The study, led by Pia Weskamp and Christoph Wallner of the Sarcoma Center at BG University Hospital Bergmannsheil, Ruhr University Bochum, drew on an institutional registry of patients with histologically confirmed extremity leiomyosarcoma who underwent their first therapeutic resection between January 2000 and December 2015 and were ultimately treated at the index center. Of 82 registry records, three were excluded because their initial surgery predated 2000, leaving 31 women and 48 men. All diagnoses were confirmed by specialist sarcoma pathologists and reviewed by a multidisciplinary tumor board, and truncal, retroperitoneal, and head-and-neck primaries were excluded to keep the analysis entity-specific. The team chose initial margin status — R1 versus the microscopically complete R0 — as the primary endpoint precisely because it captures the quality of the first operation before any subsequent surgery can blur the picture.

The headline numbers are striking. Initial R1 resection occurred in 34 of 46 men, or 73.9 percent, compared with 13 of 29 women, or 44.8 percent — an absolute difference of 29.1 percentage points, with a risk ratio of 1.65 and a nominal P value of 0.015. Men were also far more likely to have had their first operation somewhere other than the sarcoma center: 72.3 percent versus 41.4 percent for women, a 31-point gap. And they more often required re-excision of the tumor bed, 82.6 percent versus 50.0 percent. When the researchers cross-tabulated the sequence of events, the pattern became even clearer: among patients with known location and margin status, R1 followed 37 of 44 outside-center resections but only 9 of 29 index-center resections. Every one of the 47 patients with an initial R1 margin went on to re-excision.

Crucially, the study did not stop at raw percentages. Because men in the cohort also carried a different recorded tumor profile — 64.6 percent had tumors recorded as smaller than 5 centimeters, versus 35.5 percent of women, and men’s tumors were less often deep to the fascia — the investigators used Firth penalized logistic regression, a technique designed for small samples where conventional models can fail, to see how the sex association held up under adjustment. The unadjusted male-sex odds ratio for initial R1 was 3.37. After accounting for age and anatomic site it barely moved, to 3.41. But adding recorded tumor size dropped it to 2.54, and adding depth brought it down further to 2.43 — an estimate whose 95 percent confidence interval, spanning 0.82 to 7.41, comfortably included the null.

To translate those odds ratios into something clinically digestible, the team turned to Bayesian marginal standardization, expressing the association as an absolute difference in R1 probability. Standardizing over age and site yielded a male-female gap of 29.5 percentage points. But once recorded size and depth entered the model, the fully standardized difference shrank to 16.6 points, with a 95 percent credible interval running from minus 2.5 to plus 37.4 — a range compatible with a small excess in women, no difference at all, or a substantial excess in men. The authors are emphatic that this interval should not be read as evidence of a residual sex effect; it is a measure of how much genuine uncertainty remains in a cohort of this size.

Why might smaller, more superficial tumors be overrepresented among men who fared worse at first surgery? The researchers point to a plausible recognition-and-referral mechanism. Small or superficial sarcomas can masquerade as benign lumps, generating less clinical suspicion and potentially altering the entire downstream cascade — imaging, biopsy, referral, and the choice of where the first operation happens. That phenotype matches what larger multicenter studies of unplanned sarcoma excision have reported, including a 728-patient European cohort in which unplanned excision was associated with smaller, superficial tumors and a borderline male-sex association, and a Japanese registry study that linked unplanned excision to male sex as well. Neither prior study, however, isolated a single sarcoma entity the way this analysis does.

The authors are unusually candid about the limits of causal language. They constructed a directed acyclic graph during revision to lay out competing biological, social, diagnostic, and selection mechanisms, while stressing that the graph was not prespecified and is not a validated causal model. Recorded sex, they note, may mark incompletely measured biological and social processes but says nothing about gender identity, health literacy, or help-seeking behavior. Tumor size and depth need not be confounders at all — they could lie on the very pathways linking sex-related processes to surgical management. And a critical structural limitation looms over the whole enterprise: patients whose outside-center R0 surgery succeeded were never referred and never entered the cohort. Conditioning on tertiary-center treatment in this way can induce spurious associations through selection, potentially changing even the direction of the observed sex difference.

What about consequences for patients? Here the picture is more reassuring, with an important caveat. Definitive R0 resection was ultimately achieved in all 79 patients, and exploratory survival analyses showed no meaningful differences: five-year overall survival was 77.3 percent after initial R1 versus 79.3 percent after R0, with wide confidence intervals that cannot establish equivalence. Local recurrence, interestingly, trended the other way — 35.5 percent in women versus 14.6 percent in men — a contrast the authors cannot fully explain given the small number of events. The main burden of an incomplete first operation, they argue, is the additional surgery itself: re-excising a previously operated tumor bed can demand wider tissue removal and extra reconstructive or radiotherapeutic treatment, even when the oncological endpoint is eventually salvaged.

The team also probed whether calendar time explained the pattern, splitting the cohort into 2000–2007 and 2008–2015 eras. Adding an era term barely changed the pooled sex estimate, but the crude differences were notably larger in the earlier period — the outside-center gap was 48.4 percentage points in 2000–2007 versus 13.3 points in 2008–2015 — hinting, though not proving, that centralization of sarcoma care and growing referral awareness may have narrowed whatever gap existed. With strata this small, the authors warn, no firm conclusion about temporal trends is possible.

Where does this leave patients and clinicians? The study’s contribution is an exploratory, entity-specific description of how men and women with the same rare tumor entered surgical care differently, with honest quantification of the uncertainty surrounding that difference. The authors explicitly state that the findings do not establish inequity, differential access to specialist care, biological causation, or mediation, and they call for prospective multicenter studies that capture patients from first clinical assessment — recording diagnostic sequence, measurement provenance, planned versus unplanned surgery, and social and geographic context. Until such data exist, the message is less about blame and more about vigilance: any soft-tissue lump, in any patient, deserves an imaging-and-biopsy pathway that treats a possible sarcoma seriously before the first scalpel moves.

