Key Takeaways
- A large Swedish cohort study linked high milk intake (3+ glasses/day) to nearly doubled mortality risk and higher fracture rates in women, raising questions about liquid milk’s safety at high doses.
- A systematic review and meta-analysis found higher dairy intake associated with lower vertebral fracture risk and some bone mineral density benefits, though total and hip fracture effects were neutral or inconsistent.
- A 2023 meta-analysis reported small increased prostate cancer risk with higher total dairy and milk intake, but many studies were limited by PSA screening bias and could not prove causation.
- A 2025 global analysis combining major biobanks and meta-data linked total dairy—especially fermented and low-fat types—to modestly lower cardiovascular disease and stroke risks.
The question “Is milk good or bad?” has fueled decades of debate in nutrition science and popular media. For generations, public health messages promoted milk as a cornerstone for strong bones thanks to its calcium, vitamin D (when fortified), and protein. Yet periodic headlines warn that high consumption could raise risks of fractures, certain cancers, or even overall mortality. These contradictions arise not from bad science per se, but from the inherent challenges of nutritional epidemiology: studies often reach different conclusions based on population, dairy subtype (liquid milk versus yogurt or cheese), dosage, methodological rigor, and the specific health outcome examined.
To answer the core question fairly, this article weaves together four key peer-reviewed studies (or meta-analyses) that highlight genuinely conflicting findings. By examining their methods, results, limitations, potential biases, and funding sources side-by-side, we can see why the evidence feels contradictory and what a balanced, evidence-grounded perspective looks like in 2026. All studies are real, with references at the end. The synthesis shows that milk (and dairy more broadly) is neither a miracle food nor a villain; its effects are nuanced, context-dependent, and best understood through moderation and individual factors.
The story often begins with a high-profile challenge to conventional wisdom. In 2014, Swedish researchers published a pair of large prospective cohort studies that sent shockwaves through the dairy world (Michaëlsson et al., 2014). They followed over 61,000 women (aged 39–74 at baseline) for about 20 years and nearly 45,000 men for 11 years, using food-frequency questionnaires (FFQs) to assess milk, fermented dairy, and cheese intake. The standout finding: women drinking three or more glasses of milk per day had a dramatically higher all-cause mortality risk compared to those drinking less than one glass (adjusted hazard ratio [HR] 1.93, 95% CI 1.80–2.06). Each additional glass was linked to a 15% higher mortality risk in women and 3% in men. For fractures, higher milk intake showed no protection and was associated with increased hip fracture risk in women (HR 1.09 per glass). Fermented dairy and cheese, by contrast, trended protective.
The authors proposed a plausible mechanism: D-galactose (a sugar from lactose) might promote oxidative stress and inflammation, supported by biomarker data showing higher oxidative and inflammatory markers in high-milk consumers. This study stood out for its size, long follow-up, and adjustments for many confounders including age, BMI, smoking, physical activity, education, and comorbidities. It directly contradicted the long-held belief that milk builds unbreakable bones.
Yet when researchers step back and pool data across multiple studies, the picture softens. A 2019 systematic review and updated meta-analysis focused on Europeans and non-Hispanic Whites from North America examined dairy’s role in osteoporosis and fractures (Matía-Martín et al., 2019). Pooling prospective cohorts and some case-control studies, they found that highest versus lowest dairy intake showed no statistically significant reduction in total osteoporotic fractures (HR 0.95, 95% CI 0.87–1.03) or hip fractures (HR 0.87, 95% CI 0.75–1.01), with high heterogeneity (I² often >80%). However, a clearer benefit emerged for vertebral fractures (HR 0.82, 95% CI 0.68–0.99). Some analyses also linked higher youthful milk intake to better later-life bone mineral density (BMD) at certain sites, and yogurt/cheese frequently showed stronger positive signals than plain liquid milk.
Comparing these two directly reveals why results diverge. Michaëlsson et al. (2014) was a single-country cohort heavily weighted toward liquid (non-fermented) cow’s milk in a Northern European population with specific genetics, diet patterns, and lower vitamin D fortification in some eras. Matía-Martín et al. (2019) synthesized broader data, capturing varied dairy forms and fortification practices (e.g., higher vitamin D in U.S. milk). Both relied on self-reported FFQs, which are prone to recall bias and measurement error—people inaccurately remember or report intake, and single baseline assessments miss dietary changes over time. Residual confounding remains a shared weakness: high dairy consumers might differ in unmeasured lifestyle factors (socioeconomic status, overall diet quality, physical activity levels, or even frailty). Publication bias and varying study quality can also tilt meta-analyses toward more positive findings.
