Long-acting injectable antipsychotics, known as LAIs, have long been promoted as a technological fix for one of schizophrenia treatment’s most stubborn problems: non-adherence. Oral antipsychotics work well when taken consistently, but studies have repeatedly shown that many patients miss doses, stop treatment during relapses, or take medication irregularly. LAIs, administered by injection at intervals ranging from monthly to once every three months, bypass the daily decision to take a pill. The pharmacological logic is straightforward—stable drug levels in the bloodstream should mean fewer relapses, fewer crises, and, by extension, less demand placed on the family members who shoulder the consequences of treatment failure.

The research team, led by Ya-Jun Sun and Yong-Yong Teng of The Third People’s Hospital of Zhuhai, set out to test that logic in a real-world setting rather than a tightly controlled trial. Between April 2023 and December 2024, they enrolled 538 community-dwelling adults with schizophrenia who were receiving routine care through community mental health services in four cities across Guangdong Province. Of these, 270 were receiving long-acting injectable paliperidone palmitate and 268 were on oral antipsychotics. Crucially, patients were not randomized; treatment allocation reflected ordinary clinical decision-making by physicians and patients, which makes the study a portrait of practice as it actually happens. Notably, paliperidone palmitate was fully reimbursed by the government throughout the study period, removing cost as a barrier to injection therapy.

The investigators tracked three intertwined outcomes at five time points across the year. The primary measure was family caregiver burden, assessed with the 24-item Family Burden Interview Schedule, a validated instrument that captures financial strain, disruption of household routines, effects on family leisure and interaction, and the physical and psychological health of caregivers themselves. Alongside it, they measured psychiatric symptoms using the Positive and Negative Syndrome Scale, the standard clinical yardstick for psychosis severity, and social functioning with the Personal and Social Performance Scale, which grades a patient’s capacity for work, relationships, and self-care.

The headline finding is a story of two trajectories that converge. Caregiver burden fell substantially in both groups. In the oral antipsychotic group, adjusted total burden scores dropped from 9.16 to 6.32, a reduction of 31.0 percent. In the injectable group, scores fell from 9.09 to 5.53, a 39.2 percent reduction. On the surface, the injectable group appears to have fared better—but the statistical analysis tells a more careful story. Using linear mixed-effects models, the standard framework for analyzing repeated measurements nested within individuals over time, the team tested whether the two groups followed genuinely different trajectories. After adjusting for multiple comparisons, no treatment-by-time interaction was significant, and the effect sizes were vanishingly small, with Cohen’s f² values at or below 0.003.

The multiplicity-adjusted 95 percent confidence intervals tell the most decisive part of the story. Every one of them lay entirely within a ±2.4-point band around zero difference—the threshold the researchers defined as clinically meaningful on the FBIS-24 scale. In other words, the data were precise enough to rule out a difference that would matter to families. This distinction between statistical noise and clinical relevance is one of the study’s most important technical contributions. A nominal p-value below 0.05 might tempt a less rigorous analysis into declaring victory for injections, but effect sizes near zero and confidence intervals confined to the relevance bound reveal that any apparent advantage is too small for a caregiver to notice in daily life.

The researchers also examined individual burden domains on an exploratory basis. At 48 weeks, three domains—family leisure, family interaction, and the mental health of caregivers—showed nominally significant differences favoring the injectable group. Yet even these signals were negligible in magnitude, with f² values of approximately 0.005, and none survived correction across the full family of exploratory tests. The authors are careful to frame these findings as hypothesis-generating only, a caution that reflects modern standards for controlling false discoveries when many sub-measures are tested simultaneously. Psychiatric symptoms and social functioning improved in both groups over the year, with no detectable differences between formulations, and sensitivity analyses using propensity-score overlap weighting—an approach that reweights the comparison groups to mimic the balance achieved by randomization—reached the same conclusions.

Why might injections fail to outperform pills in this particular context? The study’s setting offers a plausible explanation. All participants were clinically stable adults already embedded in a structured community-based rehabilitation program, with regular contact with mental health services and full government reimbursement removing financial obstacles. In such an environment, the community system itself may be doing the work that injections are supposed to do: ensuring consistent treatment, catching early warning signs, and supporting patients before crises develop. When adherence is actively supported by outreach and supervision, the pharmacological advantage of a depot injection may simply have nothing left to add. The authors explicitly note that the open questions remain whether LAIs would reduce caregiver burden in settings without full subsidy, or among acutely ill or poorly adherent patients—the populations in whom the theoretical advantage of injectables is strongest.