Subject of Research: Sex-associated differences in initial surgical margins, treatment location, and re-excision in extremity leiomyosarcoma

Article Title: Sex-associated differences in initial resection location, margin status, and re-excision in extremity leiomyosarcoma: a retrospective tertiary-center cohort study

Article References: Weskamp, P., Schmidt, S. V., Drysch, M., Reinkemeier, F., Dadras, M., Lehnhardt, M., Puscz, F., & Wallner, C. (2026). Sex-associated differences in initial resection location, margin status, and re-excision in extremity leiomyosarcoma: a retrospective tertiary-center cohort study. Frontiers in Oncology, 16, Article 1980400. https://doi.org/10.3389/fonc.2026.1980400

Image Credits: AI Generated

DOI: 10.3389/fonc.2026.1980400

Keywords: leiomyosarcoma, soft tissue sarcoma, surgical margins, R1 resection, re-excision, sex differences, referral patterns, tertiary sarcoma center, retrospective cohort study, Bayesian standardization, Firth logistic regression, selection bias

The retrospective study, led by Malia Brennan and colleagues at the University of Washington and the Fred Hutchinson Cancer Center in Seattle, examined 103 patients diagnosed with HER2-positive de novo metastatic breast cancer between 2011 and 2023. The researchers asked two questions that have divided surgical oncologists for years: which patients in this population actually undergo breast surgery, and does the operation change the trajectory of local and distant disease progression or survival compared with systemic therapy alone?

The answer to the first question was clear. Nearly 43 percent of the cohort—42.7 percent—underwent breast surgery, and the choice was far from random. Patients who received surgery tended to be younger, with a median age at diagnosis of 48 years compared with 56 years among those managed without an operation, although that difference did not reach statistical significance. More telling, the surgery group carried a lower metastatic burden: an average of 1.3 sites of distant disease versus 1.9 sites in the non-surgery group, a difference that was statistically significant. In other words, surgeons and their patients were selecting the operations for those with less extensive spread.

The surgical details reveal how practice has evolved. Among those operated on, 70.5 percent underwent mastectomy and 29.5 percent lumpectomy. Axillary surgery was performed in 72.7 percent of surgical patients, but the mix of procedures shows the influence of modern, less invasive staging: 38.6 percent received a full axillary lymph node dissection, 27.3 percent a sentinel lymph node biopsy, and 6.8 percent a targeted axillary dissection—a technique that allows de-escalation of surgery after effective systemic treatment. The high rate of axillary intervention in a metastatic setting underscores how thoroughly the regional nodes were still being evaluated, even when distant disease was present.

Local disease progression emerged as a key selection factor. Before surgery, 39 percent of the surgical group had exhibited local progression at the primary site, compared with 23.5 percent of the non-surgery group—a difference that approached significance. This suggests that uncontrolled disease in the breast itself was one of the triggers pushing clinicians toward operative management. Yet after the operation, the tables turned: only 17.1 percent of surgical patients experienced subsequent local progression, versus 23.5 percent of those treated with systemic therapy alone, a difference that was not statistically significant. Surgery, in this dataset, did not decisively outperform modern drug therapy at controlling the primary site.

The survival signal, however, was dramatic. After adjusting for age, sites of metastases, and the number of metastatic sites in Cox proportional hazards models, patients who underwent breast surgery had significantly improved overall survival, with a hazard ratio of 0.33—meaning roughly a two-thirds reduction in the risk of death—compared with those treated without surgery. Kaplan-Meier analyses supported the same conclusion. On its face, that is the kind of result that could reshape guidelines overnight.

But the authors themselves urge restraint, and the statistical architecture of the study explains why. This was a single-institution retrospective review, not a randomized trial. Logistic regression and multivariable adjustment can account for measured confounders such as age and metastatic burden, but they cannot eliminate selection bias—the inherent tendency for healthier, lower-burden patients to be offered surgery in the first place. The very factors that predicted who got the operation, particularly the lower number of metastatic sites, are also factors that independently predict longer survival. The hazard ratio of 0.33 therefore represents an association, not proof of causation, and the authors state explicitly that the significant association between surgery and improved survival must be interpreted with caution.

The study arrives in a field crowded with competing evidence. Randomized trials have tested primary tumor resection in metastatic breast cancer for a decade: the 2015 Lancet Oncology trial by Badwe and colleagues found no survival benefit for locoregional treatment of the primary tumor, while the MF07-01 protocol by Soran and colleagues reported improved survival with surgery in stage IV disease. The EA2108 trial led by Khan and colleagues found that early local therapy for the primary site in de novo stage IV breast cancer did not improve survival overall. More recently, the Japanese JCOG1017 PRIM-BC trial and a 2024 meta-analysis of randomized controlled trials have added further nuance, and a 2025 Cochrane review systematically assessed breast surgery for metastatic breast cancer. Notably, prior analyses specifically focused on HER2-positive stage IV patients in the era of targeted therapy—such as work by Mudgway and colleagues and by Lu and colleagues—have suggested that the subtype may behave differently, which is precisely the niche this new single-institution study addresses.

What makes the HER2-positive question especially compelling is the transformation of the subtype’s prognosis by HER2-targeted therapy. Drugs directed against the human epidermal growth factor receptor 2 have significantly improved survival in de novo metastatic HER2-positive breast cancer, converting one of the most aggressive diagnoses into a chronic, manageable condition for many patients. As systemic therapy grows more effective, the calculus around local interventions shifts: if the drugs can hold distant disease at bay for years, the question of whether removing the breast adds incremental benefit—or merely risk—becomes empirically testable rather than doctrinal. The Seattle data suggest that in carefully selected patients with limited metastatic burden, surgery is being deployed, and that its association with survival deserves rigorous prospective testing.

The findings also intersect with a parallel movement in early-stage disease: trials such as those led by Kuerer and colleagues have explored eliminating breast surgery entirely for exceptional responders to neoadjuvant systemic therapy, showing that as drug treatment improves, the role of the scalpel is being renegotiated in both directions—expanded for some metastatic patients, withdrawn for some early-stage ones. For now, the practical takeaway for patients with HER2-positive de novo metastatic breast cancer is that surgery remains an option on the table, most often pursued in those with fewer sites of metastatic spread and uncontrolled local disease, and that the apparent survival advantage associated with the operation must await randomized evidence before it can be prescribed with confidence. The study was approved by the University of Washington Research IRB Committee, and the authors declare no competing interests or external funding.

