Dairy consumption and 10-y total and cardiovascular mortality: a prospective cohort study in the Netherlands^1–3

ABSTRACT

Background: The consumption of dairy products (milk, cheese, and butter) has been positively associated with the risk of ischemic heart disease (IHD), stroke, and total mortality because of the saturated fat content of these products; and protective effects against these outcomes have been attributed to the calcium content and low-fat choices of dairy products. However, robust evidence on the net effect of dairy product consumption on mortality is limited.

Objective: The objective was to investigate the association between dairy product consumption and the risk of death (from all causes, IHD, and stroke) in the Netherlands Cohort Study (NLCS).

Design: The NLCS was initiated in 120,852 men and women aged 55–69 y at baseline in 1986. After 10 y of follow-up, 16,136 subjects with complete dietary information had died. Twenty-nine percent (men) and 22% (women) of these deaths were due to IHD or stroke. The validated 150-item food-frequency questionnaire provided detailed information on dairy products.

Results: Multivariate survival analyses following a case-cohort approach showed only a few statistically significant, but mostly weak, associations. A slightly increased risk of all-cause and IHD mortality was found for both butter and dairy fat intake (per 10 g/d; rate ratio_mortality: 1.04; 95% CI: 1.01, 1.06) only in women. Fermented full-fat milk was inversely associated with all-cause and nonsignificantly with stroke mortality in both sexes.

Conclusions: The role of dairy product consumption in mortality generally appeared to be neutral in men. In women, dairy fat intake was associated with slightly increased all-cause and IHD mortality. More research is warranted on a possible protective effect of fermented milk on stroke mortality.

INTRODUCTION

The role of dairy products in the traditional diet varies greatly in different regions of the world. Populations in tropical countries have not been traditional milk consumers, whereas the more northern regions of the world—northern parts of Europe and North America—have traditionally consumed far more milk and milk products as part of their diet.

In most northwestern European countries, milk and other dairy foods contribute >25% to the total fat intake and almost 50% to the saturated fat intake (1). A diet high in saturated fatty acids and cholesterol has long been associated with cardiovascular disease and mortality, accounting for 49% of all deaths in Europe (2). Dairy product intake therefore has long been considered a risk factor for cardiovascular disease on the basis of relevant biological mechanisms and data from ecological studies (3, 4). On the other hand, a meta-analysis of cohort studies on milk and cardiovascular disease risk provided no convincing evidence that milk is harmful and suggested that milk drinking may be associated with a small but worthwhile reduction in heart disease and stroke risk (5). Growing evidence indicates that calcium and protein in milk and dairy products may be associated with reduced blood pressure and risk of stroke (6, 7). Dietary Approaches to Stop Hypertension (DASH) has shown that substantial blood pressure reductions can be achieved when low-fat dairy products were added to a healthy food and vegetable diet (8). Furthermore, studies have suggested that dairy or calcium contribute to maintaining a healthy body weight (9). Because overweight and obesity are important risk factors for a range of chronic diseases (eg, cardiovascular disease, type 2 diabetes, and many cancers), dairy consumption might have a favorable effect on chronic disease through this pathway.

Considering the postulated harmful and favorable effects of dairy consumption on a range of chronic diseases, we wanted to know how these seemingly opposite effects are balanced. Total mortality can be viewed as the final net result of all serious positive and negative influences on health. We therefore wanted to investigate how dairy product consumption, including not only dairy foods but calcium and dairy fat, is associated with total mortality. In the Netherlands, which is one of the top-5 countries in the world with the highest per capita consumption of milk and milk products, the Netherlands Cohort Study (NLCS) offers a unique opportunity to study the association between the consumption of milk and dairy foods and mortality. Because cardiovascular disease [ischemic heart disease (IHD) and stroke] is one of the most important causes of death, we also studied this cause of death to see whether the association between dairy product consumption and total mortality, if any, can be explained by cardiovascular disease mortality.

SUBJECTS AND METHODS

Study population

The study design of the NLCS has been reported in detail elsewhere (10). The NLCS was initiated in September 1986 when a self-administered questionnaire on dietary habits, lifestyle characteristics, medical history, and other potential risk factors for cancer and mortality was completed by 58,279 men and 62,573 women aged 55–69 y. After baseline exposure measurements were made, a subcohort of 5000 subjects was randomly sampled from the large cohort. Following the case-cohort approach, this subcohort was followed up for vital status and migration to estimate person time at risk accumulated in the cohort.

Data on mortality and causes of death between 1 January 1987 and 31 December 1996 in the cohort-at-large were obtained from the Dutch Central Bureau of Genealogy and the Dutch Central Bureau of Statistics. By comparing the deaths reported through this system with those identified in the subcohort through active follow-up, we estimated that mortality follow-up during this period was 99% complete for the entire cohort. The cause of death was coded according to the International Classification of Disease, ninth revision (ICD-9) for the period 1987–1995 and tenth revision (ICD-10) for 1996. Besides total mortality, the following primary causes of death were separately investigated: IHD (ICD-10: I20-I25) and stroke (ICD-10: I60-I69). In total, 18,064 deaths occurred between the 2 dates.

Subjects with very incomplete and inconsistent FFQs (11) were excluded from the analysis, which left 2254 male and 2392 female subcohort members and 10,658 male and 5478 female deaths for analysis. Of the deaths, 2408 (in men) and 826 (in women) were due to IHD and 657 (in men) and 397 (in women) were due to stroke as primary cause. The NLCS was approved by the institutional review boards of the TNO Nutrition and Food Research Institute (Zeist) and Maastricht University (Maastricht).

Questionnaire

The dietary portion of this questionnaire consisted of a 150-item semiquantitative food-frequency questionnaire (FFQ) on the usual intake of food and beverages in the year preceding the start of the study. The FFQ contained 9 items on consumption of milk and milk products (milk, yogurt, buttermilk, quark, and dishes in which these foods were used), which were further specified by fat content (full fat, low fat, or skim), 3 items on cheese, and 3 items on butter.

