Our study indicated the maternal feeding anxiety was prevalent in China and 61.4% of women at 0–3 months postpartum had feeding anxiety symptoms, with 8.6% being severe anxiety. Maternal postpartum feeding anxiety was associated with negative infant feeding practices, like bottle feeding and infant food refusal behaviors. The results showed maternal mental health may be an important factor affecting child nutrition and health. Negative emotions during feeding may affect the maternal cerebral cortex, hypothalamus, and pituitary function, decrease secretion of prolactin and oxytocin, eventually reduce the breast milk secretion [6, 9].
Pregnancy and therefore lactation are important life events for women, during which their physiological and psychological state, as well as social status undergo considerable changes. The immediate change in life roles and new responsibilities after childbirth may be pressure provoking. Therefore, postpartum women are susceptible to anxiety, fear, and even depression . Perinatal anxiety and depression are most common mental health problems in women. A systematic review summarized the prevalence and determinants of non-psychotic common perinatal mental disorders (CPMDs) in World Bank categorized low- and lower-middle-income countries, showed weighted mean prevalence was 15.6% (95% CI: 15.4–15.9) antenatally and 19.8% (19.5–20.0) postnatally, and the pooled prevalence of perinatal depression ranged from 5.2 to 32.9% during pregnancy and 4.9–59.4% after child birth . In China the prevalence of postnatal anxiety or depression varies from 2.2 to 43.6% [11,12,13,14]. A systematic review of postpartum depression prevalence published during 1996–2012 showed the pooled prevalence was 14.7% (13.1–16.3%), increasing from the east of China to the west .
Given postpartum anxiety has been noted as having independent effects just as postpartum depression, and significant comorbidity has also been noted between postpartum anxiety and depression, it is surprising that so little research has been conducted on postpartum anxiety, compared with postpartum depression. A narrative review showed  the prevalence of postpartum anxiety (ranged from 13 to 40%) has varied according to the definition, the anxiety scale used, the cut-off scores on the scales, the severity of anxiety, the timing of the assessment, the recruitment sample and the origin (country) of research. In this study, the Self-rating Feeding Anxiety Scale (SFAS) was used to assess the maternal postpartum anxiety, which was developed specifically to tap postnatal-specific anxiety-feeding anxiety, and reliability and validity has been tested . Besides assessment tools, the cultural and socio-economic settings, sample sizes, cutoff point used, and timing of assessment affect prevalence. In this study the maternal feeding anxiety was assessed at 0 ~ 3 months postpartum, which was during the Chinese traditional “doing the month” period, both limited socialization and family support could worsen maternal mental problems and lead to higher anxiety symptom prevalence. In fact, the severe FA prevalence was only 8.6% (FAS>52), that was similar to other studies. In order to achieve reliable results globally, there is a need to establish widely accepted assessment tools, cutoff scores, and timing of assessment.
Systematic review showed the correlates factors for postpartum anxiety included demographic factors, childbirth experiences, social support and history of psychiatric and psychological problems . The demographic risk factors for postnatal anxiety include being a young mother, having more education and being employed. Childbirth risk factors include being primiparous in one sample and multiparous in another, caesarean delivery, fear of the birth and of death during delivery, lack of control during labor, low self-confidence for the delivery and the delivery staff, and premature delivery. Social support problems include the lack of family support, marital/family conflict, and social health issues. Psychiatric history risk factors include prenatal depression and anxiety. Several studies have concluded that preterm birth and cesarean section are independent risk factors of postpartum anxiety [16,17,18] We found a higher FAS in women with younger age, lower household income, first-time mother and preterm birth, but without statistical significance (P > 0.05), and FA prevalence was associated with delivery mode (P < 0.05). Trumello et al.  found that mother-child separation caused by the need for specialized care of premature infants may exert a negative effect on the mothers’ emotional state; thus, women who give birth to premature infants are more likely to develop postpartum anxiety. A Chinese study  found that postpartum women aged ≥35 were more prone to anxiety or depression. Ye et al. (2014)  reported a higher prevalence of postpartum anxiety among mothers with higher education levels. Zhang et al.  found that women were more vulnerable to feeding anxiety with lower household income. Bina R et al. (2017)  found that postpartum women with lower incomes reported more symptoms of anxiety. And another study showed that postpartum women’s perceived difficulty managing on household income was associated with anxiety symptomatology . Other studies have suggested that social psychological factors, such as emotional support from family and society, and various expectations of the sex of the newborn also considerably affected the mother’s feelings during feeding [11, 24]. Because of the limited data we haven’t analyzed the relations of FA with family support and history of psychiatric and psychological problems, that is the study limitation.
Postpartum anxiety are always associated with unpleasant emotional experiences, such as worry, fear, irritability, and frustration, that can last for a few weeks or even longer . Maternal anxiety during feeding may eradicate any willingness to breastfed and even affect the composition of the breast milk, ultimately affecting the growth and development of infant. The systematic literature review included negative effects on breast-feeding, bonding, mother–infant interactions, infant temperament, sleep, mental development, health and internalizing in infants and on conduct disorder in adolescents, based on structured clinical interviews and behavior observations .
Our study indicated higher FA risk seemed to be found in women who did not feed infant colostrum or fed infant with bottle. This may be explained by that mothers who fed infant colostrum could generally let baby touch breast in mother’s arms within 1 h after delivery, which promoted emotional communication between mother and infant and helped breastfeeding. However bottle feeding might reduce mother-child contact, and hinder emotional communication during feeding. Furthermore, bottle feeding led to insufficient breast milk secretion and subsequent interruption of breastfeeding, which diminished maternal confidence in infant feeding. On the other hand, bottle feeding might help mothers control the milk intake of infants, and mothers with FA may prefer bottle feeding to breastfeeding. Lower FA prevalence was found in women who breastfed exclusively but without statistical significance. These findings are supported by two studies, in which mothers who breastfed are less likely to develop FA, as compared to mothers who chose formula feeding, and mothers suffered from FA does not emerge favorable persistence signs associated with breastfeeding [26, 27].
Moreover, mothers also had a significantly higher FAS if their babies had been exposed to sunshine outdoors. A possible explanation may be that mothers with FA might be more cautious about the infant’s health so they spent more time with their infants outdoors to synthesize vitamin D. Infant food refusal behaviors may reduce maternal confidence, and induce mother’s anxiety about infants’ nutrient intake and development. But above results needed to be confirmed by more research in the future, because the limitations of sample size and sampling method decreased the power of test.
Despite our results and the mounting evidence indicating high prevalence of maternal mental health and its adverse impact on infant feeding practices, the maternal mental health problems has not been focused. Consequently the treatment gap for mental illness is large, accounting for 76–85% patients with mental health not receiving intervention. Studies show that psychosocial, educational, and supportive interventions are effective in improving maternal mental health. That will be our next research direction, to explore the effective intervention strategy to improve maternal mental health and early child development. And theoretically the anesthesia type during delivery would have an influence on breastfeeding, but regrettably the related information has not been collected in this study. In the future, if possible we will investigate the data retrospectively and compensatively in the other relevant investigation.
Several limitations in our study deserved acknowledgements, like convenient sampling, small sample size, unpopular assessment tool, related factors (family support) data deficiency, etc.