Subject of Research: Medicine & Public Health

Article Title: Surgical management in HER2 + de novo metastatic breast cancer: disease progression and survival

Article References: Brennan, M., Javid, S. H., Flanagan, M. R., Schmicker, R., Gwin, W. R., Fleuret, S., & Palmquist, E. (2026). Surgical management in HER2 + de novo metastatic breast cancer: disease progression and survival. Breast Cancer Research and Treatment, 220(1), Article 7. https://doi.org/10.1007/s10549-026-08108-y

Image Credits: AI Generated

DOI: 10.1007/s10549-026-08108-y

Keywords: breast cancer, HER2-positive, metastatic breast cancer, surgical oncology, de novo stage IV, systemic therapy, survival outcomes, selection bias, Surgical, management, HER2, novo

The problem is deceptively simple. Studies of post-encroachment time typically rely on observed values extracted from video analysis, which limits the sample to vehicles that engaged in at least one interaction with another vehicle or vulnerable road user. Vehicles that pass through the intersection without any observable engagement are excluded from the analysis. Yet these non-engaging vehicles may represent a fundamentally different portion of the driving population—one whose behavior under higher traffic volumes or platooned arrivals could differ markedly from those that did engage. The counterfactual question—what would the post-encroachment time have been for a vehicle that never engaged?—is the critical one, because it would present a full description of the true underlying propensity for dangerous interactions in terms of both duration and frequency.

This is not a new concern in traffic safety research. Prior studies have raised the alarm about selection bias in crash severity modeling, noting that riskier drivers can be over-represented in reported crash data, leading to biased and inconsistent parameter estimates. Similar selectivity problems have been documented during the COVID-19 pandemic, where people who chose to engage in social activities carried an underlying risk proneness that also predisposed them to riskier driving. But the Texas Tech team, led by Karam Daoud, Abdallah Alrawashdeh, and Narayan Venkataraman, is among the first to tackle the issue in the context of post-encroachment time engagements at signalized intersections using a full-information maximum likelihood framework.

The conventional tool for correcting selection bias is the Heckman two-step approach, where the probability of an engagement or crash is factored into a second-stage model. But the Heckman technique relies on the Inverse Mills ratio, which provides meaningful results only in a linear regression setting. Its applicability to nonlinear settings—such as the ordered probit models needed for severity analysis—is questionable. The full-information maximum likelihood approach, by contrast, jointly models the selection and outcome equations, measuring the degree of selection directly through a correlation parameter between the error terms of the two equations. This correlation arises from unobservables shared between the selection process and the outcome process—factors like car-following behavior in downtown environments or pedestrian crossing tendencies that influence both the probability of engagement and its severity.

The computational complexity of the full-information maximum likelihood approach has historically limited its use, especially in larger datasets. The Texas Tech researchers overcame this by carefully scaling variables and re-parameterizing the correlation parameter on the real line, which supported stable and smooth convergence without the need to switch optimization techniques mid-estimation. They transformed the correlation parameter through the inverse hyperbolic tangent function, expanding the parameter space to enable smoother likelihood behavior. The result was a convergent joint model applied to 82,666 observations extracted from the Bellevue intersection, with 30,185 observations yielding measured post-encroachment times and the remaining 52,481 contributing to the selection analysis.

The severity outcome was modeled as an ordered probit with three levels: interaction (post-encroachment time greater than 3 seconds), minor engagement (greater than 2 seconds and less than or equal to 3 seconds), and critical engagement (less than or equal to 2 seconds). These thresholds align closely with the Bellevue Vision Zero conflict-classification framework, which defines post-encroachment time values below 2 seconds as critical and between 2 and 3 seconds as minor. The dataset contained 28,000 observations for interactions, 1,399 for minor engagements, and 786 for critical engagements—a stark imbalance that reflects the reality that most engagements at signalized intersections are low-severity interactions rather than near-crashes.

The joint model converged at a log-likelihood of −54,056.836, compared to −54,061.188 for the ordered probit model without selection. The likelihood-ratio test produced a chi-squared value of 8.704, demonstrating statistically significant rejection of the null hypothesis that there is no correlation due to unobservables. The correlation parameter itself was estimated at −0.174 with a t-statistic of −2.78, confirming that shared unobservables—hypothesized to include driver tendencies unique to the downtown context such as interactions with parked vehicles, non-motorized traffic, gap acceptance behavior, and perceived turning movement safety—exert a real influence on both engagement occurrence and severity.

When the researchers compared coefficients from the joint model to those from the naïve ordered probit without selection, the bias was unmistakable. Standard errors in the outcome equation were on average 14 percent higher under selection. Coefficient bias averaged 5.8 percent, but the largest distortions appeared for the speed ratio variable—measuring the ratio of average to median trajectory speed as a proxy for speed skewness—and the daytime indicator, with changes of 35 percent and 73 percent respectively. The speed ratio effect was over-estimated in magnitude in the presence of selection, while the daytime variable was under-estimated. Signs remained unchanged, but the true direction and magnitude of effects by severity level could only be recovered through a comprehensive analysis of marginal effects.

The marginal effects analysis revealed three types of behavioral insight: direct effects from variables in the severity equation, indirect effects from variables in the selection equation, and full effects combining both. The speed skew measure exerted noteworthy full marginal effects on the probability of the lowest severity level, as did lane width difference and lane count difference between departure and approach legs. The lane count difference variable—the difference between the number of lanes on the departure leg and the approach leg—had the greatest magnitude effect, contributing a −0.449 decrease in the probability of interaction engagement while increasing the probabilities of minor and critical engagements. This suggests that lane balance between approach and departure legs can yield lowered encroachment severity likelihoods, while dedicated turning movement lanes or parking-induced lane reductions may inadvertently escalate engagement severities through merging maneuvers.

The policy implications are concrete. Traffic control measures that mitigate right-skewed speed distributions can yield beneficial effects in terms of lowered post-encroachment times. Narrower approach lane widths to accommodate higher turning movement capacity may support lower encroachment severities. Approach leg medians, sharper turning radii, and departure leg right shoulders all exerted measurable effects on engagement probabilities across the severity spectrum. The researchers recommend extending the 10-second measurement technique to multiple intersections to capture richer behavioral insights across downtown, urban, suburban, and rural contexts, particularly for variables related to traffic operations and control, capacity recapture policies such as elimination of approach leg parking, and lane balance decisions. The study focused on a single intersection, so findings capture within-site variations and selection effects—but the framework itself offers a replicable template for any traffic safety analysis that relies on observed engagements rather than the full population of vehicles passing through a conflict zone.

Subject of Research: Selection bias in post-encroachment time conflict severity modeling using a full information maximum likelihood framework

Article Title: Selection bias in post-encroachment time conflict severity modeling using a full information maximum likelihood framework

Article References: Daoud, K., Alrawashdeh, A., & Venkataraman, N. (2026). Selection bias in post-encroachment time conflict severity modeling using a full information maximum likelihood framework. Frontiers in Future Transportation, 7, Article 1893114. https://doi.org/10.3389/ffutr.2026.1893114

Image Credits: AI Generated

DOI: 10.3389/ffutr.2026.1893114

Keywords: traffic safety, selection bias, post-encroachment time, full-information maximum likelihood, ordered probit, signalized intersections, trajectory analysis, Vision Zero, Selection, bias, post-encroachment, time

The study drew on 2,270 health-examination records generated in 2025 through China’s National Basic Public Health Service program, which provides annual checkups for residents aged 65 and older. The records came from a rural primary-care program in Wangkui County, Heilongjiang Province, an agricultural region where 20 examination sites contributed data. The researchers, led by Kun Xu of Henan University’s Institute of Gerontology, set out to answer a deceptively simple question: could the routine main-medication field in these records be trusted as a measure of medication exposure in secondary research?