The broader significance of the study lies in where it places the measuring stick. Most trials of antipsychotic formulations focus on relapse rates, hospitalization, or symptom scores—outcomes that matter to clinicians and health systems. By centering family caregiver burden, the Guangdong team highlights an outcome that is often invisible in pharmacological comparisons but dominates the lived experience of schizophrenia, particularly in Asian family systems where caregivers assume substantial financial, physical, and psychosocial responsibilities. Schizophrenia ranks among the leading causes of disability worldwide when measured in disability-adjusted life years, and much of that disability radiates outward from the patient to the household.

For families and clinicians, the practical message is one of reassurance and choice. In a well-supported community mental health program, the choice between a daily pill and a long-acting injection did not change how heavily caregiving weighed on families over the course of a year—both paths led to substantial relief as symptoms improved and functioning recovered. That finding does not render injectables useless; they remain a valuable option for patients who struggle with oral adherence or prefer the convenience of less frequent dosing. But it does suggest that the infrastructure around treatment—consistent contact, structured rehabilitation, and financial access—may matter more to family wellbeing than the formulation inside the syringe or the capsule. As health systems worldwide expand community-based mental health care, this study offers a data-driven reminder that supporting the system may be as powerful as upgrading the drug.

Subject of Research: Comparing family caregiver burden trajectories under long-acting injectable versus oral antipsychotic treatment in community-based schizophrenia rehabilitation

Article Title: No clinically meaningful difference detected in trajectories of family burden in adults with schizophrenia receiving community-based rehabilitation: a 48-week prospective cohort study of long-acting injectable versus oral antipsychotics

Article References: Sun, Y.-J., Teng, Y.-Y., Wang, S.-B., Xie, X.-L., Chen, C., Xie, L.-Y., Li, J.-Y., Huang, J.-J., Mao, C., & Zhang, G.-C. (2026). No clinically meaningful difference detected in trajectories of family burden in adults with schizophrenia receiving community-based rehabilitation: a 48-week prospective cohort study of long-acting injectable versus oral antipsychotics. BMC Psychiatry. https://doi.org/10.1186/s12888-026-08665-y

Image Credits: AI Generated

DOI: 10.1186/s12888-026-08665-y

Keywords: schizophrenia, long-acting injectable antipsychotics, family caregiver burden, community mental health, prospective cohort study, paliperidone palmitate, psychopharmacology, caregiving, psychosis, Guangdong China, medication adherence, BMC Psychiatry

The study’s central innovation lies in its timing. Most previous research on infection and heart defects has focused on infections contracted during pregnancy, particularly in the first trimester when the embryonic heart is forming. This new analysis shifts the window backward, asking whether the state of a woman’s immune system and reproductive tract before she conceives might also shape her child’s cardiac risk. The investigators assessed two broad categories of preconception exposure. The first comprised active reproductive tract infections, specifically candidiasis and Chlamydia trachomatis infection, diagnosed at the preconception examination. The second comprised immunoglobulin G seropositivity to three well-known pathogens: cytomegalovirus, Toxoplasma gondii, and rubella virus. Immunoglobulin G antibodies indicate prior exposure to a pathogen and persist in the blood for years, so seropositivity serves as a marker of a woman’s immunological history rather than of an active acute illness at the time of testing.

The data came from China’s National Free Preconception Health Examination Project, a government program that offers voluntary health screening to couples planning marriage or pregnancy. All participants provided written informed consent, and the retrospective analysis of anonymized records was approved by the Institutional Review Board of Guangdong Women and Children Hospital. Because the screening program covers a defined population rather than a selected sample of hospital patients, the cohort reflects a broad cross-section of women of reproductive age in one of China’s most populous provinces, which strengthens the generalizability of the findings within similar settings.

Statistically, the team used multivariable modified Poisson regression with robust error variance, a modeling approach widely favored in epidemiology when the outcome of interest is relatively common and researchers want to estimate relative risks directly rather than odds ratios, which can overstate risk in such settings. The models adjusted for potential confounding factors, and the authors reported both adjusted relative risks and risk differences per 1,000 births, each accompanied by 95 percent confidence intervals. The primary outcome, congenital heart disease, was ascertained from the birth defects registry, providing a standardized and population-wide mechanism for capturing diagnoses rather than relying on self-report or single-hospital records.

The results were striking across the board. Chlamydia trachomatis infection detected before conception was associated with an adjusted relative risk of 1.89, meaning the risk of congenital heart disease in offspring was nearly doubled among exposed women, with a confidence interval of 1.14 to 3.15. Even more pronounced was the association with cytomegalovirus IgG seropositivity, which carried an adjusted relative risk of 2.70, with a confidence interval of 2.26 to 3.22. Candidiasis was linked to an adjusted relative risk of 1.50, rubella IgG seropositivity to 1.59, and Toxoplasma gondii IgG seropositivity to 1.32, all with confidence intervals excluding the null value of 1. In other words, every one of the five preconception infection-related exposures examined showed a statistically significant association with elevated offspring heart defect risk in the fully adjusted models.