The FFQ was validated against a 9-d diet record (11). The Spearman correlation coefficients for milk (products), cheese, and added fats (which include butter) were 0.60, 0.61, and 0.57, respectively. For calcium, the (energy and gender-adjusted) Pearson correlation coefficient was 0.62. On average, the FFQ captured 86%, 64%, 103%, and 84% of the absolute intake assessed by the record of milk (products), cheese, added fats, and calcium, respectively. The capture rates are the net result of the design of the FFQ (not all possible food items were covered, eg, ice cream) on one hand and under- or overreporting by the respondents on the other hand.

The individual frequencies and serving sizes of the food items were converted into mean daily intake (g/d) for each subject. Energy and nutrient intakes and the contribution of all dairy products combined to total fat intake were calculated by using the Dutch Food Composition Table (12). Fat variables were adjusted for energy by using the residual method (13).

Other relevant risk factors available from the questionnaire were education, smoking habits (cigarettes, cigars, and pipe; duration and quantity), multivitamin use, height, weight, nonoccupational physical activity, and occupational physical activity (derived from job histories and by using a Job Exposure Matrix with 3 levels of physical activity). The questionnaire data for all cases and subcohort members were key-entered and blinded with respect to case/subcohort status to avoid random and systematic coding errors.

Statistical analyses

All analyses were conducted separately for men and women. Cox proportional hazards models were constructed to estimate hazard ratios and 95% CIs relating dairy product variables and calcium intake to all-cause mortality and to a selection of specific causes of death (Stata version 9; Stata Corporation, College Station, TX). The proportional hazards assumption was tested by using Schoenfeld residuals. When this test suggested that the assumption was violated, the -ln(-ln) survival plots to assess proportional hazard assumption were evaluated. The plotted lines were reasonably parallel in all analyses, which indicated that the proportional hazards assumption was not violated. Because of the case-cohort design, the 95% CIs were corrected for the additional variance introduced by using a random sampled subcohort instead of the complete cohort, by using the robust option (14). Linear trends for categorical variables were tested by the Wald test that contained median values of intake quintiles or categories.

Besides the main dairy groups (milk, cheese, and butter), the available dairy product items were aggregated into several variables according to their fat content (separating full-fat from low-fat and skim items) and fermentation status (yogurt, buttermilk, and quark). Also, a variable was constructed to aggregate all low-fat dairy items. Dairy product variables were analyzed as continuous variables (with increments of 100 or 10 g/d) and as categorical variables (quintiles or categories, depending on the shape of their distribution and the proportion of nonusers). Two-sided tests for trend in the rate ratios (RRs) of categorical analyses were assessed by fitting the median of exposure categories as continuous variables. We also derived a variable to assess compliance with the current Dutch recommendations for dairy product consumption for adults [500 mL milk (products) and 30 g cheese/d] (15) according to the algorithm milk/0.86 + 10 × cheese/0.64. This algorithm took into account that 1 kg cheese is made from 10 L milk and that the dairy items in the FFQ, on average, underestimated absolute intake. We classified subjects into 3 compliance categories: too low (<600 mL/d), in compliance (600–1000 mL, ie, ± 25% of 800 mL/d), and too high (>1000 mL/d).

A range of risk factors for chronic disease [ie, age at baseline (continuous); education (4 categories); smoking status (current, yes or no; duration, in years; and number smoked daily) for cigarettes, cigars, and pipe in men and for cigarettes only in women; body mass index (BMI; 4 categories); nonoccupational physical activity (4 categories); occupational physical activity (3 categories, men only); multivitamin use (yes or no); alcohol (one category for nondrinkers, 4 categories for current drinkers, and 1 category for current nondrinkers who were former drinkers 5 y before baseline); and intake of energy, energy-adjusted monounsaturated and polyunsaturated fats, and vegetable and fruit consumption (all continuous)] were included as potential confounders in the multivariate models. For occupational physical activity, a covariate with a substantial number of missing values (ie, 7.3%), a dummy variable for the missing category, was added. Cases and subcohort members with missing values for one or more other covariates (12% of the cohort, as estimated from the subcohort) were excluded from the analysis. Energy-adjusted saturated fat was added to the models in a separate analysis to investigate whether milk fat (mainly classified as saturated fat) had the same effects as saturated fat.

We conducted several sensitivity analyses. To evaluate whether disease preceding death could have influenced the results, early deaths (those that occurred within 2 y from baseline) were excluded from the analyses. We also investigated whether exclusion of subjects who reported cardiovascular disease at baseline (myocardial infarction, angina pectoris, or stroke) changed the results. To check whether BMI was an intermediate variable in the association between dairy product consumption and mortality, we excluded BMI from the multivariate models and compared the results.

RESULTS

No relevant differences in dairy product consumption and fat from dairy products or calcium were observed between deaths and the subcohort, for men or women (Table 1). Although milk and other dairy product consumption were similar for men and women, consumption of fermented low-fat milk (women drank more buttermilk) and full-fat whole milk (men drank more) differed. As a result, dairy fat intake was somewhat higher in men than in women. Slightly more than one-third of the NLCS population complied with the current Dutch recommendation for adults to drink 500 mL milk or milk products and 30 g cheese per day.