The answer turned on a comparison that initially looked compelling. Records were split by the exercise-frequency item into those with recorded physical activity, meaning daily, weekly, or occasional exercise, and those recording no exercise at all. Among the 2,226 records with usable activity information, nonblank medication fields appeared in 56.4 percent of the active group versus 42.8 percent of the inactive group, a crude difference of 13.6 percentage points. A modified Poisson regression, chosen because the outcome was common, produced a prevalence ratio of 1.317 with a confidence interval comfortably excluding the null. On its face, older adults who exercised appeared far more likely to be taking medications.

But that interpretation collapses under scrutiny of the data-generating process. Under the national health-service specification, the main-medication item was never designed as a population-wide drug inventory. It was intended to document principal medications used during the preceding year among people receiving long-term treatment for chronic disease, specifically within hypertension and diabetes management programs. When the researchers cross-tabulated field presence against chronic-disease management registration, the pattern was stark: 96.3 percent of records with hypertension and/or diabetes registration had nonblank medication fields, compared with just 2.4 percent of records with neither registration. The field was essentially a switch flipped by enrollment in a disease-management pathway.

Because exercise status was itself correlated with registration, the apparent activity contrast was a mixture of two things: any real relationship between activity and medication use, and the fact that active records were more likely to belong to the registered group where medication fields were almost always filled. The researchers built a sequence of models to watch the contrast dissolve. After adjusting for age, sex, and examination site, the prevalence ratio actually rose slightly to 1.346. But adding a four-category hypertension-diabetes registration variable, which allowed each combination of registrations its own effect, dropped the ratio to 1.001, with a confidence interval from 0.960 to 1.043. Further adjustment for body mass index, smoking, and alcohol use left it essentially unchanged at 1.002.

The absolute scale told the same story. Marginal standardization of a logistic model, which sets every record’s activity status in turn while holding all other covariates at observed values, yielded standardized field-presence proportions of 51.26 percent for the active group and 51.14 percent for the inactive group, a difference of just 0.12 percentage points with a confidence interval spanning negative 2.07 to 2.30. Within registration strata, the differences were tiny: negative 1.41 percentage points among unregistered records, 0.44 among hypertension-only records, and negative 0.56 among dual-registered records. Only the sparse diabetes-only stratum, with just 37 records, showed a large descriptive difference, and the authors explicitly declined to generalize from it.

The team did not stop at field presence. They audited the medication text itself, splitting 1,141 nonblank fields into 1,368 retained text entries after removing exact duplicates. A version-controlled rule-based parser identified medication names, doses, units, and duration information, and a random sample of 200 entries was independently reviewed by two pharmacists masked to the parser’s output. Inter-reviewer agreement on broad therapeutic class reached 99.5 percent, with a kappa of 0.982, and parser-to-manual concordance on that domain was 99.0 percent. Archived labels classified 75.9 percent of entries as antihypertensive and 14.5 percent as glucose-lowering, with 7.4 percent unresolved. Yet the authors are careful to note that this high concordance validates only the interpretation of recorded text, not actual medication exposure, because no linked prescriptions, dispensing records, or medication reconciliation data existed.

The audit also exposed structural ambiguities that would plague any secondary analysis. Nearly half of the entries, 45.0 percent, contained a bare number in the duration field with no time unit, and another 15.5 percent held timing or frequency information rather than a treatment duration. Standardized medication naming showed more variability than direct text extraction, with parser-to-reference concordance of 77.5 percent on exact normalized matching, reflecting differences in formulation qualifiers, salt wording, and dictionary granularity. A blank field, meanwhile, could mean no long-term medication, inapplicability, the question never asked, unknown status, refusal, or simple omission, all compressed into an undifferentiated empty cell.

The implications reach well beyond one county in Heilongjiang. Routine health data are products of healthcare and administrative processes rather than neutral research measurements, and informative field presence can generate associations that reflect documentation architecture rather than clinical reality. The authors sketch the specific risks: coding blank fields as medication nonuse would produce differential exposure misclassification patterned by disease registration; restricting analyses to records with nonblank fields would introduce selection bias by preferentially retaining records from chronic-disease pathways; and the intended one-year reference period may misalign with outcomes measured at a single visit. They propose a tiered approach, reserving the field for documentation audits and narrowly defined descriptive work within its intended population, while requiring validation against prescriptions, dispensing data, or structured medication reconciliation for any study of drug utilization, safety, effectiveness, or medication-outcome associations.

The study also offers concrete fixes for primary-care information systems: separating disease-management prompts from comprehensive medication lists, storing identity, indication, dose, route, frequency, and status in distinct structured elements, adopting standardized terminologies such as RxNorm and the ATC/DDD classification, and replacing undifferentiated blanks with explicit categories distinguishing nonuse, inapplicability, and missingness. The limitations are acknowledged candidly, including the single-center setting, the cross-sectional design, the absence of stable person-level identifiers, and reliance on only 20 site clusters. But the central message stands with unusual clarity: before any routinely collected variable is reused as research data, its intended population, recording workflow, and reference period must be understood. A field fit for its original operational purpose can be quietly unfit for the next question asked of it, and the ghost of that purpose can masquerade as a scientific finding.

Subject of Research: Secondary-use suitability of a routine primary-care medication field and its association with recorded physical activity among rural older adults

Article Title: Among rural older adults, an apparent physical-activity contrast in nonblank medication-field presence was markedly attenuated when chronic-disease management registration was incorporated into the analysis

Article References: Xu, K., Zhang, Y., Zhao, Y., Xi, X., Hou, G., Guo, Z., Liu, L., Wang, Y., & Zhang, D. (2026). Among rural older adults, an apparent physical-activity contrast in nonblank medication-field presence was markedly attenuated when chronic-disease management registration was incorporated into the analysis. Frontiers in Pharmacology, 17, Article 1965382. https://doi.org/10.3389/fphar.2026.1965382

Image Credits: AI Generated

DOI: 10.3389/fphar.2026.1965382

Keywords: pharmacoepidemiology, real-world data, medication data quality, electronic health records, older adults, physical activity, chronic disease management, exposure misclassification, primary care, rural China, data provenance, selection bias

The study, led by Joseph McGrath and Evelyn Fagan Burley, with senior authorship from Sarah Lonowski and Kaleb Michaud at the University of Nebraska Medical Center, set out to assess representation in the FORWARD registry, a nation-wide, patient-reported psoriasis and psoriatic arthritis registry. The team’s approach was deliberately grounded: rather than comparing FORWARD against census data or other registries, they compared it against the electronic medical record of a single academic institution, using that clinic population as a benchmark for the kinds of patients who seek care for psoriasis in a real-world setting. The work was approved under institutional review board protocol 0770-24-EP, and the authors report no funding and no conflicts of interest.

