Higher consumption of dark chocolate, but not milk chocolate, was associated with a lower risk of type 2 diabetes in a new study.
In a large prospective study published in BMJ, investigators analyzed the data from three cohort studies with 192,208 participants followed for up to 35 years, accumulating over 4.8 million person-years of follow-up. The cohorts included the 1986 to 2018 Nurses' Health Study (NHS), 1991 to 2021 Nurses' Health Study II (NHSII), and 1986 to 2020 Health Professionals Follow-Up Study (HPFS).
Participants consuming five or more servings per week of dark chocolate showed a 21% lower risk of developing type 2 diabetes (T2D) compared with rare consumption (95% confidence interval [CI] = 5%–34%). The study demonstrated a linear dose-response relationship, with each additional weekly serving of dark chocolate associated with a 3% reduction in T2D risk (95% CI = 1%–5%).
No significant associations were found between milk chocolate consumption and diabetes risk. The validation of dietary questionnaires showed correlation coefficients of 0.53 to 0.65 for dark chocolate and 0.42 to 0.47 for milk chocolate when compared with 7-day diet records.
The study identified 18,862 incident cases of T2D during follow-up. T2D diagnoses were confirmed through medical record review, with confirmation rates of 98% in NHS and 97% in HPFS.
The protective association of dark chocolate was strongest among participants with higher diet quality scores, showing a 34% lower risk (95% CI = 12%–51%) for those consuming 5 or more servings weekly. The effect was more pronounced in men, with HPFS showing a 51% lower risk (95% CI = 8%–74%) for highest consumption.
Weight change analyses revealed that increased milk chocolate consumption was associated with greater weight gain over 4-year periods: 0.68 kg (95% CI = 0.42–0.95) in patients with obesity and 0.33 kg (95% CI = 0.24–0.42) in those with normal body mass index (BMI). Dark chocolate consumption showed no significant association with weight gain.
The study controlled for multiple lifestyle and dietary factors, including age, ethnicity, smoking status, alcohol intake, physical activity, BMI, and overall diet quality. The findings persisted after adjusting for socioeconomic status and educational levels.
The researchers noted that their study population primarily consisted of non-Hispanic White health professionals over 50 years of age, which may limit generalizability. They concluded that randomized controlled trials are needed to confirm these findings and explore mechanisms.
Potential conflict of interest disclosures can be found in the study.