TABLE 1

Dairy consumption in the subcohort and in cases (death from all causes): Netherlands Cohort Study 1986–1996

	Men		Women
	Subcohort (n = 2254)	Cases (n = 10,658)	Subcohort (n = 2392)	Cases (n = 5478)
Milk products (g/d)	309 ± 2181	302 ± 215	301 ± 191	306 ± 210
Nonfermented full-fat milk2	135 ± 169	131 ± 159	89 ± 114	99 ± 131
Nonfermented low-fat milk3	89 ± 138	92 ± 140	92 ± 133	93 ± 137
Fermented full-fat milk4	16 ± 40	12 ± 36	18 ± 42	15 ± 38
Fermented low-fat milk5	70 ± 109	68 ± 110	102 ± 125	99 ± 130
Cheese (g/d)	23 ± 20	22 ± 20	23 ± 18	22 ± 18
Fat cheese	22 ± 20	21 ± 20	21 ± 18	19 ± 18
Low-fat cheese	2 ± 7	2 ± 7	2 ± 7	2 ± 8
Butter (g/d)	9 ± 16	9 ± 16	8 ± 13	9 ± 14
Low-fat dairy (g/d)	160 ± 178	161 ± 182	195 ± 183	194 ± 195
Fat from dairy (g/d)	21.6 ± 16.6	20.8 ± 16.7	18.7 ± 13.8	19.5 ± 14.6
Calcium (mg/d)	950 ± 343	928 ± 352	904 ± 310	897 ± 335
Compliance with recommendation (%)6
Too low	43.0	46.1	43.2	45.1
In compliance	36.2	34.1	38.3	36.5
Too high	20.9	19.8	18.5	18.3

1

Mean ± SD (all such values).

2

Includes whole milk (3.7% fat), cream (36%, 20% fat), condensed whole milk, whole-milk cocoa, pudding, and ice cream.

3

Includes low-fat milk (1.5% fat), skim milk (0.1% fat), condensed low-fat milk, and low-fat and skim cocoa.

4

Includes yogurt (3.5% fat), full-fat quark (fresh cheese), and sour cream.

5

Includes buttermilk, skim yogurt (0.1% fat), and nonfat quark (fresh cheese).

6

Calculated according to the following algorithm: 600 < milk/0.86 + 10 × cheese/0.64 < 1000 g/d.

Consumption of the main dairy foods across categories of the covariates is shown in Table 2. Older subjects and current smokers consumed less milk and milk products, but more butter. BMI was inversely associated with consumption of all dairy products. Energy intake was, as would be expected, positively associated with the consumption of all dairy products. Whereas consumption of milk and milk products was inversely associated with both mono- and polyunsaturated fat intakes, butter consumption was strongly inversely associated with polyunsaturated fat intake only: the quintile with the lowest intake of polyunsaturated fat consumed 4 times more butter than did the highest quintile. This was explained by the substitution of butter for margarine containing a high content of polyunsaturated fat. Fruit consumption appeared to be positively associated with milk and milk products, whereas alcohol intake showed an inverse association with these foods.

TABLE 2

Association between dairy product consumption and characteristics of the subcohort: Netherlands Cohort Study 1986–1996