FORWARD is not a small or casual undertaking. Described in a 2024 paper in the Journal of Psoriasis and Psoriatic Arthritis, it is a novel registry built around patient-reported outcomes, collecting longitudinal data directly from people living with psoriasis and psoriatic arthritis across the country. Patient-reported outcome registries occupy an important niche in dermatology because many of the outcomes that matter most to patients, such as itch, fatigue, skin-related stigma, and functional impairment, are poorly captured by routine clinical documentation. By asking patients directly and repeatedly over time, FORWARD can track disease trajectories in a way that claims data or chart reviews cannot.

Yet the very design that gives such registries their power also introduces a structural vulnerability. Enrollment in FORWARD is voluntary and, like many contemporary registries, it relies heavily on digital infrastructure for recruitment and follow-up. A well-documented body of research, including a 2019 analysis in the Journal of the American Medical Informatics Association by Toscos and colleagues, has shown that technology-based research participation is not demographically neutral: patients who sign up for app-based or web-based studies tend to differ systematically from those who do not, in ways that track both socioeconomic status and health status. If those selection pressures operate within FORWARD, the registry’s findings could quietly drift away from the broader psoriasis population it is meant to represent.

This is the concern that motivated the single-institution comparison. Registry science in dermatology has wrestled with questions of validity for more than a decade. A 2017 commentary in the British Journal of Dermatology by Egeberg and Nast urged a critical eye on registry data in psoriasis, and a 2011 systematic review in BMC Medical Research Methodology by Langham and colleagues found that large prospective observational studies in psoriasis and psoriatic arthritis varied widely in quality and in how transparently they reported potential biases. A 2016 review of patient registries in dermatology by DiMarco, Hill, and Feldman in the Journal of the American Academy of Dermatology similarly catalogued the methodological trade-offs inherent in registry design. Against that backdrop, the Nebraska team’s audit is less an indictment than a routine but essential quality check, the epidemiological equivalent of calibrating an instrument before trusting its readings.

The logic of the comparison deserves unpacking, because it illustrates a core technique in registry validation. When researchers want to know whether a national registry is representative, they need an external reference population. Census data offer demographics but no clinical detail; claims databases offer clinical breadth but shallow phenotyping. A single academic institution’s electronic medical record, by contrast, provides a richly characterized cohort of diagnosed psoriasis patients with documented demographics, comorbidities, and treatment histories. By comparing the characteristics of FORWARD enrollees against eligible patients at the institution, the researchers could identify dimensions along which the registry over- or under-represents the clinical population, whether by age, sex, disease severity, psoriatic arthritis status, or treatment intensity.

Why does this matter so much for psoriasis specifically? Psoriasis is a chronic immune-mediated disease affecting roughly two to three percent of many adult populations, and its management has been transformed over the past two decades by biologics and targeted oral therapies. Decisions about which patients should escalate to expensive systemic therapy, how to weigh cardiovascular and metabolic comorbidity in treatment planning, and how to monitor long-term drug safety all draw heavily on observational evidence. If registry cohorts skew toward younger, more digitally connected, more engaged patients, then estimates of treatment effectiveness, adverse event rates, and comorbidity burden may not transfer cleanly to older patients, those with limited internet access, or those managed outside specialty centers. Generalizability, in other words, is not an abstract statistical virtue; it determines whose care is guided by the evidence.

The study also speaks to a broader and increasingly urgent conversation about equity in data-driven medicine. Registries feed algorithms, guidelines, and policy, and any systematic under-representation within them propagates silently into those downstream products. The authors’ decision to test representation using a concrete, verifiable comparator, rather than simply asserting generalizability, models a practice that other registry-based fields would do well to emulate. It also highlights the distinctive value of academic medical centers as audit points: their patient populations are themselves imperfect samples of the general population, but they are transparent, documentable samples, which makes them useful yardsticks even with their own known biases.

It is worth emphasizing what this research letter does not claim. A single-institution comparison cannot, by itself, establish how representative FORWARD is of the entire United States psoriasis population, and the authors present their work as an assessment rather than a definitive verdict. The data underlying the comparison were drawn from the institution’s electronic medical record under IRB oversight and were deliberately not made publicly available, in order to protect patient confidentiality, a constraint that reflects the tension between methodological transparency and privacy in clinical data research. The letter’s contribution lies in its method and its framing: it demonstrates that representation audits are feasible, inexpensive, and informative, and it invites similar checks at other institutions to build a fuller picture of where the registry aligns with clinical reality and where it diverges.

For clinicians and patients, the practical takeaway is one of informed confidence. FORWARD remains a vital resource, and the fact that its stewards, including Michaud, a veteran registry scientist best known for decades of work on rheumatology registries, subjected their own cohort to external scrutiny is a mark of methodological seriousness rather than weakness. The study, published as volume 318, article 423 of the Archives of Dermatological Research, arrives at a moment when real-world evidence is being asked to carry more weight than ever in dermatology. Ensuring that the patients inside the database resemble the patients in the exam room is the unglamorous, foundational work on which everything else, from drug safety signals to treatment guidelines, ultimately depends. As registries grow larger and more influential, expect more studies of this kind, quietly checking the mirror to make sure it reflects everyone it claims to.

