HR 1.93) while the abruption risk faded to non-significance (HR 1.31), pointing to acute complications as the main driver of abruption deaths and a more persistent process behind retention deaths.
- Why it matters. The finding that retention keeps raising death risk years after delivery argues for looking past the traditional six-week postpartum window and building longer follow-up and preventive care for women who have had an abnormal placental separation. Because this is observational and relies on diagnostic codes, it cannot prove cause or pin down mechanisms, so the distinct risk patterns and health-condition clusters we describe are meant to guide, not settle, future research.
2024
Wahl, B., Butin, G., Gombe, S., , & Schwalbe, N. (2024). Beyond "business as usual": lessons from FIFA for fair benefit-sharing in global health. Health Affairs Scholar, 2(7), qxae068.
- Why we did it. Scientists and agencies in low- and middle-income countries supply much of the surveillance data and biospecimens that make pandemic prevention, preparedness, and response possible, yet they often receive neither fair compensation nor guaranteed access to the tools built from their contributions. Existing frameworks like the International Health Regulations, the Pandemic Influenza Preparedness Framework, and the Nagoya Protocol lack the enforcement and incentives to close this gap, so we looked beyond global health for a working model of fair reward.
- What we found. This commentary examines how FIFA, football's governing body, compensates the smaller clubs that develop players through training compensation and solidarity payments, routed through a centralized Clearing House that FIFA anticipates will pay out about $400 million a year across roughly 14,000 transactions. We map these mechanisms onto global health, showing how a trusted clearing house, a web-based platform that tracks who provides and uses data, and a dispute-resolution body could reward the countries and institutions that share pandemic-relevant data.
- Why it matters. The model is not directly transferable, since no health body holds FIFA's hegemonic control over its field, pandemic data and football players gain value on very different timelines, and even FIFA's scheme has been faulted for widening the gap between rich and poor clubs. Still, as countries negotiate a new pandemic accord and revise the International Health Regulations, the football analogy offers a concrete blueprint for making data-sharing incentives fairer instead of relying on goodwill alone.
2023
Copur, S., , Basile, C., & Kanbay, M. (2023). Endocrinological disorders in acute kidney injury: An often overlooked field of clinical research. Journal of Nephrology, 36(1), 49–63.
- Why we did it. Acute kidney injury affects about one in five hospitalized adults, and the kidney is where most hormones are made, metabolized, or cleared, yet the hormonal fallout of AKI has drawn far less attention than its cardiovascular, electrolyte, and neurological effects. We wrote this narrative review to pull together what is known, and what is still unknown, about how AKI disturbs the endocrine system.
- What we found. Across the published evidence, AKI commonly comes with thyroid abnormalities (mostly low-T3 euthyroid sick syndrome, meaning altered thyroid blood tests without true thyroid disease), swings in blood sugar and insulin resistance, disordered bone-mineral hormones such as FGF-23 and PTH, and shifts in erythropoietin and natriuretic peptide levels. Trials of giving erythropoietin, thyroxine, or natriuretic peptides to prevent or treat AKI have produced mixed and mostly unconvincing results, and the effects of AKI on some core hormones like cortisol and insulin have essentially never been studied.
- Why it matters. These hormonal changes are not just laboratory curiosities: markers like FGF-23 rise early and track with worse outcomes, and glucose targets need adjusting in AKI patients to avoid dangerous hypoglycemia. The honest takeaway is that this area is thinly studied, and large randomized trials are needed to learn whether the endocrine disturbances persist after recovery and whether correcting them changes patient outcomes.
Copur, S., , Cherney, D., Tuttle, K., & Kanbay, M. (2023). Tirzepatide decreases systolic and diastolic blood pressure. European Journal of Internal Medicine, 114, 135–137.
- Why we did it. Tirzepatide, a dual GIP and GLP-1 receptor drug for type 2 diabetes, was already known to lower blood sugar and body weight, with early signals of heart and kidney benefit. Its effect on blood pressure was less clear, because most trials reported it only as a secondary measure and leaned on single office readings rather than 24-hour ambulatory monitoring.
- What we found. In this letter to the editor, we gathered blood pressure data from the main tirzepatide trials, the SURPASS and SURMOUNT programs, and found that most showed meaningful drops in both systolic and diastolic pressure, with systolic declines running from roughly 3 to 13 mmHg versus placebo or comparator drugs and usually a small 1 to 2 beat-per-minute rise in heart rate. One phase 2 trial was the exception, showing no significant difference between groups.
