Community singing intervention reduces maternal salivary cytokines and enhances maternal-infant cortisol synchrony in postnatal depression: biological findings from the SHAPER-PND trial

The arts, specifically music interventions, have emerged as a promising therapeutic strategy for postnatal depression (PND). The Scaling-up Health-Arts Programmes: Implementation and Effectiveness Research-Postnatal Depression (SHAPER-PND) randomized controlled trial showed that Breathe Melodies for Mums (M4M), a 10-week community singing program for PND developed by the not-for-profit Breathe Arts Health Research, is both effective and feasible compared with an active control of community-based, non-musical mother-baby activities. To validate its scientific basis and refine its application, it is important to examine the biological mechanisms underlying these effects. This study investigated salivary cytokines and cortisol in mothers and their babies taking part in SHAPER-PND. Salivary samples (passive drool for cytokines, swabs for cortisol) were collected from the mother before and after the first (week 1) 1-hour M4M singing (or control) session. Additional samples were collected for cortisol: before and after the last (week 10) M4M singing (or control) session; and, at home, to assess awakening and diurnal cortisol rhythms in both mothers and infants, before and after the whole 10-week M4M singing (or control) intervention. In total, 145 participants (M4M group: 110, control group: 35) contributed data for biological analysis. Inflammatory cytokines were measured using the Meso Scale Discovery (MSD) proinflammatory panel, while cortisol was measured using the Salimetrics enzyme immunoassay kit. Only interleukin-1β (IL-1β), interleukin-6 (IL-6), IL-8 and tumour necrosis factor-alpha (TNF-α), were reliably detected in the saliva. During the week 1 session, IL-6 and TNF-α levels decreased only in the M4M group (IL-6: -26%, p < 0.001; TNF-α: -29%, p = 0.001), but not in the control group (all p values > 0.36), however a significant time x group interaction was seen only for IL-6 (p = 0.026). Cortisol levels fell in both groups at week 1 (control: -41%, p < 0.001, M4M: -22%, p < 0.001), but this effect was sustained only in the M4M group by week 10 (control: +4%, p = 0.19, M4M: -28%, p < 0.001). At home, infant evening cortisol levels decreased from Week 1 to Week 10 in the M4M group (-47%, p = 0.022) but not the control (-25%, p = 0.29). Moreover, the M4M group showed increased mother-infant cortisol synchrony at week 10 (p = 0.017), and closer mother-infant cortisol synchrony was associated with more infant-focused maternal speech (r(21) = -0.43, p = 0.039), less parent focused speech (r(26) = 0.58, p = 0.001), and less inattentive maternal engagement (r(21) = 0.51, p = 0.012)). These results demonstrate that participation in Breathe M4M is associated with acute reductions in stress-related biomarkers and in the regulation of infant stress physiology, which may mirror improvements in mother-infant communication.