Subject of Research: Evaluation of demographic and clinical representation in the FORWARD national psoriasis registry using a single-institution electronic medical record comparison

Article Title: Assessing representation in a nation-wide psoriasis registry: a single-institution comparison

Article References: McGrath, J., Burley, E. F., Sullivan, C., Kilton, M., Michaud, K., & Lonowski, S. (2026). Assessing representation in a nation-wide psoriasis registry: a single-institution comparison. Archives of Dermatological Research, 318(1), Article 423. https://doi.org/10.1007/s00403-026-04860-9

Image Credits: AI Generated

DOI: 10.1007/s00403-026-04860-9

Keywords: FORWARD registry, psoriasis, psoriatic arthritis, patient registries, representation, generalizability, selection bias, patient-reported outcomes, dermatology, electronic medical records, real-world evidence, registry methodology

The review, published in the journal Trends in Psychology, was conducted by Nicolas de Oliveira Cardoso, Thaiane Moreira de Oliveira, Luisa Massarani, Ketlin da Rosa Tagliapietra, and Wagner de Lara Machado, researchers affiliated with Federal Fluminense University, the Oswaldo Cruz Foundation, and the Pontifical Catholic University of Rio Grande do Sul. The team searched seven databases and applied the Cochrane risk-of-bias tool for randomized trials, known as RoB-2, to each of the twelve studies that met their inclusion criteria. RoB-2 is the current gold standard for evaluating whether the results of a randomized trial can be trusted, examining domains such as the randomization process, deviations from intended interventions, missing outcome data, measurement of outcomes, and selective reporting. Using it in the context of climate communication research is relatively unusual, and that is precisely what makes the review notable.

The trials the reviewers examined represent the leading edge of belief-change research. Several tested psychological inoculation, a strategy in which people are pre-exposed to weakened forms of misleading arguments so they can recognize and resist them later. Work by researchers such as Sander van der Linden and colleagues has suggested that inoculation messages can confer resistance to climate misinformation, and a 2024 study in Nature Human Behaviour tested inoculation strategies across twelve countries. Other trials in the review focused on correcting misinformation after exposure, comparing different correction formats and placements, including on social media platforms such as Instagram. Still others examined whether communicating the overwhelming scientific consensus on climate change—through plain facts, visual displays, or metaphors—could shift beliefs, building on the influential idea that perceived consensus acts as a gateway belief.

What the reviewers found was not that these ideas are wrong in principle, but that the evidence base supporting them is far shakier than the field’s enthusiasm would suggest. All but one of the twelve randomized trials received a judgment of some concern or high risk of bias under RoB-2. In practical terms, this means that for nearly every study, there were credible methodological reasons to doubt whether the reported effects reflected the true impact of the intervention. The reviewers highlight two biases as particularly prevalent: selection bias and instrumentation bias. Selection bias arises when the way participants are recruited or allocated produces groups that differ before the intervention even begins, so any observed difference in climate beliefs may stem from those pre-existing differences rather than from the message itself. Instrumentation bias, meanwhile, occurs when the measurement tools used to assess outcomes are flawed, unstable, or poorly adapted, distorting the apparent size of change between pre-test and post-test.

The connection between these biases and the field’s underwhelming results is more than coincidental, the authors argue. When trials are methodologically fragile, effects tend to shrink or vanish under more rigorous scrutiny, and the predominance of interventions with small or no effects may partly be a symptom of the biases themselves rather than a pure reflection of how malleable climate beliefs really are. This reframing matters. A pessimistic reading of the literature would conclude that people’s climate beliefs are simply resistant to change. The review suggests a more nuanced interpretation: the true effectiveness of belief interventions cannot yet be established with confidence, because the studies designed to measure that effectiveness are themselves compromised.

The reviewers also point to a striking geographic imbalance. The included studies came overwhelmingly from high-income countries, with the United States dominating the sample. This matters for generalizability. Climate beliefs are shaped by political systems, media environments, cultural values, and direct experience of environmental change, all of which vary enormously across contexts. An inoculation message calibrated for American audiences steeped in polarized media discourse may behave very differently in Brazil, India, or Nigeria, where the drivers of skepticism or concern about climate change can be entirely different. The authors call for future randomized trials to evaluate interventions in low- and middle-income populations, a gap they describe as essential if the field wants its findings to speak to the global public rather than a narrow slice of it.

On the measurement side, the review offers a constructive lead. The authors identified four psychometrically sound instruments for measuring climate change beliefs, tools that have undergone proper validation procedures for reliability and validity. Adopting such instruments, they argue, is essential for reducing instrumentation bias in future trials. The broader point echoes long-standing principles from psychometric theory: an intervention study is only as credible as the scale used to detect change. If a questionnaire’s wording, response alternatives, or scoring shift in meaning between measurement points, or if it was never validated in the population being studied, then apparent pre-test to post-test gains may be artifacts rather than genuine belief change. Research on question wording effects in climate surveys has shown how sensitive these measures can be, making the choice of instrument a first-order methodological decision rather than an afterthought.

To tackle selection bias, the reviewers recommend pre-test/post-test designs combined with clearly established cut-off points for climate belief levels. The logic is familiar from classic experimental design: measuring participants before randomization allows researchers to verify that groups are comparable at baseline and to adjust statistically for any residual differences, while predefined thresholds for what counts as a given belief level reduce researcher discretion in interpreting outcomes. The authors also sketch a forward-looking agenda, outlining directions for the design of future randomized controlled trials and for meta-analyses that would synthesize the evidence more rigorously. Their hope is that integrating these recommendations—better designs, better instruments, broader populations—could improve both the accuracy and the generalizability of the next generation of climate belief interventions.

The stakes of this methodological reckoning are considerable. Climate change is already producing measurable harms to physical and mental health, and the scientific consensus on its human causes is essentially complete. Yet public belief lags, sustained in part by organized disinformation and conspiracy narratives, which meta-analytic work has linked to reduced willingness to engage in climate-friendly behavior and political action. If the research community cannot demonstrate, with trials that survive bias assessment, which communication strategies actually work, then practitioners, educators, and policymakers are left choosing among interventions on the basis of weak evidence. The review does not claim that inoculation, consensus messaging, or debunking are useless; it claims something more unsettling—that the field does not yet know, because the trials testing them are too fragile to say.

There is, ultimately, a constructive message embedded in this critical one. The fact that eleven of twelve trials raised concerns is a sign of a young, fast-moving field that has prioritized creative ideas over methodological armor. The RoB-2 framework, validated measurement instruments, pre-registered designs with baseline measurement, and deliberate attention to underrepresented populations are all available tools; what the review demands is that they be used systematically. If future trials meet that standard, the field will finally be able to distinguish between interventions that genuinely move climate beliefs and those that merely looked effective in flawed studies. For a problem as urgent as climate change, knowing the difference may be one of the most valuable contributions psychology can make.