Neither study was directly industry-funded. Michaëlsson’s work was supported by Swedish public research councils, reducing sponsorship bias concerns. Matía-Martín et al. reported no major conflicts, though broader dairy research sometimes involves industry ties that have been shown in some reviews to correlate with more favorable outcomes. The key lesson: liquid milk at very high intakes may carry different signals than mixed or fermented dairy, and fracture sites matter—hip fractures (more debilitating) often show weaker or null associations compared to vertebral ones.
Cancer concerns add another layer of complexity. A 2023 systematic review and meta-analysis synthesized cohort and case-control studies on dairy products and prostate cancer risk (Zhao et al., 2023). In dose-response analyses, higher total dairy intake (per 400 g/day) was associated with a small increased risk (RR 1.02, 95% CI 1.00–1.03); similar modest elevations appeared for total milk (per 200 g/day) and cheese. Butter showed a slightly stronger link, while whole milk had a slight inverse association in some models. The authors highlighted a common limitation: many studies suffered from PSA screening bias—men who consume more dairy might undergo more frequent prostate-specific antigen testing, leading to overdiagnosis of slow-growing cancers rather than true causal increases.
This finding appears to conflict with bone and heart data, but mechanisms offer possible reconciliation. Dairy can elevate insulin-like growth factor-1 (IGF-1), a hormone linked to cell proliferation that might theoretically promote hormone-sensitive cancers. Yet the absolute risk increase is small, evidence quality is often rated “limited-suggestive” by bodies like the World Cancer Research Fund, and causation remains unproven due to observational design. Again, FFQ inaccuracies, incomplete adjustment for family history, obesity, or physical activity, and heterogeneity across studies weaken strong conclusions. Funding for this meta-analysis was academic with no reported commercial grants, enhancing credibility on bias grounds. When viewed alongside the Swedish mortality data, it suggests potential downsides for very high intakes in men, but not a blanket condemnation.
On the positive side, recent large-scale evidence tilts toward cardiovascular benefits. A 2025 global analysis integrated data from major biobanks (including UK Biobank and China Kadoorie Biobank) with an updated systematic review and meta-analysis of prospective studies (Zhuang et al., 2025). Total dairy consumption was associated with a 3.7% lower risk of cardiovascular disease (CVD) and a 6% lower stroke risk. Cheese and low-fat dairy often showed the strongest protective signals; plain milk was more neutral. Fermented products frequently outperformed non-fermented ones, possibly due to bioactive peptides, probiotics, or favorable effects on blood pressure and lipids.
Placing this in narrative context with the earlier studies highlights important distinctions. While Michaëlsson et al. (2014) flagged mortality risks (which include CVD but also other causes), the broader 2025 synthesis suggests that when dairy is consumed moderately and includes fermented varieties, net cardiometabolic effects may be neutral to beneficial. Methodological strengths in Zhuang et al. include massive sample sizes, long follow-ups, and adjustments for energy intake, smoking, and other dietary factors. Shared limitations persist: observational nature means residual confounding (healthier people may choose yogurt/cheese), self-reported intake, and regional differences in dairy processing or baseline diets. Funding was primarily institutional/academic.
An umbrella review published in the same year (covering 33 prior meta-analyses) reinforces this nuanced view, finding mostly weak-to-suggestive evidence for modest CVD benefits from dairy (especially yogurt) alongside mixed bone outcomes—higher BMD but inconsistent fracture protection, with fermented dairy often faring best (as summarized in related 2025 analyses).
Why the contradictions?
Several factors explain the apparent chaos. First, dairy is not monolithic: liquid milk (high in lactose and D-galactose) can differ biologically from fermented products like yogurt and cheese, which contain probiotics, lower lactose, and unique peptides that may support gut and metabolic health. Second, dose matters—benefits or risks often appear at extremes (very low or very high intake). Third, populations differ: Swedish women in the 2014 study may have had unique dietary patterns or genetics compared to diverse global cohorts. Fourth, outcomes vary: a study focused on hip fractures or all-cause mortality may miss vertebral or CVD-specific signals. Fifth, universal weaknesses in nutrition research—recall bias in FFQs, residual confounding by overall healthy lifestyles, and challenges proving causation without long-term RCTs (which are rare and expensive in nutrition)—mean single studies or even metas rarely give the final word.
