Heart rate variability monitoring identifies asymptomatic toddlers exposed to Zika virus during pregnancy

Although Zika virus (ZIKV) seems to be prominently neurotropic, there are some reports of involvement of other organs, particularly the heart. Of special concern are those children exposed prenatally to ZIKV and born with no microcephaly or other congenital anomaly. Electrocardiogram (ECG) - derived heart rate variability (HRV) metrics represent an attractive, low cost, widely deployable tool for early identification of such children. We hypothesized that HRV in such children would yield a biomarker of fetal ZIKV exposure. We investigated the HRV properties of 21 infants aged 4 to 25 months from Brazil. The infants were divided in two groups, the ZIKV-exposed (n=13) and controls (n=8). Single channel ECG was recorded in each child at ~15 months of age and HRV was analyzed in 5 min segments to provide a comprehensive characterization of the degree of variability and complexity of the heart rate. Using a cubic Support Vector Machine (SVM) classifier we identified babies as Zika cases or controls with negative predictive value of 92% and positive predictive value of 86%. Our results show that HRV metrics can help differentiate between ZIKV-affected, yet asymptomatic, and non-ZIKV exposed babies. We identified the Grid Count as the best HRV measure in this study allowing such differentiation, regardless the presence of microcephaly. We show that it is feasible to measure HRV in infants and toddlers using a small non-invasive portable ECG device and that such approach may uncover memory of in utero exposure to ZIKV. This approach may be useful for future studies and low-cost screening tools involving this challenging to examine population.

💡 Research Summary

This pilot study investigates whether heart‑rate variability (HRV) derived from a single‑channel electrocardiogram (ECG) can serve as a low‑cost, non‑invasive biomarker of in‑utero exposure to Zika virus (ZIKV) in infants who appear asymptomatic at birth. Twenty‑one Brazilian infants aged 4–25 months were recruited: 13 had confirmed maternal ZIKV infection during pregnancy (two of these presented with severe microcephaly) and 8 served as controls with other diagnoses (including congenital infections, a genetic syndrome, small‑for‑gestational‑age, and two healthy infants).

Each child underwent a ~15‑minute resting ECG recording using a portable Zephyr device while being held upright by the mother. After automatic R‑peak detection and visual quality control, noisy segments were discarded, leaving on average 15 minutes of usable data per subject. HRV was quantified using a moving 5‑minute window (2.5‑minute overlap), generating 54 time‑domain, frequency‑domain, and non‑linear metrics for each window; these were then averaged across windows to produce a single value per metric per infant.

A cubic support‑vector‑machine (SVM) classifier was trained and evaluated with 100 repetitions of 5‑fold cross‑validation. Because of the limited sample size, only a single HRV metric was selected as the classifier input. The metric that yielded the highest discriminative performance was “Grid Count,” a complexity measure derived from a grid‑based transformation of the phase‑space trajectory of the R‑R interval series. Using Grid Count alone, the classifier achieved an area under the receiver‑operating‑characteristic curve (AUC) of 94.5 %, sensitivity of 85.7 %, specificity of 92.3 %, a negative predictive value of 92 %, and a positive predictive value of 86 %. Excluding the two microcephalic infants from the ZIKV group did not materially alter these results.

The authors interpret the findings in the context of chronic fetal hypoxia, a known consequence of ZIKV placental infection. Experimental work in fetal sheep suggests that chronic hypoxia enhances sympathetic drive, leading to increased synchrony among sino‑atrial node pacemaker cells and a reduction in HRV complexity. Post‑natally, a withdrawal of this heightened sympathetic input may manifest as a rebound increase in complexity, captured by a higher Grid Count. Thus, Grid Count may reflect a “memory” of in‑utero hypoxic stress that persists after birth.

Limitations include the small cohort, age heterogeneity (which could confound HRV maturation effects), and variability in the gestational timing of maternal ZIKV symptoms. The study also relied on a single HRV feature rather than multivariate patterns, which may limit robustness. Future work should involve larger, age‑matched cohorts, stratification by trimester of exposure, and the exploration of multivariate HRV signatures combined with other physiological data.

In conclusion, the study demonstrates the feasibility of acquiring reliable HRV data from toddlers using a portable ECG device and provides preliminary evidence that HRV, particularly the Grid Count metric, can differentiate asymptomatic ZIKV‑exposed infants from unexposed controls. If validated in larger populations, this approach could become a scalable screening tool for identifying children at risk of subtle, long‑term sequelae of congenital Zika infection.