- Why it matters. The blood pressure benefit strengthens the case that tirzepatide may protect the heart and kidneys beyond glucose control, but because the trials leaned on single office readings and short follow-up, and the drop could in theory worsen orthostatic hypotension in patients with diabetic nerve damage, longer studies with 24-hour monitoring and dedicated cardiovascular and renal endpoints are needed.
Zavalichi, M. A., Ionescu, G., Arsenescu Georgescu, C. M., Mihaescu, A., Cimpoesu, C. D., Cimpoesu, G., Zavalichi, S. D., Statescu, C., , Kanbay, M., Covic, A., & Nistor, I. (2023). The use of extracorporeal membrane oxygenation in COVID-19: A systematic review. Archives of Medical Science.
- Why we did it. Severe COVID-19 can push the lungs and heart into failure that a ventilator alone cannot fix, and extracorporeal membrane oxygenation (ECMO), a machine that oxygenates the blood outside the body, was offered as a last-resort rescue. Early reports on how well it worked in COVID-19 patients were scattered and inconsistent, so we set out to pull them together and assess ECMO's real impact.
- What we found. We reviewed 33 studies from 10 countries covering 4,760 COVID-19 patients treated with ECMO, and survival to hospital discharge ranged widely, from 9% to 90.6%. Serious complications were common, including acute kidney injury (up to 87% of patients), major bleeding (up to 92.1%), and stroke or cerebral hemorrhage (up to 34%), while events such as pulmonary embolism, peripheral bleeding, and sepsis further worsened survival.
- Why it matters. ECMO can be a useful rescue option for the sickest COVID-19 patients, but the wide swing in outcomes across centers and the heavy burden of complications mean it is no guaranteed save. Better patient selection and larger, more consistent studies are needed before its role can be pinned down.
2022
Afsar, B., Afsar, R. E., , Altay, S., Korkmaz, H., Yildiz, A., Covic, A., Ortiz, A., & Kanbay, M. (2022). Sodium–glucose cotransporter inhibition in polycystic kidney disease: Fact or fiction. Clinical Kidney Journal, 15(7), 1275–1283.
- Why we did it. Autosomal dominant polycystic kidney disease (ADPKD), the most common inherited kidney disease, still has few good treatments, and its underlying biology (disordered metabolism, impaired cellular housekeeping, mitochondrial dysfunction, and inflammation) overlaps with the pathways that SGLT2 inhibitors act on. These drugs are now proven to protect the heart and kidneys, but people with ADPKD were left out of the trials that showed kidney benefit, so we asked whether they might help here too.
- What we found. Searching MEDLINE, Cochrane, Embase, and PubMed, we found no clinical studies in ADPKD and only a handful of animal studies with conflicting results: SGLT inhibition was protective in one rat model of proximal-tubule cysts but failed to slow, and in one case worsened, cyst growth in a recessive-PKD rat model and an accelerated Pkd1 mouse model. None of these animal models faithfully reproduces human ADPKD, which limits how far the results can be trusted.
- Why it matters. Whether SGLT2 inhibitors preserve kidney function in ADPKD is genuinely unknown, and there are real safety questions, including the risk of cyst infection and a possible rise in vasopressin, especially alongside tolvaptan. Since these patients may already receive the drugs for heart conditions, we lay out a roadmap, including better animal models, a patient registry, and re-analysis of existing cardiovascular trials, to settle the safety and benefit questions.
Copur, S., , & Kanbay, M. (2022). Uric acid in metabolic syndrome: Does uric acid have a definitive role? European Journal of Internal Medicine, 103, 4–12.
- Why we did it. High serum uric acid, the end product of purine breakdown, shows up alongside metabolic syndrome and tracks with hypertension, fatty liver, kidney disease, and cardiovascular disease, but whether it actually causes these problems or is just a bystander marker has stayed unsettled. We wrote this review to pull together what is known about how uric acid acts inside cells and across organ systems and to ask whether it plays a definitive role.
- What we found. We describe uric acid as a double-edged molecule: it can mop up oxidative stress outside cells yet drive it inside them, and through mechanisms touching inflammation, insulin resistance, atrial fibrillation, and cell death it can both trigger metabolic damage and result from it. We also gather newer evidence linking uric acid to the gut microbiome and to vitamin D, and note that studies disagree on where to set a healthy target, with some finding a U-shaped risk where both high and very low levels (below about 2 mg/dL) carry harm.
- Why it matters. Because the cause-and-effect direction is still a chicken-and-egg question and no consensus target exists, uric acid is not yet ready to guide treatment of asymptomatic patients, though lowering it may help prevent several metabolic and cardiovascular conditions. We argue that large, well-designed randomized trials are needed before uric acid can move from a risk marker to a practical tool for personalized care.