Subject of Research: Effectiveness and risk of bias in randomized controlled trials of interventions to change climate change beliefs

Article Title: Limited Impact of Interventions for Climate Belief: A Systematic Review Assessing Risk of Bias

Article References: de Oliveira Cardoso, N., de Oliveira, T. M., Massarani, L., da Rosa Tagliapietra, K., & de Lara Machado, W. (2025). Limited Impact of Interventions for Climate Belief: A Systematic Review Assessing Risk of Bias. Trends in Psychology. https://doi.org/10.1007/s43076-025-00485-5

Image Credits: AI Generated

DOI: 10.1007/s43076-025-00485-5

Keywords: climate change beliefs, systematic review, risk of bias, randomized controlled trials, psychological inoculation, misinformation, scientific consensus, environmental psychology, selection bias, instrumentation bias, psychometrics, climate communication

The study, published in the journal Current Psychology by Anna Włodarska, Anna Zajenkowska, and Izabela Kaźmierczak of VIZJA University in Warsaw, addresses a long-standing methodological worry known as volunteer bias or self-selection bias. Scientific research is vulnerable to both random error, which scatters measurements unpredictably, and systematic error, which pushes them consistently in one direction. Random error can be tamed with larger samples and better instruments. Systematic error is more insidious, because it can masquerade as a genuine finding. If the people who volunteer for studies differ from the population they are meant to represent, then every downstream statistic inherits that distortion, no matter how elegant the analysis.

Volunteer bias has been documented for decades in psychology, where researchers have repeatedly noticed that participants in behavioral studies tend to score differently on personality and affect measures than non-volunteers. Earlier work by some of the same Polish team, published in PLOS ONE in 2023, found that personality and affective disturbances were prevalent among participants in psychological studies, hinting that the problem extended beyond simple demographic skew. The new study extends that logic into clinical research, where the stakes are considerably higher. Clinical trials generate the evidence on which drug approvals, treatment guidelines, and ultimately patient care depend. If trial volunteers carry a distinctive psychological profile, the safety and efficacy data produced by those trials may not generalize cleanly to the broader patient population.

To investigate, the team drew on a representative community sample of Polish adults, a design choice that matters enormously. Most studies of volunteer bias compare actual trial participants with actual non-participants at a single research site, which limits what can be said about the wider population. By surveying a representative cross-section of the community and simply asking respondents about their interest in participating in psychological studies and clinical trials, the researchers could estimate how personality pathology is distributed across the willing and the unwilling in the population at large. Participants completed validated measures, including the Polish adaptations of the Personality Inventory for ICD-11 and the Level of Personality Functioning Scale-Brief Form 2.0, alongside the Patient Health Questionnaire-9 for depressive symptoms.

The theoretical framework underlying the measurement is worth understanding, because it represents a major shift in how psychiatry conceptualizes personality disorder. Both the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders and the eleventh edition of the International Classification of Diseases have moved toward dimensional models, in which personality pathology is understood as a matter of degree rather than a set of discrete categories. The ICD-11 model assesses two things: a level of personality functioning, which captures how severely a person’s sense of self and relationships are impaired, and a set of pathological trait domains, including negative affectivity, disinhibition, dissociality, and others. This dimensional approach allowed the Polish team to measure personality pathology continuously across their entire sample rather than only in people with formal diagnoses.

The results were striking in their consistency. Individuals who said they were interested in participating in clinical trials showed lower levels of personality functioning, meaning greater impairment in self-organization and interpersonal capacities, than those who were not interested. They also scored higher on negative affectivity, the trait domain encompassing anxiety, emotional lability, and a tendency to experience distressing emotions, and on disinhibition, which reflects impulsivity and difficulty with planning and restraint. Dissociality, a trait involving callousness and disregard for the rights and feelings of others, was likewise elevated. On top of the personality findings, interested respondents reported more depressive symptoms than their uninterested peers, suggesting that mood disturbance and personality pathology travel together in the population of potential volunteers.

Perhaps the most intriguing finding concerned the type of trial. The highest scores on personality pathology and depressive symptoms appeared among respondents interested specifically in psychiatric drug trials. The authors suggest a plausible explanation rooted in perceived benefit: people struggling with psychological distress may see psychiatric drug trials as a route to new treatments and structured care that they might not otherwise access. This creates a paradox with real methodological consequences. The very trials designed to evaluate psychiatric medications may disproportionately attract participants whose personality functioning is impaired, meaning the samples on which psychiatric drug evidence rests could be psychologically distinctive in ways that current screening rarely captures.

Why should this matter for the integrity of trial results? Prior research offers several concrete mechanisms. Psychopathological traits have been shown to influence participant behavior in ways that complicate protocol adherence and data accuracy. Neuroticism, a close cousin of negative affectivity, is associated with larger and more prolonged physiological responses to emotional stimuli, altered cortisol reactivity, and even differences in antibody response to vaccination, all of which could interact with the outcomes a trial is designed to measure. Depressive symptoms can bias decision-making, and personality pathology is linked to poorer treatment outcomes and higher treatment utilization in clinical populations. Participants with elevated disinhibition may struggle with the demands of strict dosing schedules and follow-up visits, while those with high negative affectivity may report side effects differently, skewing safety data in either direction.

The study’s authors argue that their findings highlight the importance of screening trial participants for personality pathology, not to exclude them from research, but to account for their characteristics in the design and analysis of studies. There is a delicate ethical balance here. Volunteers with psychological distress are precisely the people who may benefit most from access to experimental treatments, and excluding them wholesale would be both unfair and scientifically wasteful. But ignoring the pattern risks a subtler form of harm: trial results that look generalizable but are actually calibrated to a psychologically atypical subset of the population, with safety profiles and efficacy estimates that may not transfer to typical patients.

The research, which was part of a larger project on hostile attribution and was approved by the Maria Grzegorzewska University Research Ethics Committee, adds a compelling new dimension to an old methodological conversation. It suggests that the gate through which every participant passes, the decision to volunteer, is itself psychologically patterned, and that the pattern runs toward pathology rather than away from it. For a field that has invested enormously in randomization, blinding, and statistical rigor, the message is humbling: some of the most consequential selection into the evidence base happens before a single protocol procedure begins. As clinical science grapples with replication problems and questions about generalizability, understanding who volunteers, and why, may prove to be one of the most underappreciated levers for improving the reliability of medicine itself.