	Men					Women
	n	%	Milk	Cheese	Butter	n	%	Milk	Cheese	Butter
			g/d	g/d	g/d			g/d	g/d	g/d
Age
55–59 y	704	37.3	322 ± 2221	23 ± 19	9 ± 16	819	36.7	310 ± 202	23 ± 19	7 ± 12
60–64 y	672	35.6	310 ± 218	24 ± 21	9 ± 15	747	33.5	294 ± 177	22 ± 18	8 ± 12
65–69 y	513	27.2	301 ± 216	22 ± 19	10 ± 16	663	29.7	291 ± 187	23 ± 18	8 ± 13
Cigarette smoking
Never	255	13.5	338 ± 228	23 ± 19	9 ± 15	1320	59.2	303 ± 182	24 ± 19	7 ± 12
Former	996	52.7	306 ± 205	24 ± 20	9 ± 15	453	20.3	311 ± 199	23 ± 16	8 ± 12
Current	638	33.8	311 ± 235	22 ± 20	10 ± 17	456	20.5	276 ± 199	19 ± 17	8 ± 13
Occupational physical activity
<8 kJ/min	1073	56.8	300 ± 212	24 ± 20	10 ± 16
8–12 kJ/min	442	23.4	317 ± 211	21 ± 20	7 ± 15
>12 kJ/min	236	12.5	353 ± 223	23 ± 20	9 ± 16
Missing	138	7.3	318 ± 279	21 ± 18	8 ± 15
Nonoccupational physical activity
<30 min/d	283	15.0	322 ± 258	22 ± 17	11 ± 17	596	26.7	282 ± 196	20 ± 18	7 ± 12
30–59 min/d	550	29.1	309 ± 220	23 ± 20	9 ± 15	786	35.3	306 ± 188	23 ± 19	7 ± 12
60–89 min/d	671	35.5	307 ± 200	24 ± 20	9 ± 16	604	27.1	298 ± 182	24 ± 18	8 ± 13
≥90 min/d	385	20.4	319 ± 217	23 ± 21	8 ± 15	243	10.9	321 ± 191	25 ± 18	9 ± 13
Education
Primary	436	23.2	321 ± 218	22 ± 19	8 ± 8	733	33.0	300 ± 183	21 ± 19	6 ± 12
Extended primary	393	20.9	321 ± 213	22 ± 19	7 ± 14	518	23.3	285 ± 180	20 ± 17	6 ± 12
Secondary (including vocational)	689	36.6	309 ± 237	23 ± 20	10 ± 16	772	34.8	300 ± 197	24 ± 18	9 ± 14
College/university	363	19.3	300 ± 190	26 ± 20	12 ± 17	196	8.8	323 ± 198	29 ± 17	11 ± 13
BMI
≤22 kg/m2	399	21.1	332 ± 250	24 ± 22	11 ± 17	610	27.4	304 ± 195	24 ± 18	9 ± 14
23–24 kg/m2	612	32.4	322 ± 212	23 ± 19	9 ± 16	625	28.0	313 ± 196	24 ± 20	8 ± 13
25–26 kg/m2	504	26.7	299 ± 207	24 ± 20	8 ± 15	436	19.6	288 ± 174	21 ± 17	6 ± 11
≥27 kg/m2	374	19.8	292 ± 210	22 ± 18	9 ± 15	558	25.0	286 ± 185	21 ± 17	6 ± 11
Energy
Quintile 1 (M: 6486 kJ; F: 5045 kJ)	372	19.7	229 ± 164	17 ± 15	6 ± 11	454	20.4	211 ± 143	15 ± 13	4 ± 8
Quintile 2 (M: 7847 kJ; F: 6089 kJ)	381	20.2	292 ± 187	20 ± 16	8 ± 13	441	19.8	270 ± 171	19 ± 14	6 ± 10
Quintile 3 (M: 8861 kJ; F: 6914 kJ)	375	19.9	309 ± 203	22 ± 19	9 ± 15	445	20.0	306 ± 182	25 ± 17	8 ± 12
Quintile 4 (M: 9956 kJ; F: 7779 kJ)	383	20.3	326 ± 205	26 ± 20	12 ± 18	447	20.1	323 ± 177	24 ± 18	9 ± 14
Quintile 5 (M: 11,757 kJ; F: 9175 kJ)	378	20.0	403 ± 282	31 ± 24	11 ± 19	442	19.8	387 ± 220	30 ± 23	11 ± 16
Multivitamin use
No	18.4	96.6	313 ± 220	23 ± 20	9 ± 16	2097	94.1	298 ± 187	22 ± 18	7 ± 12
Yes	65	3.4	289 ± 200	19 ± 16	13 ± 16	132	5.9	313 ± 221	27 ± 20	10 ± 14
Monounsaturated fat, energy-adjusted
Quintile 1 (M: 28 g/d; F: 21 g/d)	372	19.7	364 ± 259	27 ± 22	7 ± 14	451	20.2	358 ± 214	25 ± 20	6 ± 11
Quintile 2 (M: 32 g/d; F: 25 g/d)	377	20.0	301 ± 208	22 ± 17	8 ± 15	443	19.9	318 ± 196	22 ± 16	8 ± 13
Quintile 3 (M: 35 g/d; F: 27 g/d)	393	20.8	306 ± 196	21 ± 19	12 ± 17	452	20.3	283 ± 173	23 ± 20	7 ± 13
Quintile 4 (M: 38 g/d; F: 30 g/d)	371	19.6	292 ± 202	22 ± 18	9 ± 16	439	19.7	282 ± 167	21 ± 16	9 ± 13
Quintile 5 (M: 44 g/d; F: 33 g/d)	376	19.9	299 ± 220	24 ± 21	9 ± 15	444	19.9	253 ± 175	23 ± 18	7 ± 12
Polyunsaturated fat, energy-adjusted
Quintile 1 (M: 28 g/d; F: 23 g/d)	380	20.1	385 ± 281	26 ± 22	25 ± 20	441	19.8	368 ± 215	27 ± 22	21 ± 16
Quintile 2 (M: 33 g/d; F: 27 g/d)	371	19.6	306 ± 196	23 ± 20	12 ± 16	441	19.8	300 ± 184	23 ± 17	9 ± 12
Quintile 3 (M: 36 g/d; F: 29 g/d)	366	19.4	291 ± 191	21 ± 18	6 ± 11	446	20.0	273 ± 168	20 ± 17	4 ± 8
Quintile 4 (M: 40 g/d; F: 32 g/d)	383	20.3	292 ± 203	22 ± 18	2 ± 7	449	20.1	273 ± 183	22 ± 17	3 ± 7
Quintile 5 (M: 47 g/d; F: 38 g/d)	389	20.6	287 ± 195	24 ± 20	1 ± 5	452	20.3	282 ± 178	21 ± 17	1 ± 4
Vegetable consumption
Quintile 1 (M: 100 g/d; F: 105 g/d)	370	19.6	318 ± 259	21 ± 19	9 ± 15	430	19.3	275 ± 191	21 ± 18	8 ± 13
Quintile 2 (M: 144 g/d; F: 149 g/d)	376	19.9	320 ± 213	22 ± 18	9 ± 15	446	20.0	285 ± 183	21 ± 16	7 ± 12
Quintile 3 (M: 178 g/d; F: 182 g/d)	374	19.8	308 ± 215	23 ± 21	9 ± 15	452	20.3	312 ± 197	22 ± 16	8 ± 13
Quintile 4 (M: 218 g/d; F: 222 g/d)	374	19.8	308 ± 196	23 ± 19	11 ± 17	454	20.4	309 ± 186	24 ± 19	7 ± 12
Quintile 5 (M: 293 g/d; F: 299 g/d)	395	20.9	307 ± 209	26 ± 21	8 ± 16	447	20.1	312 ± 186	25 ± 22	8 ± 13
Fruit consumption
Quintile 1 (M: 29 g/d; F: 59 g/d)	368	19.5	285 ± 242	20 ± 20	8 ± 16	432	19.4	245 ± 170	19 ± 17	7 ± 12
Quintile 2 (M: 88 g/d; F: 126 g/d)	370	19.6	290 ± 212	22 ± 20	10 ± 16	446	20.0	282 ± 180	22 ± 18	7 ± 12
Quintile 3 (M: 134 g/d; F: 179 g/d)	377	20.0	322 ± 203	25 ± 19	9 ± 15	448	20.1	304 ± 191	23 ± 17	8 ± 13
Quintile 4 (M: 189 g/d; F: 239 g/d)	386	20.4	321 ± 203	22 ± 18	9 ± 16	450	20.2	320 ± 185	23 ± 17	8 ± 13
Quintile 5 (M: 297 g/d; F: 353 g/d)	388	20.5	341 ± 228	26 ± 21	9 ± 16	453	20.3	340 ± 204	26 ± 21	8 ± 13
Alcohol intake
0 g/d	241	12.9	363 ± 266	22 ± 20	9 ± 17	663	31.1	304 ± 188	22 ± 19	6 ± 12
0.1–4 g/d	395	21.1	325 ± 213	23 ± 19	9 ± 16	766	36.0	308 ± 184	22 ± 18	7 ± 12
5–14 g/d	503	26.9	324 ± 208	24 ± 21	9 ± 15	392	18.4	302 ± 196	24 ± 17	9 ± 13
15–29 g/d	418	22.3	284 ± 198	22 ± 18	8 ± 14	194	9.1	256 ± 174	22 ± 17	10 ± 14
≥30 g/d	275	14.7	259 ± 205	25 ± 21	11 ± 17	79	3.7	241 ± 191	23 ± 17	12 ± 14
Former drinker	39	2.1	365 ± 301	24 ± 18	7 ± 13	35	1.6	343 ± 258	18 ± 16	9 ± 12