, Afsar, B., Covic, A., Kuwabara, M., Ferro, C. J., Lanaspa, M. A., Johnson, R. J., & Kanbay, M. (2022). The role of uric acid in acute myocardial infarction: A narrative review. Angiology, 73(1), 9–17.
- Why we did it. High blood levels of uric acid, the final product of purine breakdown, had long been tied to heart disease, but whether it independently drives acute myocardial infarction (a heart attack) or is just a bystander that travels with other risk factors stayed unsettled, with studies pointing both ways. We set out to pull together the recent evidence on uric acid and heart attack, along with the biology that might connect them.
- What we found. Across the studies we reviewed, higher uric acid tended to track with worse outcomes after a heart attack, including larger infarct size, heart failure, and higher short- and long-term mortality, and one meta-analysis reported a 2- to 4-fold higher death rate in patients in the top uric acid quartile compared with lower ones. The plausible mechanisms are oxidative stress, inflammation, and endothelial dysfunction (damage to the blood-vessel lining), though uric acid also has antioxidant effects, a two-sided "Janus" behavior that may explain why some studies found no link.
- Why it matters. Uric acid may be a useful, cheap marker for stratifying risk after a heart attack, and some uric-acid-lowering drugs such as allopurinol are associated with fewer cardiovascular events, but the evidence is mixed and no guideline yet recommends treating high uric acid to protect the heart. Randomized trials are still needed to show whether lowering uric acid actually improves outcomes and to define the target level, since very low levels may carry their own harms.
, Kanbay, A., & Kanbay, M. (2022). Long-term effect of COVID-19 infection on hemodialysis patients: Should we follow hemodialysis patients more closely? Clinical Kidney Journal, 15(3), 369–371.
- Why we did it. Hemodialysis patients, whose kidneys have failed and who rely on a machine to filter their blood several times a week, were known to die at high rates from COVID-19 during their first hospital stay. What no one had shown was whether that danger ended at discharge or kept going, since almost all earlier reports stopped counting at one month.
- What we found. In this editorial we discuss a prospective, observational study by Carriazo and colleagues that followed 56 hemodialysis patients with COVID-19 for a full year and found that 20 of them (35.7%) died, only 6 during the initial admission, a threefold higher risk of death than dialysis patients who never had COVID-19 (hazard ratio 3.00). We also flag their finding that protective anti-SARS-CoV-2 antibodies faded fast, present in 73.4% of tested patients early on but only 38.8% at 12 months.
- Why it matters. The excess deaths were concentrated in the first three months after diagnosis, so we argue that dialysis patients who recover from COVID-19 should be watched closely well beyond hospital discharge rather than considered out of danger. Because their antibodies drop so quickly, we also make the case for prioritizing these patients for vaccination and tracking their antibody levels to catch waning protection and possible reinfection.
, & Kanbay, M. (2022). The assessment of hypertension in kidney transplant patients: Time to change our approach? Clinical Kidney Journal, 15(1), 1–4.
- Why we did it. Hypertension is the most frequently encountered comorbidity after kidney transplantation, and how well it is controlled is strongly tied to how long both the patient and the transplanted kidney survive. Yet current transplant guidelines still rely on office blood pressure readings, which miss a lot, leaving open the question of whether a fuller 24-hour picture should become the standard.
- What we found. In this editorial we discuss a meta-analysis by Pisano et al. of 42 studies and 4115 transplant patients, which found that office readings and 24-hour ambulatory blood pressure monitoring (a wearable cuff that records pressure every 10 to 30 minutes, day and night) disagreed on whether a patient's hypertension was controlled 44% of the time. The ambulatory approach also exposed patterns office visits cannot catch, including a 54% average rate of non-dipping (blood pressure that fails to fall normally overnight), masked hypertension in 26%, and white-coat hypertension in 10%.
- Why it matters. Because these hidden patterns carry real cardiovascular and graft risk, especially soon after transplant, we argue that ambulatory monitoring should be the first-line tool for diagnosing and managing hypertension in transplant patients and that guidelines should be updated to reflect that. It also allows chronotherapy, timing medications to a patient's own daily rhythm, though head-to-head data comparing ambulatory with cheaper home monitoring is still thin and worth pursuing.
, Öztosun, Ç., Acar, S., Sakarya, S., Gürsoy, T., & Ata, B. (2022). Transition from locally developed, faculty-written examinations to the National Board of Medical Examiners (NBME) subject examinations in clinical medical education. SN Social Sciences, 2(10), 200.