Subject of Research: Volunteer bias in clinical trials linked to personality pathology and depressive symptoms

Article Title: Personality pathology and willingness to participate in psychological and clinical trials

Article References: Włodarska, A., Zajenkowska, A., & Kaźmierczak, I. (2026). Personality pathology and willingness to participate in psychological and clinical trials. Current Psychology, 45(19), Article 1566. https://doi.org/10.1007/s12144-026-10067-y

Image Credits: AI Generated

DOI: 10.1007/s12144-026-10067-y

Keywords: volunteer bias, selection bias, personality pathology, clinical trials, depressive symptoms, ICD-11, negative affectivity, disinhibition, psychiatric drug trials, research methodology, personality functioning, self-selection

The scale of Nigeria’s tomato paradox is striking. The country produced an estimated 3.84 million tonnes of tomatoes in 2020 and 3.58 million tonnes in 2021 across more than 844,000 hectares, making it Africa’s second-largest producer after Egypt. Yet national yields average a mere 4 to 7 metric tonnes per hectare, compared with 38.7 tonnes in Egypt and a remarkable 78.7 tonnes in South Africa. That shortfall forces Nigeria to spend roughly 170 million dollars each year importing tomatoes to meet domestic demand, a bitter irony for a nation with fertile rainforest soils and an equatorial climate well suited to the crop.

The roots of the yield gap lie in chronic underinvestment. Fewer than 10 percent of Nigeria’s smallholder farmers have access to formal agricultural credit, and even those who do often face punishing interest rates, stringent collateral demands, and slow administrative procedures. Without reliable financing, farmers cannot purchase improved seeds, fertilizers, irrigation equipment, or post-harvest handling technology. Credit constraints also block adoption of climate-smart and risk-mitigating practices, leaving growers exposed to environmental and market shocks that can wipe out an entire season’s income.

Measuring the true effect of credit is harder than it sounds, and this is where the study makes its methodological contribution. Farmers who obtain loans are not a random sample of the farming population. They tend to be more experienced, better educated, more commercially oriented, and more willing to take risks, and these same traits independently boost productivity. A naive comparison of yields between credit users and non-users would therefore conflate the effect of money with the effect of farmer ability, producing biased estimates. The researchers tackled this problem with an endogenous treatment regression model, which jointly models the decision to access credit using a Probit regression and the resulting yield outcome in a second-stage linear regression, correcting for both observable and unobservable differences between the two groups.

The data came from a cross-sectional survey of 300 smallholder tomato farmers in Oyo and Ondo States, two prominent tomato-producing areas in Southwestern Nigeria. Using a multistage sampling design, the team randomly selected agricultural zones, local government areas, and villages, ultimately interviewing farmers across 24 rural communities during the peak cropping season between August and September 2023. The questionnaire was pretested in non-sampled communities, and reliability checks yielded a Cronbach’s alpha of 0.78, indicating acceptable internal consistency. Respondents averaged 44 years of age, 13 years of farming experience, and 3.72 hectares of farmland, with a mean tomato yield of 518.56 kilograms per hectare.

The raw differences between credit users and non-users were substantial. Farmers with credit access harvested 587.23 kilograms per hectare on average, versus 412.64 kilograms for those without, a gap of nearly 175 kilograms per hectare. Credit users were also more likely to be male, married, better educated, members of agricultural cooperatives, and connected to extension services, and they lived closer to markets. Eighty-three percent of credit users belonged to cooperatives, compared with a significantly lower share among non-users, and 79 percent of credit recipients reported access to extension services. These systematic differences confirmed that simple comparisons would be misleading, and that a rigorous correction for selection bias was essential.

After applying the endogenous treatment framework, the headline result held firm. The average treatment effect of credit access on yield was 8.229, and the average treatment effect on the treated was 8.728, both statistically significant at the 1 percent level. In practical terms, farmers who accessed credit produced roughly eight to nine times more tomatoes than they would have without it, holding other factors constant. To guard against the possibility that this finding was an artifact of one particular model, the researchers ran two robustness checks. Propensity score matching, which compares credit users with statistically similar non-users, produced an ATT of 8.514, while two-stage least squares estimation using cooperative membership, extension access, and formal education as instruments yielded a coefficient of 8.381. The convergence of all three methods strengthens the case that credit genuinely drives productivity rather than merely accompanying it.

The diagnostic tests underpinning the analysis were equally rigorous. A Wu-Hausman test rejected the assumption that credit access is exogenous, confirming that ordinary regression would have produced biased estimates. The Hansen J statistic failed to reject the validity of the instruments, and the first-stage F-statistic of 19.84 exceeded the conventional threshold for instrument strength. The first-stage Probit model also revealed who gets credit and who does not: farming experience, market proximity, extension access, and cooperative membership all increased the likelihood of obtaining loans, while older age and larger numbers of dependents reduced it. The findings expose a troubling pattern in which older farmers and households with heavy dependency burdens are systematically excluded from formal finance, and where women, who made up only 32 percent of the sample and a smaller share of credit recipients, face structural barriers to the resources that drive yields.

The mechanism linking credit to productivity is straightforward but powerful. Liquidity allows farmers to buy improved seeds, fertilizers, and agrochemicals at the right time, hire labor during critical operations, and invest in irrigation, greenhouse technology, and integrated pest management. Credit also supports post-harvest storage and transportation, reducing the spoilage that plagues perishable crops like tomatoes. Beyond the farm gate, the authors argue that expanding credit access would generate benefits across the broader farming population, not just current beneficiaries, and they frame the findings within the Sustainable Development Goals on poverty reduction and zero hunger.

The policy implications are clear. The researchers recommend strengthening extension systems, promoting cooperative membership and group lending, and designing flexible loan products with seasonal repayment schedules and lower collateral requirements tailored to smallholder realities. Gender-sensitive and age-responsive credit schemes are essential to close the productivity gaps the study documented. The authors also acknowledge limitations: the sample is confined to two states and a single crop, the cross-sectional design limits causal inference over time, and unobserved factors such as risk preferences and informal lending arrangements may still play a role. Even so, the message for Nigeria’s policymakers is hard to ignore. With the country bleeding 170 million dollars a year on tomato imports, helping smallholder farmers borrow affordably may be one of the cheapest ways to grow more food at home.

Subject of Research: The effect of agricultural credit access on tomato yield among smallholder farmers in Nigeria

Article Title: Unlocking yield potential through credit access: insights from smallholder tomato farmers in Nigeria

Article References: Kehinde, A. D., Akinola, A., & Tijani, A. (2026). Unlocking yield potential through credit access: insights from smallholder tomato farmers in Nigeria. BMC Agriculture, 2(1), Article 3. https://doi.org/10.1186/s44399-025-00025-7

Image Credits: AI Generated

DOI: 10.1186/s44399-025-00025-7

Keywords: agricultural credit, tomato yield, smallholder farmers, Nigeria, endogenous treatment regression, selection bias, propensity score matching, two-stage least squares, cooperative membership, extension services, food security, rural finance