1

Mean ± SD (all such values).

Results of the survival analyses for all-cause mortality are presented in Table 3 (for men) and Table 4 (for women). Because of the high number of deaths (twice as many men than women), the CIs around the estimated RRs were very narrow. Almost no associations between consumption of dairy products and all-cause mortality were observed among men or women. However, a few clear trends were present: consumption of fermented full-fat milk showed a small, but consistent, inverse and statistically significant association with mortality, both for men (RR_continuous: 0.91; 95% CI: 0.86, 0.97 per 100 mL/d) and women (RR_continuous: 0.92; 95% CI: 0.85, 1.00 per 100 mL/d). The very low consumption of low-fat cheese was associated with higher mortality among men (RR_continuous: 1.05; 95% CI: 1.02, 1.08 per 10 g/d), but not among women. This association became slightly weaker, but remained statistically significant after exclusion of men with preexisting cardiovascular conditions at baseline (RR_continuous: 1.04; 95% CI: 1.00, 1.08 per 10 g/d). Furthermore, consumption of butter increased the risk of death slightly and significantly, but only among women; similar results were found for fat from dairy products (correlation between butter and fat from dairy products: 0.73). The risk of death was 8% (95% CI: 1, 15%) higher for women who ate butter than for women who did not eat butter, most of whom substituted butter for margarine. In men, a slightly increased risk was observed in the highest quintile of calcium intake (median: 1387 mg/d; RR: 1.06; 95% CI: 0.99, 1.15). Exclusion of deaths that occurred in the first 2 y of follow-up did not change the results, nor did the exclusion of subjects with preexisting cardiovascular conditions (data not shown).

TABLE 3

Risk of death associated with dairy product variables (by quintile or category or continuous) in men: Netherlands Cohort Study 1986–1996¹

