Infant sleep is one of the most closely watched indicators of early neurodevelopment—and one of the hardest to measure reliably. Total night sleep, longest continuous sleep stretch, and frequency of night wakings are tracked in studies worldwide, yet comparisons across that body of work have been hampered by inconsistent methods, small samples, and sparse longitudinal data. A new mini-scoping review published in Frontiers in Neuroscience takes a systematic look at what two decades of research using nightly measurements actually shows.
The study authors searched PubMed for studies published between 2000 and 2024 that used actigraphy, sleep diaries, or both to measure three core metrics during the first six months of life: total night sleep (TNS), longest sleep stretch (LSS), and night wakings (NW).
Of 800 studies initially identified, 35 met inclusion criteria. The authors aggregated data across studies and fit regression curves to characterize developmental trajectories for each metric.
What They Found
Across both methods, the broad trajectory of infant sleep consolidation was consistent: Total night sleep increases over the first several months, the longest uninterrupted stretch grows longer, and night wakings decrease.
For total night sleep, actigraphy and diary studies showed similar curves: Starting around 7–8 hours per night in the first weeks of life and converging to roughly 9.5 hours by week 24.
The picture became considerably messier for the other two metrics.
For longest sleep stretch, actigraphy-based studies (which derive sleep from infant movement) reported roughly 4 hours at 24 weeks, while studies using caregiver diaries reported approximately 6.5 hours at the same age, a meaningful gap the authors attribute largely to how each method handles brief arousals and movement during sleep.
Night wakings showed the sharpest divergence of all: Diary studies reported a gradual decline from about 3 to 2 wakings per night, while actigraphy studies in some cases reported 10 or more wakings per night in early infancy. This illustrates the sensitivity of findings to experimenter-defined movement thresholds, and the care required to interpret and compare such results to studies using other methods.
Compounding these method-level differences was a structural problem in the literature itself. Of the 35 included studies, 28 reported data at only one or two time points, with sample sizes ranging from just 13 to 320 infants. That sparse sampling makes it difficult to describe developmental trajectories with confidence, let alone establish population norms.
Why It Matters
The field lacks reliable benchmarks for normal infant sleep development—benchmarks that are foundational for identifying sleep disturbance, evaluating interventions, and understanding the downstream effects on infant neurodevelopment and caregiver mental health.
This review provides a practical reference for the approximate range of TNS, LSS, and NW values researchers can expect at different ages, while making explicit the methodological context that any such values carry with them.
The authors call for greater standardization across future studies: consistent operational definitions of “nighttime,” larger and more diverse cohorts, more frequent longitudinal sampling, and transparent reporting of actigraphy parameters. They also point toward emerging technologies, as promising avenues for improving both the objectivity and resolution of infant sleep data, including video-based sleep monitoring and continuous physiological sensors.
The full review is open access in Frontiers in Neuroscience.