Milk and vitamin D – the more, the better?

Chair: Haakon E. Meyer 

Calcium and vitamin D are of great importance for skeletal health, and are routinely included in recommendations for the prevention of osteoporosis and fractures. However, there are substantial controversies concerning the optimal intake levels and doses. Whereas some claim that we need high doses of both nutrients, the results from systematic literature reviews are ambiguous and some studies have even reported adverse effect of high intakes. Not least, there is a current debate on whether a high milk intake is beneficial or not and whether high-dose vitamin D supplementation should be recommended. In this symposium, three researchers within nutrition, endocrinology and epidemiology will share their results and we will discuss which recommendations should be given for fracture prevention in older adults.

S1.5.1 Milk drinking and risk of hip fracture: Findings from Sweden

Liisa Byberg

Uppsala University, Sweden

Background: Consumption of milk has long been promoted to strengthen bone and reduce fracture risk although this has been difficult to demonstrate. Milk products may differ in their pro-oxidant properties and may therefore have different effects on fracture risk. Methods: Participants of a population-based cohort in Central Sweden (61,240 women born 1914-48) answered food frequency and lifestyle questionnaires at baseline in 1987-90. Repeat questionnaires were administered in 1997 (38,071 women contributed). Incident hip fracture events between baseline and 31 December 2014 were collected from the Swedish Patient Register with no loss to follow-up. Results: During follow-up, 5827 women had a hip fracture. Milk intake was associated with higher fracture rate; multivariable adjusted hazard ratio (HR) 1.07 (95% CI 1.04-1.10) per 200 ml milk. Compared with an intake of <1 glass per day, those women who consumed 3 or more glasses of milk per day had a HR of hip fracture of 1.51 (95% CI 1.33-1.71). In contrast, a higher intake of fermented milk (yogurt or soured milk - ”filmjölk”) was associated with lower hip fracture rate: HR 0.89 (95% CI 0.86-0.92) per 200 ml/day. Results were consistent in strata of potential confounders such as BMI and educational level. Associations between milk and fermented milk products and incident hip fractures (n=1166) in a cohort of 45,000 men were weaker or null. Potential explanations for the different results include biological differences in galactose degradation and perhaps more importantly that the male cohort had only one diet assessment, follow-up was shorter and hip fracture rate is lower among men compared to women.  Conclusion: Our results question the value of recommending high consumption of milk for the prevention of hip fractures. A moderate consumption of fermented milk products are however associated with lower hip fracture rates. 

S1.5.2 Milk drinking and risk of hip fracture: Findings from Norway 

Kristin Holvik 

Norwegian Institute of Public Health

Background: Findings from Sweden challenged previous meta-analyses of observational studies that had found no clear associations between milk drinking and risk of hip fracture. We aimed to study this association in Norwegians, who are similar to Swedes both in genetics and lifestyle factors, with a high incidence of hip fracture and a traditionally high milk consumption. Methods: Baseline data from two cohorts of population-based health studies were used: The third wave of the Norwegian Counties Study in Oppland, Sogn og Fjordane and Finnmark 1985-88, and regional health studies in Oslo, Hedmark, Oppland, Troms, and Finnmark 2000-2002 (referred to as ‘five counties’). Milk consumption and lifestyle and health information was self-reported through questionnaires. Height and weight were measured and demographic characteristics were obtained from Statistics Norway. The data was linked to the Norwegian Epidemiologic Osteoporosis Studies (NOREPOS) hip fracture database, which provides nationwide data on hip fractures treated in Norwegian hospitals 1994-2013. Subjects were followed until first hip fracture, death, emigration, or end of follow-up 31.12.2013. Data were analysed in Cox proportional hazards regression. Results: In the Norwegian Counties cohort, 1,870 incident hip fractures occurred among 35,165 participants. Hazard ratios for hip fracture per daily glass of milk in fully adjusted analysis were 0.97 (95% CI 0.91 to 1.02) in men and 1.03 (95% CI 0.98 to 1.08) in women (p<0.05 for statistical interaction between sex and milk consumption). In the ‘five counties’ cohort, 1,480 incident hip fractures occurred among 23,415 participants. Hazard ratio for hip fracture per daily glass of milk in fully adjusted analysis was 1.01 (95% CI 0.97 to 1.06, no sex-milk-interaction). Subgroup analyses revealed an increased risk in women with low body mass index drinking 4 or more glasses per day in both cohorts. This subgroup was small and characterized by high prevalence of cigarette smoking, physical inactivity, and poor self-rated health, suggestive of reverse causation. Conclusions: There was no overall association, neither protective nor risk-increasing, between milk drinking and subsequent risk of hip fracture in two Norwegian cohorts. 

S1.5.3 Vitamin D and bone – the search for the optimal dose 

Haakon E. Meyer

Norwegian Institute of Public Health and University of Oslo

Background: Since the discovery that cod liver oil cured rickets back in 1918, bone diseases have been the hallmark of vitamin D deficiency. Still, hundreds year later, we do not know the optimal vitamin D intake or levels for our skeleton.  As for most nutrients, the effects of vitamin D intake in the body follow a U-shaped pattern – with an increased risk of detrimental effects at both very low and very high intakes, with a broad range of intakes in between regarded safe and sufficient. What the tresholds for too little and too much should be for vitamin D, are still not clarified. Methods: We performed a one year randomized controlled trial comparing the effect of two different doses of vitamin D3 - an average daily dose of 800 IU versus 6500 IU - on bone mineral density (BMD) and bone turnover markers in 297 postmenopausal women with osteopenia. Results: After one year, serum 25-hydroxyvitamin D increased from 64 to 163 nmol/l in the 6500 IU group, and from 64 to 81 nmol/l in the 800 IU group. There was no differences between the two treatment groups regarding change in BMD. However, the bone turnover marker P1NP was more reduced in the 800 IU group. In addition, 1,25(OH)2D increased in the 6500 IU group, and decreased in the 800 IU group. Conclusions: Our results could indicate detrimental effects of high dose vitamin D on bone. Based on these results and results from other dose-comparing studies for BMD and fracture prevention, possible tresholds for vitamin D in relation to optimal bone health will be discussed. Also, challenges in comparing results from different studies will be highlighted.