- Why we did it. Many medical schools outside North America write their own multiple-choice exams, but with cohorts of 20 to 25 students it is hard to show those exams are reliable and valid. When Koç University School of Medicine began replacing its faculty-written clerkship exams with standardized NBME subject examinations, third-party tests produced by the US National Board of Medical Examiners, no one knew whether the two would grade students consistently.
- What we found. We compared 286 paired scores from ten exams where students sat both tests and found fair to moderate agreement (kappa 0.41, a chance-corrected measure of consistency), with about 88 percent concordance on pass or fail decisions. Students preferred the NBME exam in every clerkship except internal medicine, mainly for its clearer language, and areas where our students lagged other NBME test takers exposed real curriculum gaps, such as thin coverage of adolescent care.
- Why it matters. For English-taught medical schools in non-English speaking countries, third-party exams can replace in-house tests with acceptable grading agreement while saving faculty time and providing external benchmarks against other schools. Discrepancies between exam content and the local curriculum are best read as a prompt to improve the curriculum rather than a reason to avoid the switch.
Ertuglu, L. A., , Afsar, B., Ortiz, A., & Kanbay, M. (2022). The use of healthy eating index 2015 and healthy beverage index for predicting and modifying cardiovascular and renal outcomes. Current Nutrition Reports, 11, 526–535.
- Why we did it. Nutrition research has moved from studying single nutrients to scoring overall dietary patterns, yet the newest Healthy Eating Index, HEI-2015, and the recently introduced Healthy Beverage Index (HBI), which scores the quality of what people drink, had not yet been assessed for how well they track heart and kidney outcomes. We set out to gather the existing evidence on both indices and guide their use.
- What we found. In this narrative review, higher HEI scores were consistently tied to lower cardiometabolic risk, less inflammation, and lower rates of cardiovascular disease, diabetes, chronic kidney disease, and all-cause death; in the ARIC cohort the top fifth of HEI-2015 had a 16 percent lower risk of new cardiovascular disease and a 17 percent lower risk of new kidney disease. On the beverage side, better HBI scores were linked to lower hyperlipidemia and hypertension risk, a closely related beverage score predicted lower all-cause mortality in people with kidney disease, and modeling suggested that swapping one daily sugary drink for water could raise HBI scores by 9 to 21 percent and predict modest weight loss of about 0.4 to 2 kg.
- Why it matters. Because these indices capture overall diet and drink quality rather than isolated foods, their components give clinicians a practical way to counsel patients, for example eat more vegetables, fruit, and whole grains, cut added sugar and saturated fat, and replace sweet drinks with water. Their value in people with advanced kidney disease and their link to newer biomarkers still need confirmation, and the field lacks one simple, agreed-upon index that is easy to apply in routine care.
Kanbay, M., Copur, S., , Sag, A. A., Covic, A., Ortiz, A., & Tuttle, K. R. (2022). Fatty kidney: A possible future for chronic kidney disease research. European Journal of Clinical Investigation, 52(6), e13748.
- Why we did it. Obesity and metabolic syndrome keep getting more common, and fat can build up not just in the liver but inside the kidney and in the tissue around it, a condition called fatty kidney. People with metabolic syndrome develop chronic kidney disease more often, but whether the fat in the kidney actually drives the damage, and how, was still an open question, so we set out to review the current evidence.
- What we found. In this narrative review of preclinical and clinical work, we identified oxidative stress and inflammation, Klotho deficiency, endoplasmic reticulum stress, mitochondrial dysfunction, and disrupted cellular energy balance as the main routes by which accumulated fat appears to injure kidney tissue. We also noted that the Framingham Heart Study found imaging consistent with fatty kidney in about 30 percent of subjects, and that current clinical studies point to fatty kidney as an independent risk factor for chronic kidney disease and cardiovascular events, though large-scale confirmatory studies are still missing.
- Why it matters. If fat in the kidney is a modifiable driver of disease, it opens a new target for prevention and treatment through lifestyle change and drugs such as PPAR-alpha agonists, statins, GLP-1 receptor agonists, and SGLT-2 inhibitors, though no agreed treatment algorithm yet exists. Clearer imaging standards, ideally MRI, and dedicated studies of the mechanisms, outcomes, and therapies are needed before fatty kidney can guide everyday clinical care.
Kanbay, M., Copur, S., , & Tuttle, K. R. (2022). Cardiorenal metabolic consequences of nighttime snacking: Is it an innocent eating behavior? Current Nutrition Reports, 11(2), 347–353.
