This article is more than 5 years old. Information may no longer be current.
Breast-feeding reduces diabetes incidence in mothers, offspring
Nonindigenous Canadian mothers who initiated breast-feeding in the hospital reduced their risk for diabetes by 23%, whereas their offspring experienced an 18% risk reduction for developing the disease, according to study findings presented at the World Diabetes Congress.
In a population-based analysis of more than 330,000 live births during a 24-year period in Manitoba, researchers also found that indigenous women saw a 14% risk reduction for diabetes when they initiated breast-feeding.
“Every mother should try to initiate breast-feeding if she can, no matter how long the duration, to receive beneficial effects ... including reducing the risk for diabetes,” Garry X. Shen, MD, PhD, professor of internal medicine and human nutritional sciences co-chair, Endocrine Research Group, University of Manitoba, told Endocrine Today. “This is one of the most cost-effective ways to reduce type 2 diabetes in women and their offspring.”
Shen and colleagues analyzed hospital records from 334,553 deliveries (60,088 First Nation births) taking place in Manitoba between 1987 and 2011. Within the cohort, 262,124 women initiated breast-feeding in the hospital (mean age, 28 years; 13% First Nations; 2.9% with gestational diabetes), and 72,429 did not initiate breast-feeding (mean age, 26 years; 36% First Nations; 4.2% with gestational diabetes), according to hospital records that note breast-feeding initiation at discharge. Researchers also determined diabetes status through hospital records.
Breast-feeding initiation was associated with significantly lower incidence of diabetes in mothers and offspring without a history of gestational diabetes, although First Nations women maintained a higher risk than non-First Nations women (HR = 0.888 vs. 0.745). Women with a history of gestational diabetes also saw a reduced risk for subsequent diabetes, with First Nations women again maintaining a higher risk vs. non-First Nations women (HR = 0.826 vs. 0.821).
Researchers found that First Nations status did not significantly affect the association between breast-feeding and incident diabetes for offspring and, therefore, pooled analyses, finding that children of mothers who initiated breast-feeding had an 18% reduced risk for diabetes (adjusted HR = 0.83; 95% CI, 0.69-0.99).
The protective effects of breast-feeding observed in mothers and children were independent of First Nations status, gestational diabetes, gestational hypertension, maternal age, birth weight and socioeconomic factors, according to researchers.
“We need to determine the impact of duration and exclusive/partial breast-feeding on diabetes and obesity,” Shen told Endocrine Today. “We plan to conduct a long-term, prospective study on breast-feeding and the development of diabetes in mothers and offspring.” – by Regina Schaffer
Reference:
Martens PJ, et al. Abstract #0511. Presented at: World Diabetes Congress; Nov. 30-Dec. 4, 2015; Vancouver, British Columbia.
Disclosure: The researchers report no relevant financial disclosures.
Perspective
Back to Top
Breast-feeding initiation in the hospital may be influenced by the health status of the mother and baby after delivery, including the severity of glucose intolerance, maternal obesity, and/or adverse delivery and newborn health outcomes, such as cesarean delivery, preterm birth or NICU admission. The health conditions and inpatient care practices may interfere with the initiation of breast-feeding as well as affect future diabetes risk. For example, this study was unable to account for reverse causation (ie, more favorable metabolic profiles lead to greater rates of breast-feeding initiation) and could not rule out the possibility that pre-existing metabolic profiles reduced future diabetes risk in women rather than breast-feeding initiation. The study also did not address major confounders of breast-feeding and diabetes risk in women and children, such as maternal obesity, and adverse maternal and infant perinatal outcomes (eg, cesarean delivery, gestational age at birth, preterm birth) and lifestyle behaviors that influence breast-feeding success.
Very large population-based databases that rely solely on hospital discharge records and diagnostic codes have limitations for diabetes studies because of the lack of detailed measures of gestational diabetes severity, maternal body size and other clinical parameters of metabolic risk, and may be characterized by detection bias due to higher screening rates for diabetes among high-risk women. Large sample sizes may provide statistically significant results, but unless sources of bias that affect breast-feeding are carefully evaluated, they may not provide valid estimates of the true association. Given the long study period, 1987 to 2011, there are likely to be strong secular trends affecting not only breast-feeding initiation rates, but screening for gestational and type 2 diabetes in the population, but these trends were not addressed in this study.
Our recently published study (Gunderson EP, et al. Ann Intern Med. 2015;doi:10.7326/M15-0807) utilized a carefully designed study to address specific potential sources of bias. The SWIFT study is a large prospective cohort of more than 1,000 women with gestational diabetes that assessed pregnancy clinical characteristics, including prepregnancy weight and metabolic status, perinatal outcomes and lifestyle behaviors, and conducted in-person research quality assessments of breast-feeding and screened women from 2 months postpartum and annually thereafter via the 2-hour, 75-g oral glucose tolerance test. This rigorous research study design evaluated the impact of breast-feeding intensity and duration on progression to diabetes after pregnancy and was able to minimize both reverse causation and confounding based on the methodologies for detailed collection of clinical, biochemical, behavioral and other risk factor data.