Trust in vaccines and in those who deliver them is often shaped by individuals' social networks. Social networks determine exposure to information, norms, and reinforcement processes that influence vaccine decision‑making. The authors investigated how the composition and content of personal networks among young people in rural KwaZulu‑Natal, South Africa, related to willingness to receive a COVID‑19 vaccine.
The analysis used data collected in 2022–23 from the Sixhumene social network study conducted primarily among young adults in three rural communities in KwaZulu‑Natal. The study assessed individual and interpersonal factors alongside network characteristics to evaluate associations with vaccine willingness.
Respondents were predominantly female with a reported median age of 21 years (interquartile range 18–28). Participants identified a median of three key social contacts (alters). The social networks were characterised by strong kinship connections and a high degree of local ties. Network variables examined included network size, kinship composition, geographic proximity of alters (household, same village, outside village), gender homophily, and measures of support received (emotional and physical), quantified in contact‑days.
The primary outcome was a six‑level ordinal variable combining respondents' past vaccination status with their hypothetical willingness to vaccinate. To evaluate associations between network characteristics and the ordinal vaccine willingness outcome, the authors used multivariable ordinal logistic regression. Models adjusted for respondent characteristics alongside the network measures under study.
The sample skewed female and young (median age 21). Respondents reported small, kin‑centred networks with most contacts local to the respondent’s household or village. Median network size was three alters.
Vaccine willingness varied by the geographic composition of respondents' networks. Among urban respondents, those whose networks included alters in the same village had higher willingness compared with urban respondents whose networks were concentrated within the household (adjusted odds ratio [aOR] 2.54, 95% CI 2.40–2.70). Urban respondents with alters outside the village had still higher willingness (aOR 3.26, 95% CI 2.04–5.20).
Among rural respondents, compared with the reference network composition, willingness was also higher when networks were household‑based (aOR 1.95, 95% CI 1.57–2.41), village‑based (aOR 1.63, 95% CI 1.40–1.90), or extended beyond the village (aOR 1.64, 95% CI 1.50–1.79). These associations indicate that geographic dispersion and the presence of contacts beyond the immediate household or village were associated with greater vaccine willingness in both urban and rural respondent groups.
Different types of support from network contacts showed divergent associations with vaccine willingness. Emotional support — measured as contact‑days of emotional support — was positively associated with willingness; specifically, each 30 contact‑day increase in emotional support was associated with higher odds of greater willingness (aOR 1.15, 95% CI 1.01–1.32).
In contrast, physical support had an inverse association with willingness: higher levels of physical support from network members were associated with lower odds of greater vaccine willingness (aOR 0.84, 95% CI 0.72–0.98).
The study shows that both individual factors and network dynamics influence vaccination willingness in this rural South African setting. Network geographic composition — the extent to which alters are in the household, village, or beyond — was consistently associated with willingness across respondent groups. The positive association with emotional support suggests that supportive social ties that provide reassurance or encouragement may promote vaccine willingness. The inverse relationship with physical support could reflect complex social or practical dynamics in which reliance on material or instrumental help associates with different risk perceptions or constraints; the source does not provide further causal explanation.
The abstract reports methods and associations but does not detail all potential limitations such as measurement specifics, representativeness beyond the three communities, or potential residual confounding; those details were not reported in the source abstract. The authors declared no known competing financial interests or personal relationships that could have influenced the work.
In these rural KwaZulu‑Natal communities, small kin‑centred networks predominated, and network composition and the type of support received were associated with an ordinal measure of vaccine willingness. Emotional support showed a modest positive association with willingness, while physical support was inversely associated. These findings suggest that leveraging emotionally supportive ties within personal networks may help promote vaccine acceptance in similar rural settings.