Abstract
Background
Previous studies on dairy intake and cardiovascular disease (CVD) risk reported heterogeneous associations for dairy types. While guidelines recommend low-fat dairy to reduce saturated fatty acid intake, emerging evidence suggests that health effects depend on the dairy type and its food matrix rather than on fat content alone. Some studies found beneficial associations for fermented dairy, while others observed neutral or adverse associations for other dairy types. The inconsistencies in findings may partly result from differences in unspecified dairy substitutions and the background diet. While previous studies have examined dairy-to-non-dairy substitutions, within-dairy substitution in relation to CVD or coronary heart disease (CHD) are lacking, highlighting the need to assess both absolute intake and the potential impact of replacing on dairy type for another.
Purpose
To examine the associations between total dairy intake and its subtypes with incident CVD and CHD and to evaluate the impact of substituting dairy types on disease risk.
Methods
We analysed data from the European Prospective Investigation into Cancer and Nutrition in the Netherlands (EPIC-NL) cohort including 30,217 participants (median age 51.9 y [interquartile range 43.1-58.3], 22.2% male). Baseline dietary intake was assessed using a validated 178-item Food Frequency Questionnaire. Cox proportional hazards models computed hazard ratios (HRs) and 95% confidence intervals for dairy intake (total, fermented, milk, yoghurt, cheese, cream, ice cream – and their high- and low-fat variants) and CVD and CHD incidence in servings/day and quartiles, adjusting for demographic, energy, lifestyle, dietary, and cardiometabolic factors. The leave-one-out substitution method was used to assess the theoretical replacement among dairy types in serving/day.
Results
During a median follow-up of 14.8 years, 3,658 CVD cases (12.1%) and 2,011 CHD cases (6.7%) were recorded. Higher intake of high-fat yoghurt was associated with a lower risk of CHD (HRserving/day 0.52, 0.28-0.97; HRQ4vs.Q1 0.87, 0.76-0.99) and CVD (HRQ4vs.Q1 0.90, 0.81-0.99). In contrast, a higher intake of low-fat cheese was associated with a higher risk of CVD (HRserving/day 1.12, 1.02-1.22). Substituting low-fat cheese with high-fat yoghurt, low-fat yoghurt, high-fat milk, or high-fat cheese was associated with lower CVD risk (HRs 0.61-0.89). Substituting low-fat cheese with high-fat yoghurt was also associated with lower CHD risk (HRserving/day 0.51, 0.27-0.96).
Conclusion
EPIC-NL findings support the idea that the health effects of dairy depend on its type rather than fat content alone and reveal novel results regarding within-dairy substitutions in relation to CVD and CHD. High-fat yoghurt intake was inversely associated with the risk of CHD and CVD, whereas low-fat cheese intake was associated with a higher CVD risk. Substituting low-fat cheese with yoghurt and other dairy types may lower CVD risk.
Previous studies on dairy intake and cardiovascular disease (CVD) risk reported heterogeneous associations for dairy types. While guidelines recommend low-fat dairy to reduce saturated fatty acid intake, emerging evidence suggests that health effects depend on the dairy type and its food matrix rather than on fat content alone. Some studies found beneficial associations for fermented dairy, while others observed neutral or adverse associations for other dairy types. The inconsistencies in findings may partly result from differences in unspecified dairy substitutions and the background diet. While previous studies have examined dairy-to-non-dairy substitutions, within-dairy substitution in relation to CVD or coronary heart disease (CHD) are lacking, highlighting the need to assess both absolute intake and the potential impact of replacing on dairy type for another.
Purpose
To examine the associations between total dairy intake and its subtypes with incident CVD and CHD and to evaluate the impact of substituting dairy types on disease risk.
Methods
We analysed data from the European Prospective Investigation into Cancer and Nutrition in the Netherlands (EPIC-NL) cohort including 30,217 participants (median age 51.9 y [interquartile range 43.1-58.3], 22.2% male). Baseline dietary intake was assessed using a validated 178-item Food Frequency Questionnaire. Cox proportional hazards models computed hazard ratios (HRs) and 95% confidence intervals for dairy intake (total, fermented, milk, yoghurt, cheese, cream, ice cream – and their high- and low-fat variants) and CVD and CHD incidence in servings/day and quartiles, adjusting for demographic, energy, lifestyle, dietary, and cardiometabolic factors. The leave-one-out substitution method was used to assess the theoretical replacement among dairy types in serving/day.
Results
During a median follow-up of 14.8 years, 3,658 CVD cases (12.1%) and 2,011 CHD cases (6.7%) were recorded. Higher intake of high-fat yoghurt was associated with a lower risk of CHD (HRserving/day 0.52, 0.28-0.97; HRQ4vs.Q1 0.87, 0.76-0.99) and CVD (HRQ4vs.Q1 0.90, 0.81-0.99). In contrast, a higher intake of low-fat cheese was associated with a higher risk of CVD (HRserving/day 1.12, 1.02-1.22). Substituting low-fat cheese with high-fat yoghurt, low-fat yoghurt, high-fat milk, or high-fat cheese was associated with lower CVD risk (HRs 0.61-0.89). Substituting low-fat cheese with high-fat yoghurt was also associated with lower CHD risk (HRserving/day 0.51, 0.27-0.96).
Conclusion
EPIC-NL findings support the idea that the health effects of dairy depend on its type rather than fat content alone and reveal novel results regarding within-dairy substitutions in relation to CVD and CHD. High-fat yoghurt intake was inversely associated with the risk of CHD and CVD, whereas low-fat cheese intake was associated with a higher CVD risk. Substituting low-fat cheese with yoghurt and other dairy types may lower CVD risk.
| Original language | English |
|---|---|
| Article number | ehaf784.3857 |
| Journal | European Heart Journal: The Journal of the European Society of Cardiology |
| Volume | 46 |
| Issue number | 1 |
| DOIs | |
| Publication status | Published - Nov 2025 |
