Enclosed neighborhood streets are linked to better sleep, study finds

A study of employed Japanese adults found that greater enclosure in neighborhood landscapes was associated with less severe insomnia symptoms, less daytime sleepiness, and longer sleep. The study authors report that this may operate through higher perceived safety and liveliness of an area. The paper was published in Building and Environment.

Sleep occupies roughly one-third of our lives and is essential for physical recovery, mental health, and normal social functioning. However, sleep problems are increasingly recognized as a major public health concern and are associated with conditions ranging from cardiovascular and metabolic disease to depression and dementia. At the same time, sleep is influenced not only by individual habits and biology but also by the environments in which people live.

Features of urban neighborhoods such as greenery, noise, air pollution, street lighting, and overall community quality have all been linked to sleep health. What people actually see at street level may be particularly important because the visual environment can influence stress, perceived safety, mood, and other psychological processes.

Research has found, for example, that visible greenery and street trees can be associated with sleep quality, although the effects are not always straightforward or consistent. Negative neighborhood features such as litter and broken windows may also contribute to poorer sleep by reducing feelings of safety and increasing environmental stress. Traditionally, these environmental characteristics have been measured through surveys, expert assessments, satellite imagery, or labor-intensive street audits. However, newer approaches use computer vision to make more objective assessments and to make much more extensive evaluations of environmental characteristics.

Study author Xiaorui Wang and their colleagues conducted a study in which they examined the associations between objective physical and subjective perceptual qualities of landscapes at eye level and sleep health. Their study was conducted in the Greater Tokyo Area of Japan, a megacity region characterized by high population density and extensive urban development surrounding the Tokyo Metropolis.

The study authors initially recruited 2000 employed adults to be participants in their study. However, after excluding those living above the third floor and participants whose sleep characteristics or other measured properties were very unusual, the analyses were performed on 1089 individuals.

The study authors report that while participants living on the third floor or below were similar in age and education to those living above the third floor (who were excluded), they differed significantly in gender, household income, and parental status. Participants were 40-59 years of age.

The study authors assessed characteristics of participants’ neighborhood eye-level visual environment using street-view images linked to residential postal-code areas. They sampled street intersections within each neighborhood and collected images in four directions. In total, they analyzed more than 214,000 street-view images.

The study authors used a procedure called computer-vision semantic segmentation—an artificial intelligence technique that categorizes individual pixels in an image to detect specific objects—to identify visible elements such as trees, roads, sidewalks, buildings, and fences. From these, they derived measures of the environment – greenness, walkability, imageability, enclosure and complexity. In parallel, they used the methodology developed by the Place Pulse project to estimate how these landscapes would be perceived in terms of safety, beauty, liveliness, wealthiness, lack of boredom, and lack of depressiveness.

The study participants self-reported their bedtime and wake time over the previous month, and completed assessments of daytime sleepiness (the 8-item Japanese version of the Epworth Sleepiness Scale), and insomnia symptoms (the Athens Insomnia Scale). The study authors calculated participants’ sleep duration from the bedtime and wake times they reported. They then integrated this information with averaged assessments of their neighborhood characteristics derived from the analyses of street-view images.

The results showed that participants who lived in areas with greater enclosure tended to have less severe insomnia symptoms, experience less daytime sleepiness, and sleep longer. Enclosure was a measure of how strongly buildings, trees, and other vertical elements visually bounded the street space relative to more open elements such as roads, sidewalks, earth, and grass. Further analyses indicated that this relationship may operate through higher perceived safety and liveliness of a neighborhood. Greenness was associated with higher perceived safety and, in turn, longer sleep. On the other hand, walkability was associated with lower perceived safety and, in turn, shorter sleep.

Greenness was the proportion of the streetscape occupied by visible vegetation such as trees, plants, flowers, palms, and grass. Safety was an estimate of how safe the neighborhood streetscape would generally be perceived to look. Walkability was a measure based mainly on the amount of sidewalk and fencing relative to road area, intended to capture how supportive the immediate street environment was for walking. Liveliness was an estimate of how lively or active a neighborhood streetscape would generally be perceived to look.

“Together, these findings broaden explanations for how urban environments relate to sleep health and provide actionable candidate indicators for future causal and intervention research on urban design that supports sleep,” the study authors concluded.

The study sheds light on the links between a person’s physical environment and sleep quality. However, it should be noted that the study participants were all employed, middle-aged adults living on the lower floors of a single metropolitan area in Japan. Findings may not generalize to residents of rural areas, other urban areas, or individuals from other cultures, age groups, employment statuses, and socioeconomic backgrounds.

The paper, “Urban eye-level landscapes and sleep health: Cross-sectional evidence from a megacity using machine learning for objective and perceptual assessment,” was authored by Xiaorui Wang, Jihui Yuan, Weixin Huang, and Daisuke Matsushita.