Funding transparency helps: the four highlighted studies were largely supported by public or academic sources rather than direct dairy industry grants, which reduces one common source of bias. However, the broader literature shows that industry-funded nutrition studies sometimes report more favorable results, underscoring the need for independent replication. Publication bias (positive findings more likely to be published) and small-study effects can further distort the picture.
FAQs
Is milk essential for strong bones? No. While milk provides calcium and supports bone mineral density (BMD) in some studies, meta-analyses show inconsistent links to reduced fracture risk overall. Higher dairy intake often shows neutral effects on total/hip fractures, with modest benefits only for vertebral fractures in certain populations.
Why do some studies link high milk intake to higher fracture risk? Observational cohorts (e.g., Swedish data) found associations with higher hip fractures and mortality at 3+ glasses/day, possibly due to D-galactose from lactose promoting inflammation/oxidative stress, residual confounding (unmeasured lifestyle factors), or reverse causation in frail individuals.
Does dairy really increase prostate cancer risk? Some meta-analyses show small increased risk (RR ~1.02 per increment) with higher dairy/milk intake, potentially via elevated IGF-1 (insulin-like growth factor-1) promoting cell growth. However, many studies are limited by PSA screening bias (more screening in dairy consumers leads to overdiagnosis), so evidence is limited-suggestive, not causal.
Is dairy good or bad for heart health? Recent global analyses link total dairy (especially fermented like yogurt/cheese) to modestly lower CVD and stroke risk (3–6% reduction). Full-fat dairy shows neutral or no harm in many modern reviews, unlike older concerns about saturated fat.
What about full-fat vs. low-fat milk? Full-fat milk adds saturated fat but recent evidence suggests neutral cardiometabolic effects; fermented/full-fat forms often show similar or better outcomes for blood pressure/lipids than low-fat in some trials. No major difference for most outcomes.
Why do studies conflict so much on dairy? Observational designs dominate (no long-term RCTs), with issues like self-reported FFQs (recall/measurement error), residual confounding, population differences (e.g., fortification levels, genetics), dairy subtypes (liquid milk vs. fermented), dose (high vs. moderate), and outcome-specific effects.
Does industry funding bias dairy research? Some positive studies have dairy industry ties, which correlate with more favorable conclusions in broader reviews. However, the four key studies cited were largely publicly/academically funded, reducing direct bias concerns.
How much dairy is “moderate” intake? Typically 1–2 servings/day (e.g., 8 oz milk + yogurt/cheese), aligning with patterns showing neutral-to-beneficial associations across outcomes.
Is lactose intolerance common, and what does it mean for dairy? Yes, ~65–70% of adults worldwide have reduced lactase enzyme activity. Symptoms (bloating, gas, diarrhea) vary; many tolerate small amounts or fermented/low-lactose options like yogurt/cheese.
Can I get enough calcium without dairy? Yes—leafy greens (kale, broccoli), fortified plant milks, nuts, seeds, tofu, and supplements provide calcium. Dairy is convenient but not required.
Does milk help prevent osteoporosis? It supports BMD in youth/adulthood, but evidence for fracture prevention is mixed/inconsistent. Weight-bearing exercise, vitamin D, and overall diet matter more.
Is raw milk healthier or safer? No—raw milk carries higher risk of pathogens (e.g., Salmonella, E. coli, Listeria), linked to outbreaks and serious illness. Pasteurization preserves nutrients without these risks.
Does milk cause inflammation or mucus/phlegm? No strong evidence; myths about increased phlegm during colds lack support in studies. Some feel it thickens mucus subjectively, but it’s not physiological.
Is dairy linked to other cancers (e.g., breast, colorectal)? Mixed: Some data show reduced risk for colorectal/bladder; inconsistent or no link for breast/ovarian; small potential increase for prostate in high intakes.
Does milk affect weight or obesity? Moderate dairy intake shows neutral or modest benefits for body composition; no consistent link to weight gain when calories are controlled.
Should kids drink milk for growth? It supports growth in children via protein/calcium/IGF-1, with some associations to better height/body composition. Not essential—alternatives work if tolerated.