Perhaps the most eye-catching numbers emerged when the researchers examined women exposed to more than one factor. Among women with both Chlamydia trachomatis infection and cytomegalovirus IgG seropositivity, the adjusted relative risk reached 5.14, with a confidence interval of 2.92 to 9.06. Women with candidiasis combined with cytomegalovirus IgG seropositivity showed an adjusted relative risk of 3.72, with a confidence interval of 2.53 to 5.49. These joint-exposure estimates suggest that women carrying multiple infection-related markers before conception may face substantially higher absolute risks than those with any single marker. However, the authors are careful to frame these findings appropriately: the elevated estimates in jointly exposed groups describe joint exposure patterns and should not be interpreted as evidence of biological interaction, meaning the study cannot establish that the two exposures amplify each other’s effects at a mechanistic level.

The biological plausibility of a link between maternal infection and fetal cardiac development has a long scientific pedigree. Congenital heart disease arises from disruptions in the intricate choreography of embryonic heart formation, which occurs largely within the first eight weeks of gestation. Maternal immune activation, whether triggered by active infection or by the lingering immunological landscape left by past pathogens, is thought in some contexts to influence early embryonic environments through inflammatory signaling, altered cytokine profiles, and effects on placental development. Rubella is a classic and well-established teratogen, capable of causing a constellation of congenital defects including cardiac anomalies when infection occurs in early pregnancy. Cytomegalovirus, the most common congenital infection worldwide, and Toxoplasma gondii, a parasite transmitted through undercooked meat or contact with cat feces, are both recognized causes of congenital disease, though their specific contribution to structural heart defects has been less clearly quantified. This study adds a preconception dimension to that picture, suggesting that the immunological terrain established before pregnancy may matter as much as events during gestation.

The methodological strengths of the study deserve emphasis. With more than 630,000 pregnancies and over 1,300 outcome cases, the analysis had the statistical power to detect associations of modest magnitude that smaller studies would miss. The prospective timing of the exposure assessment is another advantage: infection status and serology were measured at preconception examinations, before any pregnancy outcome was known, which reduces the risk of recall bias and reverse causation that plague studies relying on postnatal reporting. The use of a birth defects registry for outcome ascertainment provides systematic case identification across the province. Sensitivity analyses, in which the researchers re-ran their models under alternative assumptions, produced results consistent with the primary findings, adding confidence that the associations are not artifacts of particular modeling choices.

At the same time, the authors and readers alike must weigh the inherent limits of observational research. An association, however robust, does not prove causation. Seropositivity to cytomegalovirus, rubella, and Toxoplasma gondii reflects past exposure and may correlate with unmeasured characteristics, such as socioeconomic conditions, occupational exposures, dietary patterns, or broader immune profiles, that could themselves influence cardiac development risk. The study was conducted in a single province of China, and seroprevalence of these pathogens varies considerably across regions and income settings, so the absolute risks and even the magnitude of associations may differ elsewhere. The authors also note that the published version was shared early as a citable, peer-reviewed accepted manuscript subject to further editorial refinement, a reminder that details may be updated in the final version of record.

Even with those caveats, the public health implications are considerable. Congenital heart disease imposes a lifelong burden on affected children and families, often requiring surgery, ongoing cardiology care, and in severe cases resulting in early mortality. If preconception infection status genuinely contributes to risk, then the period before pregnancy becomes a critical and currently underused window for intervention. Screening and treating chlamydial infections and candidiasis before conception, counseling women on toxoplasmosis prevention, and ensuring rubella immunity through vaccination well ahead of pregnancy are all established, low-cost measures already embedded in many preconception care frameworks. The findings from Guangdong suggest that such measures may carry benefits beyond the infections themselves, potentially reducing the burden of the most common birth defect. For the millions of couples planning pregnancy each year, the message from this research is that the health of the next generation’s heart may begin taking shape long before the first heartbeat.

Subject of Research: Preconception maternal infection-related exposures and the risk of congenital heart disease in offspring

Article Title: Preconception maternal infection-related exposures and risk of congenital heart disease in offspring: evidence from a large population-based cohort in South China

Article References: Chen, Y., Xu, R., Wei, H., Gu, Q., Liu, Y., Zhu, J., Yang, L., Zhang, Y., & Tan, G. (2026). Preconception maternal infection-related exposures and risk of congenital heart disease in offspring: evidence from a large population-based cohort in South China. BMC Public Health. https://doi.org/10.1186/s12889-026-29782-z

Image Credits: AI Generated

DOI: 10.1186/s12889-026-29782-z

Keywords: congenital heart disease, preconception health, Chlamydia trachomatis, cytomegalovirus, rubella, Toxoplasma gondii, candidiasis, IgG seropositivity, population-based cohort, birth defects registry, maternal infection, Guangdong China