	Q1/C1	Q2/C2	Q3/C3	Q4/C4	Q5/C5	P for trend2	Continuous
Milk products
Median intake (g/d)	78	180	268	373	577		Per 100 g/d
Person-years	4101	4061	4094	4124	4100
Number of cases	2305	1986	2114	2232	2021
Age-adjusted RR (95% CI)	1.00 (Ref)	0.97 (0.91, 1.03)	0.96 (0.91, 1.02)	1.00 (0.95, 1.06)	0.98 (0.92, 1.04)	0.885	1.00 (0.99, 1.01)
Multivariate RR (95% CI)3	1.00 (Ref)	1.01 (0.94, 1.08)	0.98 (0.92, 1.05)	1.05 (0.99, 1.12)	1.02 (0.95, 1.09)	0.372	1.00 (0.99, 1.01)
Nonfermented full-fat milk
Median intake (g/d)	0	34	90	156	342		Per 100 g/d
Person-years	4113	4783	3333	4163	4088
Number of cases	2318	2518	1650	2041	2131
Age-adjusted RR (95% CI)	1.00 (Ref)	0.97 (0.92, 1.03)	0.97 (0.91, 1.03)	0.97 (0.92, 1.03)	1.00 (0.94, 1.06)	0.772	1.00 (0.99, 1.01)
Multivariate RR (95% CI)3	1.00 (Ref)	0.99 (0.93, 1.05)	1.00 (0.94, 1.08)	1.01 (0.94, 1.08)	1.02 (0.95, 1.09)	0.427	1.00 (0.99, 1.02)
Nonfermented low-fat milk
Median intake (g/d)	0	32	186				Per 100 g/d
Person-years	9119	4578	6782
Number of cases	4696	2305	3657
Age-adjusted RR (95% CI)	1.00 (Ref)	0.97 (0.93, 1.02)	1.01 (0.97, 1.05)			0.511	1.01 (0.99, 1.02)
Multivariate RR (95% CI)3	1.00 (Ref)	0.97 (0.92, 1.03)	1.02 (0.97, 1.07)			0.257	1.01 (1.00, 1.03)
Fermented full-fat milk
Median intake (g/d)	0	53					Per 100 g/d
Person-years	15,679	4801
Number of cases	8732	1926
Age-adjusted RR (95% CI)	1.00 (Ref)	0.90 (0.85, 0.94)					0.89 (0.85, 0.95)
Multivariate RR (95% CI)3	1.00 (Ref)	0.93 (0.88, 0.98)					0.91 (0.86, 0.97)
Fermented low-fat milk
Median intake (g/d)	0	21	146				Per 100 g/d
Person-years	7666	5962	6851
Number of cases	4440	2705	3513
Age-adjusted RR (95% CI)	1.00 (Ref)	0.90 (0.86, 0.95)	0.95 (0.91, 0.99)			0.160	1.00 (0.98, 1.01)
Multivariate RR (95% CI)3	1.00 (Ref)	0.92 (0.87, 0.97)	0.97 (0.93, 1.03)			0.893	1.00 (0.98, 1.02)
Cheese
Median intake (g/d)	1	11	19	37	56		Per 10 g/d
Person-years	3697	3279	5253	5342	2909
Number of cases	2239	1614	2787	2610	1408
Age-adjusted RR (95% CI)	1.00 (Ref)	0.91 (0.86, 0.97)	0.93 (0.88, 0.99)	0.93 (0.88, 0.98)	0.94 (0.88, 1.00)	0.126	1.00 (0.99, 1.01)
Multivariate RR (95% CI)3	1.00 (Ref)	0.94 (0.87, 1.01)	0.99 (0.93, 1.05)	1.01 (0.95, 1.08)	1.04 (0.96, 1.12)	0.083	1.01 (1.00, 1.02)
Fat cheese
Median intake (g/d)	0	7	19	29	56		Per 10 g/d
Person-years	3736	3911	4556	5788	2498
Number of cases	2294	1990	2413	2762	1199
Age-adjusted RR (95% CI)	1.00 (Ref)	0.94 (0.89, 1.00)	0.94 (0.89, 1.00)	0.92 (0.87, 0.97)	0.93 (0.86, 0.99)	0.015	0.99 (0.98, 1.00)
Multivariate RR (95% CI)3	1.00 (Ref)	0.93 (0.87, 0.99)	0.96 (0.90, 1.02)	0.97 (0.91, 1.03)	0.99 (0.91, 1.07)	0.715	1.01 (0.99, 1.02)
Low-fat cheese
Median intake (g/d)	0	7					Per 10 g/d
Person-years	18,100	2379
Number of cases	9325	1333
Age-adjusted RR (95% CI)	1.00 (Ref)	1.08 (1.02, 1.15)					1.03 (1.00, 1.06)
Multivariate RR (95% CI)3	1.00 (Ref)	1.13 (1.06, 1.20)					1.05 (1.02, 1.08)
Butter
Median intake (g/d)	0	25					Per 10 g/d
Person-years	13,588	6892
Number of cases	7235	3423
Age-adjusted RR (95% CI)	1.00 (Ref)	0.97 (0.93, 1.01)					1.00 (0.99, 1.01)
Multivariate RR (95% CI)3	1.00 (Ref)	0.96 (0.91, 1.01)					1.00 (0.99, 1.02)
Low-fat dairy
Median intake (g/d)	0	28	119	211	380		Per 100 g/d
Person-years	4278	3964	4314	3813	4110
Number of cases	2320	1983	2363	1758	2234
Age-adjusted RR (95% CI)	1.00 (Ref)	0.96 (0.90, 1.02)	1.01 (0.95, 1.07)	0.93 (0.87, 0.99)	1.00 (0.94, 1.06)	0.855	1.00 (0.99, 1.01)
Multivariate RR (95% CI)3	1.00 (Ref)	0.95 (0.89, 1.01)	1.02 (0.96, 1.09)	0.94 (0.87, 1.00)	1.01 (0.95, 1.08)	0.584	1.01 (1.00, 1.02)
Fat from dairy
Median intake (g/d)	5.4	10.9	16.8	26.0	44.7		Per 10 g/d
Person-years	4074	4123	4103	4073	4107
Number of cases	2464	2189	2041	1912	2052
Age-adjusted RR (95% CI)	1.00 (Ref)	0.97 (0.92, 1.03)	0.94 (0.89, 1.00)	0.94 (0.89, 1.00)	0.96 (0.90, 1.02)	0.191	0.99 (0.98, 1.01)
Multivariate RR (95% CI)3	1.00 (Ref)	0.99 (0.93, 1.06)	0.98 (0.92, 1.05)	0.97 (0.90, 1.04)	1.00 (0.93, 1.09)	0.956	1.01 (0.99, 1.02)
Calcium
Median intake (mg/d)	549	757	907	1083	1387		Per 100 mg/d
Person-years	4004	4144	4094	4182	4056
Number of cases	2509	2095	2024	2030	2000
Age-adjusted RR (95% CI)	1.00 (Ref)	0.94 (0.89, 1.00)	0.95 (0.90, 1.01)	0.93 (0.88, 0.98)	0.96 (0.90, 1.02)	0.141	1.00 (0.99, 1.00)
Multivariate RR (95% CI)3	1.00 (Ref)	0.99 (0.93, 1.05)	1.01 (0.94, 1.08)	1.02 (0.95, 1.09)	1.06 (0.99, 1.15)	0.082	1.01 (1.00, 1.02)

1

Q1–Q5, quintiles 1–5 based on distribution in the subcohort; C1–C5, categories; C1, nonusers; RR, rate ratio; Ref, reference.

2

Wald test of linear estimate of variable containing median values of intake categories.

3

Adjusted for age (y); education (4 categories); cigarette, cigar, and pipe smoking (status, number, and duration); nonoccupational physical activity (4 categories); occupational physical activity (3 categories plus 1 missing); BMI (4 categories); multivitamin use (yes or no); alcohol (5 categories plus former drinkers); energy (kJ); energy-adjusted mono- and polyunsaturated fat intakes (continuous); and vegetable and fruit consumption (continuous).