- Why we did it. Metabolic syndrome and obesity keep climbing, and eating late at night, whether as casual snacking or as night eating syndrome (daytime loss of appetite, evening overeating, and trouble sleeping), has been raised as a possible contributor. But nighttime eating has never been tightly defined, and most of the evidence comes from cross-sectional snapshots rather than long-term or interventional work, so we set out to gather what is actually known.
- What we found. This review pulls together the literature linking nighttime eating to a broad set of harms: hyperlipidemia, high triglycerides, elevated blood sugar, weight gain, obesity, metabolic syndrome, higher blood pressure, atherosclerosis, and kidney signals such as declining eGFR and protein in the urine. We describe how these effects may trace back to a disrupted circadian clock, shifted levels of hormones like leptin, ghrelin, melatonin, and cortisol, and impaired cellular repair under a high-glucose state, while noting that in one study nighttime eaters also simply took in more total daily calories (about 4,758 versus 4,244).
- Why it matters. Because the timing of food, not just its content, may carry its own cardiorenal cost, clinicians have reason to ask patients when they eat and not only what. The evidence base is still thin, so we argue for longer-term interventional studies that separate the effect of meal timing from nutritional content, and we point to candidate treatments for night eating syndrome such as SSRIs, melatonin, and cognitive behavioral therapy.
Kanbay, M., Tapoi, L., Ureche, C., Tanriover, C., Cevik, E., , Afsar, B., Cherney, D. Z. I., & Covic, A. (2022). Effect of sodium–glucose cotransporter 2 inhibitors on hemoglobin and hematocrit levels in type 2 diabetes: A systematic review and meta-analysis. International Urology and Nephrology, 54(4), 827–841.
- Why we did it. SGLT2 inhibitors, a class of glucose-lowering drugs, improve heart and kidney outcomes in type 2 diabetes, and several trials had noticed that patients taking them showed higher hemoglobin and hematocrit, the standard measures of red blood cell mass. We set out to pool this scattered evidence and quantify how much these drugs raise red blood cell parameters, since anemia is common in diabetes and kidney disease and worsens prognosis.
- What we found. We combined seventeen randomized, double-blind, placebo-controlled trials covering 14,748 patients on canagliflozin, dapagliflozin, empagliflozin, or ipragliflozin. Compared with placebo, SGLT2 inhibitors significantly raised hemoglobin (mean difference 5.60 g/L) and hematocrit (mean difference 1.32 percent), and the increase held across the individual drugs studied, though the trials varied widely among themselves.
- Why it matters. The rise in red blood cell measures is a consistent class effect that may help correct the anemia seen in diabetes and kidney disease and could contribute to the drugs' known heart and kidney benefits, possibly through better tissue oxygen delivery. Because plasma volume contraction, restored erythropoietin production, and anti-inflammatory effects likely all play a part, more work is needed to sort out which mechanism matters most and whether it differs between diabetic and non-diabetic patients.
2021
Afsar, B., Afsar, R. E., , Covic, A., & Kanbay, M. (2021). Deciphering nutritional interventions for podocyte structure and function. Pharmacological Research, 172, 105852.
- Why we did it. Chronic kidney disease often progresses through proteinuria, which damages the glomerular filtration barrier and its specialized cells, the podocytes. The standard drugs for these conditions, steroids and calcineurin inhibitors, act non-specifically and carry heavy body-wide side effects, so as interest in "food as medicine" has grown we gathered what is known about whether specific dietary compounds can protect podocyte structure and function.
- What we found. This narrative review of laboratory and animal studies found that a wide range of nutritional agents, including retinoic acid, grape seed procyanidin, curcumin, epicatechin, ferulic acid, resveratrol, and taurine, improved podocyte health across models. Across these studies the compounds raised protective slit diaphragm proteins such as nephrin, podocin, and CD2AP, stabilized the podocyte actin skeleton, and reduced harmful processes including the shift of podocytes toward a scar-forming state, apoptosis, and endoplasmic reticulum stress.
- Why it matters. Diet-based compounds could one day offer a gentler way to treat podocyte injury, but the evidence remains preclinical, so safe human doses, toxicity to other organs, interactions with existing drugs, and effects on hard outcomes like kidney function decline are all still unknown, and clinical trials are needed before any of these interventions can be recommended in practice.
Dogan, O., Vatansever, C., Atac, N., Albayrak, O., Karahuseyinoglu, S., Sahin, O. E., Kilicoglu, B. K., , Ergonul, O., Gönen, M., & Can, F. (2021). Virulence determinants of colistin-resistant K. pneumoniae high-risk clones. Biology, 10(5), 436.
