New research supports the universality of maternal sensitivity in shaping child attachment

A new study published in Child Development provides the strongest longitudinal evidence to date that sensitive caregiving in early life predicts children’s attachment security in China, a country with cultural traditions and parenting styles often seen as different from those in the West. Drawing from a nine-year study of over 350 mother-child pairs in Nanjing, researchers found that maternal sensitivity in infancy was associated with children’s secure attachment representations at age 10. This finding lends support to the idea that maternal sensitivity plays a universal role in shaping attachment.

Attachment theory argues that children form expectations about caregivers based on their early experiences of support, and these expectations—called internal working models—go on to influence social and emotional development. A central idea in the theory is the “sensitivity hypothesis,” which states that responsive and emotionally attuned caregiving fosters secure attachments. But critics have long questioned whether this claim holds true in cultures with different parenting practices and values, particularly in East Asian societies like China, where overt expressions of warmth may be less common and autonomy is not always emphasized.

The research team, led by Theodore E. A. Waters and Rui Yang from New York University, set out to test whether the link between early maternal sensitivity and later attachment security would still hold in this distinct cultural context. To do this, they used data from the Nanjing Child Development Project, a long-term study tracking parent-child relationships from infancy to middle childhood.

The study followed 356 Chinese mother-child pairs from when the children were 14 months old until they were 10 years old. The mothers, mostly well-educated and from middle- to upper-middle-class families, participated in lab-based sessions where researchers observed their interactions with their children. These sessions were videotaped when the children were 14 and 24 months old. Trained observers rated each mother’s sensitivity based on how responsive, warm, and non-intrusive she was during play and structured tasks like reading and bead threading.

When the children turned 10, they completed a Chinese-adapted version of the Attachment Script Assessment. This task measures a child’s secure base script knowledge, or how well they can tell stories that reflect the expectation that caregivers provide help when needed. Children were given story outlines involving common situations, such as getting lost in a mall or preparing for an exam, and asked to tell first-person stories involving a parent. Coders rated these stories based on how clearly they reflected a secure pattern—asking for help, receiving support, resolving distress, and returning to normal.

The researchers also collected background information, including maternal education levels and the child’s sex, to account for other possible influences on attachment development.

The results showed a clear and consistent pattern: the more sensitive a mother was during her child’s infancy and toddlerhood, the more secure the child’s attachment representations were at age 10. Children whose mothers were warm, responsive, and not overly controlling during early interactions were more likely to tell stories that reflected a secure attachment script. This link held even after accounting for the mother’s education level and the child’s sex.

The effect size of this relationship (β = 0.20) was in line with findings from similar studies in Western countries, suggesting a comparable strength of association. Girls tended to score higher on the attachment measure than boys, and children with more educated mothers also scored higher, possibly reflecting the influence of language exposure or other cognitive supports.

Importantly, the Chinese version of the Attachment Script Assessment showed strong internal consistency and fit well with models used in Western research. The researchers found that the measure tapped into a generalized sense of attachment security that was not limited to relationships with a particular parent, mirroring findings from previous studies.

While the findings support the universality of the sensitivity hypothesis, the study had some limitations. First, the observations of maternal sensitivity only occurred in low-stress situations like free play and structured tasks. Because attachment theory emphasizes how caregivers respond during distress, future research should also observe parent-child interactions in more emotionally intense situations.

Second, the research only included maternal caregiving. Although mothers are often the primary caregivers in China, other adults—especially fathers and grandparents—also play important roles in children’s development. Future work could explore how sensitivity from different caregivers contributes to attachment security.

Nevertheless, the study offers evidence that the link between maternal sensitivity and children’s attachment security extends beyond the Western world. Even in a cultural context like China—where parenting practices are shaped by different values around authority, emotion, and independence—early sensitivity from caregivers still predicted the formation of secure attachment expectations.

The study, “Maternal Sensitivity Predicts Child Attachment in a Non-Western Context: A 9-Year Longitudinal Study of Chinese Families,” was authored by Theodore E. A. Waters, Rui Yang, Yufei Gu, Victoria Zhu, Lixian Cui, Xuan Li, Niobe Way, Hirokazu Yoshikawa, Xinyin Chen, Sumie Okazaki, Kristen Bernard, Guangzhen Zhang, and Zongbao Liang.