TABLE 4

Risk of death associated with dairy product variables (quintiles or categories and continuous) in women: Netherlands Cohort Study 1986–1996¹

	Q1/C1	Q2/C2	Q3/C3	Q4/C4	Q5/C5	P for trend2	Continuous
Milk products
Median intake (g/d)	72	189	278	370	537		Per 100 g/d
Person-years	4475	4558	4569	4590	4580
Number of cases	1143	1078	1077	1000	1180
Age-adjusted RR (95% CI)	1.00 (Ref)	0.98 (0.90, 1.06)	0.97 (0.89, 1.05)	0.92 (0.85, 1.00)	1.05 (0.97, 1.14)	0.385	1.01 (1.00, 1.03)
Multivariate RR (95% CI)3	1.00 (Ref)	1.01 (0.92, 1.11)	1.00 (0.92, 1.10)	0.96 (0.87, 1.05)	1.09 (0.99, 1.20)	0.141	1.01 (1.00, 1.03)
Nonfermented full-fat milk
Median intake (g/d)	0	21	52	107	238		Per 100 g/d
Person-years	4994	4126	4673	4417	4562
Number of cases	1266	892	1095	995	1230
Age-adjusted RR (95% CI)	1.00 (Ref)	0.92 (0.85, 1.01)	0.95 (0.88, 1.03)	0.94 (0.87, 1.02)	1.04 (0.96, 1.12)	0.056	1.03 (1.01, 1.06)
Multivariate RR (95% CI)3	1.00 (Ref)	0.96 (0.87, 1.05)	0.96 (0.88, 1.04)	0.94 (0.86, 1.04)	1.00 (0.91, 1.09)	0.678	1.02 (1.00, 1.05)
Nonfermented low-fat milk
Median intake (g/d)	0	40	201				Per 100 g/d
Person-years	9576	5555	7642
Number of cases	2396	1255	1827
Age-adjusted RR (95% CI)	1.00 (Ref)	0.98 (0.92, 1.05)	0.99 (0.93, 1.05)			0.846	1.01 (0.99, 1.03)
Multivariate RR (95% CI)3	1.00 (Ref)	1.05 (0.97, 1.13)	1.00 (0.93, 1.07)			0.699	1.00 (0.98, 1.02)
Fermented full-fat milk
Median intake (g/d)	0	53					Per 100 g/d
Person-years	16,628	6145
Number of cases	4197	1281
Age-adjusted RR (95% CI)	1.00 (Ref)	0.91 (0.85, 0.97)					0.90 (0.84, 0.97)
Multivariate RR (95% CI)3	1.00 (Ref)	0.93 (0.87, 1.00)					0.92 (0.85, 1.00)
Fermented low-fat milk
Median intake (g/d)	0	60	192				Per 100 g/d
Person-years	7560	7595	7617
Number of cases	2037	1620	1821
Age-adjusted RR (95% CI)	1.00 (Ref)	0.89 (0.83, 0.95)	0.95 (0.89, 1.01)			0.602	1.00 (0.98, 1.02)
Multivariate RR (95% CI)3	1.00 (Ref)	0.94 (0.88, 1.01)	1.02 (0.95, 1.09)			0.265	1.01 (1.00, 1.04)
Cheese
Median intake (g/d)	1	11	19	37	56		Per 10 g/d
Person-years	3718	3645	6283	6229	2896
Number of cases	1072	915	1390	1458	634
Age-adjusted RR (95% CI)	1.00 (Ref)	0.92 (0.84, 1.00)	0.86(0.79, 0.93)	0.91(0.84, 0.98)	0.91 (0.83, 1.01)	0.138	0.99 (0.97, 1.00)
Multivariate RR (95% CI)3	1.00 (Ref)	1.01 (0.92, 1.11)	0.94 (0.86, 1.02)	0.96 (0.88, 1.05)	0.98 (0.88, 1.10)	0.607	0.99 (0.98, 1.01)
Fat cheese
Median intake (g/d)	0	8	19	37	56		Per 10 g/d
Person-years	4442	4217	5782	5886	2445
Number of cases	1261	1068	1260	1386	503
Age-adjusted RR (95% CI)	1.00 (Ref)	0.93 (0.86, 1.01)	0.86 (0.80, 0.93)	0.92 (0.85, 0.99)	0.87 (0.78, 0.96)	0.012	0.98 (0.97, 1.00)
Multivariate RR (95% CI)3	1.00 (Ref)	0.98 (0.90, 1.07)	0.93 (0.86, 1.01)	0.96 (0.88, 1.04)	0.92 (0.82, 1.03)	0.167	0.99 (0.97, 1.01)
Low-fat cheese
Median intake (g/d)	0	7					Per 10 g/d
Person-years	18,988	3784
Number of cases	4465	1013
Age-adjusted RR (95% CI)	1.00 (Ref)	1.09 (1.02, 1.16)					1.04 (1.01, 1.07)
Multivariate RR (95% CI)3	1.00 (Ref)	1.07 (0.99, 1.15)					1.02 (0.99, 1.06)
Butter
Median intake (g/d)	0	20					Per 10 g/d
Person-years	14,692	8081
Number of cases	3361	2117
Age-adjusted RR (95% CI)	1.00 (Ref)	1.03 (1.00, 1.12)					1.03 (1.01, 1.05)
Multivariate RR (95% CI)3	1.00 (Ref)	1.08 (1.01, 1.15)					1.04 (1.01, 1.06)
Low-fat dairy
Median intake (g/d)	0	63	171	268	439		Per 100 g/d
Person-years	4547	4562	4598	4835	4232
Number of cases	1257	1029	1052	1076	1064
Age-adjusted RR (95% CI)	1.00 (Ref)	0.91 (0.84, 0.99)	0.91 (0.84, 0.99)	0.89 (0.82, 0.97)	0.99 (0.91, 1.07)	0.956	1.01 (0.99, 1.02)
Multivariate RR (95% CI)3	1.00 (Ref)	0.97 (0.89, 1.06)	0.98 (0.90, 1.07)	0.98 (0.89, 1.07)	1.05 (0.96, 1.15)	0.246	1.01 (0.99, 1.03)
Fat from dairy
Median intake (g/d)	4.8	9.7	14.3	22.5	37.8		Per 10 g/d
Person-years	4481	4576	4585	4545	4586
Number of cases	1101	966	1051	1124	1236
Age-adjusted RR (95% CI)	1.00 (Ref)	0.95 (0.87, 1.03)	1.01 (0.93, 1.10)	1.03 (0.95, 1.12)	1.06 (0.97, 1.14)	0.036	1.02 (1.01, 1.04)
Multivariate RR (95% CI)3	1.00 (Ref)	0.96 (0.88, 1.06)	1.05 (0.96, 1.15)	1.07 (0.97, 1.18)	1.11 (0.99, 1.24)	0.024	1.04 (1.01, 1.06)
Calcium
Median intake (mg/d)	538	724	871	1030	1296		Per 100 mg/d
Person-years	4487	4585	4591	4544	4566
Number of cases	1217	1123	1000	1047	1091
Age-adjusted RR (95% CI)	1.00 (Ref)	0.94 (0.87, 1.02)	0.91 (0.84, 0.99)	0.94 (0.87, 1.02)	0.98 (0.90, 1.06)	0.772	1.00 (0.99, 1.01)
Multivariate RR (95% CI)3	1.00 (Ref)	0.98 (0.90, 1.08)	0.98 (0.89, 1.08)	1.00 (0.90, 1.10)	1.02 (0.91, 1.13)	0.682	1.01 (1.00, 1.02)