- Why we did it. Colistin is a last-resort antibiotic, and colistin-resistant Klebsiella pneumoniae kills about half of the patients it infects, yet we understood little about why certain "high-risk" strains spread so aggressively or how they slip past the immune system. We set out to identify the main virulence traits of these strains and to watch how they interact with neutrophils, the white blood cells that are a first line of defense.
- What we found. Across 142 patients with invasive colistin-resistant infection, the ST101 and ST395 strains carried iron-scavenging genes (the yersiniabactin gene ybtS and the ferric uptake gene kfu) far more often than other strains (99% and 94%), and these strains were tied to higher 30-day death rates (58% and 75%); carrying kfu, being in intensive care, and belonging to the ST101 clone each independently predicted death. In the lab, iron-gene-positive bacteria were taken up by neutrophils more often (78% vs. 65%) but survived that engulfment better and triggered heavy release of neutrophil extracellular traps, the sticky DNA webs neutrophils cast to catch microbes.
- Why it matters. Iron uptake systems appear to help these high-risk clones outlast the neutrophil defense and drive damaging inflammation, which helps explain their high fatality. That makes drugs aimed at bacterial iron uptake a promising direction for treating colistin-resistant K. pneumoniae, a possibility the authors note should next be tested in animal studies.
Ertuglu, L. A., Afsar, B., Yildiz, A. B., , Ortiz, A., Covic, A., & Kanbay, M. (2021). Substitution of sugar-sweetened beverages for other beverages: Can it be the next step towards healthy aging? Current Nutrition Reports, 10(4), 399–412.
- Why we did it. As life expectancy rises, the gap between how long people live and how long they stay disease-free has widened, and diet has emerged as a strong lever on that "healthspan." Sugar-sweetened beverages are a major source of added sugar in the American diet, about a quarter of all added sugar intake, and are tied to chronic disease, yet how they act on the biology of aging itself, and whether simply swapping them out helps, had not been pulled together.
- What we found. In this review we trace how sugar-sweetened beverages appear to accelerate aging through several linked pathways: chronic low-grade inflammation (inflammaging), oxidative stress and the buildup of advanced glycation end products, shifts in gut bacteria toward a profile resembling the aged gut, and the extra calories that drive weight gain and metabolic syndrome. Drawing on large cohort and trial data, we find that replacing these drinks with water, coffee, tea, or milk is tied to lower diabetes risk and weight loss, while artificially sweetened substitutes carry their own signals of harm such as altered microbiota and links to stroke and dementia, making water the preferred swap.
- Why it matters. Because the biological processes that these drinks disturb are the same ones that push aging forward, cutting sugar-sweetened beverages, ideally by switching to water, is a simple and low-cost step that could ease the chronic-disease burden in older adults and possibly extend healthy years of life. The review also points to future studies testing whether beverage substitution actually slows the biological rate of aging, and to policy tools such as sugar taxes and front-of-package warnings.
Ertuglu, L. A., , Basile, C., Afsar, B., Covic, A., & Kanbay, M. (2021). Sodium and ultrafiltration profiling in hemodialysis: A long-forgotten issue revisited. Hemodialysis International, 25(4), 433–446.
- Why we did it. Intradialytic hypotension, a sudden blood pressure drop during dialysis, is a common complication with a prevalence as high as 40 percent and is tied to worse cardiovascular outcomes and death. Sodium and ultrafiltration profiling, older techniques that vary the dialysate salt concentration and the fluid removal rate across a session to steady the patient, were largely set aside over worries about salt loading, thirst, and fluid gain between sessions, so we revisited what the evidence actually shows.
- What we found. Drawing on small crossover trials, observational studies, and large cohort data, we found that sodium profiling reliably reduces hypotension and muscle cramps, with the stepwise pattern the most effective, but that salt-positive versions raise interdialytic weight gain and thirst, and one large DOPPS cohort linked routine profiling to higher cardiovascular and all-cause mortality. Those adverse effects largely disappear when the profile is kept sodium-balance neutral or negative, meaning no net salt is added over the session, and automated biofeedback systems can hold that balance while preserving the hemodynamic benefit.
- Why it matters. Profiling can make dialysis sessions safer and more comfortable, but only if clinicians avoid a net salt gain, since positive sodium balance appears to drive the fluid overload and the mortality signal. Large long-term trials are still needed to settle the best method and to test whether ultrafiltration profiling adds any value on its own.
Ertuglu, L. A., Porrini, E., Hornum, M., , Afsar, B., Ortiz, A., Covic, A., Rossing, P., & Kanbay, M. (2021). Glucagon-like peptide-1 receptor agonists and sodium-glucose cotransporter 2 inhibitors for diabetes after solid organ transplantation. Transplant International, 34(8), 1341–1359.