Are there benefits to fermented dairy (yogurt/cheese)? Yes—often stronger protective signals for CVD, stroke, bone health, and gut microbiome due to probiotics, lower lactose, and bioactive peptides.
Can high dairy intake cause other issues (e.g., acne, hormones)? Possible links to acne in some (via IGF-1/hormones), but evidence is inconsistent. No broad causal proof for most adults.
What if I have a milk allergy? True allergy (to proteins like casein) is different from intolerance; avoid all dairy and seek alternatives. Affects ~2–3% of young children, often outgrown.
Bottom line: Should I drink milk? It depends on you. If tolerated, moderate amounts fit a balanced diet with potential benefits (nutrients, possible CVD/bone support). Limit extremes, prioritize variety, and consult a doctor/dietitian for personal factors like allergies, conditions, or family history.
Related reading:
Breastfeeding: Pumping Breast Milk Changes the Composition of Its Microbiota
Drinking Cow’s Milk Does Not Increase the Risk of Breast Cancer
Study Shows That Probiotics Can Evolve Making Them Highly Unpredictable
Final Thoughts
After digging deeply into these conflicting studies, the most evidence-based answer to “Is milk good or bad?” is: it depends. The weight of current research does not support extreme positions—neither “milk is poison” nor “drink unlimited amounts for perfect health.” Moderate intake (roughly 1–2 servings daily, ideally including fermented options like yogurt and cheese) appears compatible with good health for most people and may offer modest benefits for bone density, cardiovascular risk factors, and nutrient intake (calcium, protein, vitamin D when fortified). High intakes of plain liquid milk, particularly in certain populations or contexts, show occasional signals of harm (e.g., fractures or prostate cancer risk in some analyses), but these are often small, non-causal, or offset by other factors.
Common limitations across the evidence base—reliance on observational data, dietary reporting inaccuracies, confounding variables, and outcome-specific effects—mean we cannot claim definitive causation in either direction. Biological mechanisms (IGF-1 for cancer concerns, anti-inflammatory effects of fermented dairy for heart health, calcium’s role in bones) provide plausible explanations but require more randomized controlled trials in diverse populations to clarify.
For readers: Focus on your overall dietary pattern rather than obsessing over one food. If you tolerate dairy well (no lactose intolerance or allergies), include a variety of forms as part of a balanced, plant-rich diet. Those with specific conditions—history of prostate cancer, high fracture risk, or cardiovascular concerns—should discuss intake levels with a doctor or registered dietitian, who can consider personal factors like age, sex, genetics, medications, and lab results. Plant-based fortified alternatives can help meet nutrient needs when dairy is limited.
More long-term, high-quality studies in diverse groups are needed. In the meantime, evidence supports moderation, variety, and personalization over fear or overconsumption. Milk and dairy can be part of a healthy lifestyle for many, but they are not a one-size-fits-all solution.
(Word count: approximately 2490, excluding references and takeaways. This article is for informational purposes and is not medical advice.)
References
Matía-Martín, P., Torrego-Ellacuría, M., Larrad-Sainz, A., Fernández-Pérez, C., Cuesta-Triana, F., & Rubio-Herrera, M. Á. (2019). Effects of milk and dairy products on the prevention of osteoporosis and osteoporotic fractures in Europeans and non-Hispanic Whites from North America: A systematic review and updated meta-analysis. Advances in Nutrition, 10(Suppl_2), S120–S143. https://doi.org/10.1093/advances/nmy097
Michaëlsson, K., Wolk, A., Langenskiöld, S., Basu, S., Warensjö Lemming, E., Melhus, H., & Byberg, L. (2014). Milk intake and risk of mortality and fractures in women and men: Cohort studies. BMJ, 349, g6015. https://doi.org/10.1136/bmj.g6015
Zhao, Z., Wu, D., Gao, S., Zhou, D., et al. (2023). The association between dairy products consumption and prostate cancer risk: A systematic review and meta-analysis. British Journal of Nutrition, 129(10), 1714–1731. https://doi.org/10.1017/S0007114522002380
Zhuang, P., Liu, X., Li, Y., et al. (2025). A global analysis of dairy consumption and incident cardiovascular disease. Nature Communications, 16, 437. https://doi.org/10.1038/s41467-024-55585-0