- Why we did it. Diabetes that appears after an organ transplant, or that was present beforehand, is a major cause of illness and death in transplant recipients, yet insulin is the standard choice only for the first weeks after surgery and no drug is agreed on for long-term glucose control. We set out to weigh two newer diabetes drug classes, GLP-1 receptor agonists and SGLT2 inhibitors, that already protect the heart and kidneys in the general diabetic population but have barely been studied after transplantation.
- What we found. Pulling together the small retrospective and exploratory studies available, we found that GLP-1 receptor agonists improved blood sugar and produced meaningful weight loss without changing tacrolimus levels or kidney function, and preclinical work suggests their mechanism may even counteract the beta-cell toxicity caused by anti-rejection drugs. SGLT2 inhibitors showed similar metabolic benefits but lose potency as kidney function falls and carry a real risk of urinary and genital infections, with dropout in both classes driven by side effects rather than lack of efficacy.
- Why it matters. Both drug classes look safe and useful enough to consider for transplant patients with diabetes, and we offer a practical algorithm for choosing between them alongside a roadmap for the randomized trials that are still missing. Because almost all the current evidence comes from observational data, we argue that kidney transplant recipients, who face the highest cardiovascular and kidney risk, are the population where proper trials should start.
Kanbay, M., , Afsar, B., Covic, A., Tapoi, L., Ureche, C., & Ortiz, A. (2021). Role of klotho in the development of essential hypertension. Hypertension, 77(3), 740–750.
- Why we did it. Essential hypertension accounts for almost 95 percent of adult cases, yet it has no single identified cause and its mechanisms remain poorly understood after decades of study. Klotho, an antiaging protein whose blood levels fall with normal aging and with diseases like hypertension and kidney disease, kept surfacing in these conditions, so we set out to critically review whether its decline actually helps drive high blood pressure.
- What we found. Drawing on animal experiments, genetic studies, and human observations, we describe how low Klotho promotes hypertension through several converging routes: greater salt sensitivity, stiffening and aging of the arteries, endothelial dysfunction (impaired function of the blood vessel lining), excess aldosterone, and disturbed phosphate handling via the FGF23 pathway. In mice, Klotho supplementation prevented or reversed the rise in blood pressure, which points to a possible causal role rather than a bystander association.
- Why it matters. If Klotho deficiency truly sits upstream of essential hypertension, it could serve both as an early biomarker of subclinical kidney and vascular injury and as a treatment target, moving beyond drugs that only lower pressure toward ones that address a root cause. These links still rest largely on animal and small human data, so we outline the large clinical studies needed before Klotho-based diagnosis or therapy could reach patients.
Kanbay, M., , Afsar, B., Karakus, K. E., Ortiz, A., Hornum, M., Covic, A., Sarafidis, P., & Rossing, P. (2021). Sodium-glucose cotransporter 2 inhibitors for diabetes mellitus control after kidney transplantation: Review of the current evidence. Nephrology (Carlton), 26(12), 1007–1017.
- Why we did it. Diabetes is both a leading cause of kidney failure and a common complication after a kidney transplant, yet clinicians have been cautious about using SGLT2 inhibitors (a newer class of glucose-lowering drugs that also protect the heart and kidneys in the general population) in transplant recipients because of possible interactions with anti-rejection drugs and worries about infection. We reviewed the current evidence to see whether these drugs are safe and useful for controlling diabetes after a kidney transplant.
- What we found. Across 11 studies covering 214 diabetic transplant recipients, SGLT2 inhibitors lowered HbA1c (a marker of average blood sugar) by roughly 0.6 percentage points and cut body weight by about 2.4 kg, similar to what is seen in non-transplant patients. Kidney filtration dipped in the short term but returned to baseline by 3 to 12 months, and the main safety signals were 23 urinary tract infections, 2 genital yeast infections, 1 acute kidney injury, and 1 mild low-blood-sugar episode, with no cases of ketoacidosis or graft rejection.
- Why it matters. The early evidence suggests these drugs are safe and effective for glucose and weight control in transplant recipients, but the larger heart and kidney benefits proven in other patients, along with the blood pressure and protein-in-urine effects, still need to be confirmed by bigger, longer trials before this becomes standard practice.
Preprints & Under Review
, Skolnick, S., Cheng, C.-L., Jones, M., & Tam, J. (2026). Maternal morbidity, infant mortality, and health-care costs attributable to prenatal smoking in the USA during the e-cigarette era: A simulation modelling study. SSRN.
- Why we did it. Smoking during pregnancy is a known, preventable cause of miscarriage, placental complications, hypertensive disorders, and infant death, yet the most influential US burden estimates come from an earlier era, before prenatal smoking declined and e-cigarettes became common. No existing US model had combined cigarette and e-cigarette transitions, quitting during pregnancy, and postpartum relapse to project how much harm still lies ahead.
- What we found. We built a microsimulation model, a computer simulation that follows individual women year by year, calibrated to national survey data, and projected that smoking among pregnant women falls to about 2.9 percent by 2040 while e-cigarette use rises to 8.5 percent. Even so, prenatal smoking from 2027 to 2100 accounts for 1.62 million miscarriages, 414,000 hypertensive disorders of pregnancy, 87,000 infant deaths, $7.7 billion in direct medical costs, and 62,000 quality-adjusted life-years lost among mothers.
- Why it matters. Falling prevalence does not close the problem: roughly 1,170 excess infant deaths per year remain preventable, so tobacco control during pregnancy and the postpartum period still matters. The model also gives policymakers a benchmark that reflects today's nicotine-use patterns, including vaping, for evaluating future interventions.
, Skolnick, S., & Tam, J. (2025). Maternal and infant health benefits of a nicotine product standard in the United States. medRxiv preprint, 2025.09.11.25335605.
- Why we did it. Smoking during pregnancy is a well-established driver of miscarriage, ectopic pregnancy, placental complications, hypertensive disorders, and infant death, yet about 44% of people who smoked before pregnancy keep smoking through it. The FDA has proposed capping nicotine in cigarettes at minimally addictive levels, but no one had projected what such a rule would mean specifically for maternal and infant health.
- What we found. We built the Smoking, E-cigarette use, and Pregnancy (SEP) microsimulation, an individual-level model of U.S. women of reproductive age, and ran a nicotine product standard starting in 2027 against a status-quo scenario through 2100. Smoking in pregnancy fell from 6.0% in 2027 to 1.2% by 2040, and cumulatively the policy averted roughly 950,000 miscarriages, 167,000 ectopic pregnancies, 200,000 hypertensive disorders, and 64,000 infant deaths, while gaining about 103,000 maternal quality-adjusted life years and avoiding around $4.9 billion in pregnancy-related medical costs.
- Why it matters. This is the first study to quantify a nicotine standard's effects on pregnancy, and it shows durable perinatal gains and cost savings even under conservative assumptions, strengthening the case for timely implementation. The benefits could be amplified by pairing the rule with prenatal cessation counseling and postpartum relapse prevention.
Conference Presentations
, Skolnick, S., Cheng, C.-L., & Tam, J. (2026). Projecting the maternal and infant health impact of a federal nicotine product standard. Poster, Society for Medical Decision Making (SMDM) 48th Annual Meeting, Oslo, Norway, June 28 – July 1, 2026.
, & Yaesoubi, R. (2024). Cost-effectiveness of metformin vs. SGLT2 inhibitors as the first-line treatment in type 2 diabetes management. Poster, Society for Medical Decision Making (SMDM) 46th Annual North American Meeting, Boston, MA, October 27–30, 2024.
, Öztosun, Ç., Acar, S., Gürsoy, T., Sakarya, S., & Ata, B. (2022). The use of third-party multiple-choice examinations for measurement and evaluation in clinical medical education. Poster, 12th National Medical Education Congress (XII. Ulusal Tıp Eğitimi Kongresi), Türkiye, May 19–22, 2022.
Dural, D., Cunedioglu, B. Ö., Yurtseven, E., , Gürsoy, E., Aytekin, S., & Aytekin, V. (2021). Association between lipoprotein (a) level and chronic cardio-renal syndrome in patients with coronary artery disease. Abstract, World Congress on Heart Failure 2021, June 29 – July 1, 2021.
Dogan, O., Atac, N., Vatansever, C., Albayrak, O., Sahin, Ö. E., Kilicoglu, B. K., , Ergönül, Ö., Gönen, M., & Can, F. (2019). Yersiniabactin as a promoter of mortality of infections caused by colistin-resistant Klebsiella pneumoniae ST101. Poster, 29th European Congress of Clinical Microbiology & Infectious Diseases (ECCMID), Amsterdam, the Netherlands, April 13–16, 2019.
Master’s Thesis
(2023). Regulation of Artificial Intelligence in Medicine: Upholding Public Health. Erasmus University Rotterdam. Awarded Thesis of the Year in Health Economics, Policy and Law.