1

Q1–Q5, quintiles 1–5 based on distribution in the subcohort; C1–C5, categories; C1, non-users; RR, rate ratio; Ref, reference.

2

Wald test of linear estimate of variable containing median values of intake categories.

3

Adjusted for age (y), education (4 categories), cigarette smoking (status, number, and duration), nonoccupational physical activity (4 categories), BMI (4 categories), multivitamin use (yes or no), alcohol (5 categories plus former drinkers), energy (kJ), energy-adjusted mono- and polyunsaturated fat intakes (continuous), and vegetable and fruit consumption (continuous).

To investigate whether the observed inverse association of fermented full-fat milk (which mainly includes full-fat yogurt) could be explained by yogurt as such, we constructed a posteriori a yogurt variable by aggregating the available yogurt items: full-fat (3.6% fat) and skim yogurt (0.1% fat). A total of 75% of the women and 61% of the men consumed yogurt; yogurt eaters consumed, on average, 66 mL (men) and 71 mL (women) yogurt/d. In men, yogurt was inversely associated with mortality (P for trend = 0.039; RR_continuous: 0.96; 95% CI: 0.92, 1.00 per 100 mL/d), but less strongly than was fermented (full-fat) milk. Exclusion of the first 2 y of follow-up or prevalent cases with cardiovascular disease did not change these estimates. In women, no association between yogurt and mortality was observed (RR_continuous: 1.00; 95% CI: 0.95, 1.05 per 100 mL/d).

The addition of energy-adjusted saturated fat to the multivariate models did not attenuate the associations between milk and milk fat intake and total mortality (data not shown). Exclusion of BMI from the multivariate model did not change the estimates (data not shown), which indicated that it was not a strong intermediate variable in the pathway between dairy product consumption and mortality.

Results of the survival analyses for specific causes of death are presented in Tables 5 and 6. A total of 1997 (in men) and 692 (in women) deaths due to primary IHD were included in the multivariate analysis (Table 5). For men, no association with milk, low-fat dairy products, or calcium was found. Weak inverse, statistically significant associations were observed for fermented milk (RR_continuous: 0.75; 95% CI: 0.61, 0.92 per 100 mL/d) and fat from dairy products (RR_continuous: 0.94; 95% CI: 0.89, 1.00 per 10 g/d). These associations were slightly reduced after exclusion of prevalent cardiovascular disease patients. For women, milk (RR_continuous: 1.07; 95% CI: 1.00, 1.13 per 100 mL/d) and butter (RR_continuous: 1.11; 95% CI: 1.01, 1.21 per 10 g/d) consumption slightly increased the risk of IHD death, which was also reflected in the increased risk observed for consumption of low-fat dairy products, calcium, and fat from dairy products. All types of dairy products seemed to contribute to the increased risk, which did not appear to be caused by dairy fat as such. After exclusion of prevalent cardiovascular disease patients, the increased risks tended to diminish for milk, but not for butter. The addition of energy-adjusted saturated fat to the multivariate models did not attenuate the associations between milk and milk fat intake and IHD mortality (data not shown). The exclusion of BMI from the multivariate model did not change the estimates (data not shown), which indicated that it was not a strong intermediate variable in the pathway between dairy consumption and IHD mortality.

TABLE 5

Risk of death due to ischemic heart disease for dairy product variables in men and women: Netherlands Cohort Study 1986–19961

	Q1/C1	Q2/C2	Q3/C3	Q4/C4	Q5/C5	P for trend2	Continuous3
Men (1997 deaths)
Milk products	1.00 (Ref)	1.03 (0.82, 1.29)	1.08 (0.86, 1.35)	1.06 (0.84, 1.34)	1.03 (0.81, 1.31)	0.806	1.00 (0.97, 1.04)
Prevalent CVD excluded4	1.00 (Ref)	1.05 (0.80, 1.39)	1.15 (0.87, 1.52)	1.21 (0.91, 1.61)	1.03 (0.77, 1.39)	0.655	1.01 (0.97, 1.05)
Nonfermented full-fat milk	1.00 (Ref)	0.98 (0.79, 1.22)	1.10 (0.87, 1.41)	0.90 (0.71, 1.14)	1.02 (0.79, 1.31)	0.931	0.98 (0.93, 1.03)
Prevalent CVD excluded4	1.00 (Ref)	0.89 (0.68, 1.16)	1.05 (0.78, 1.42)	0.89 (0.67, 1.20)	1.08 (0.80, 1.46)	0.425	1.02 (0.96, 1.09)
Nonfermented low-fat milk	1.00 (Ref)	1.09 (0.90, 1.31)	1.15 (0.98, 1.36)			0.117	1.05 (1.00, 1.10)
Prevalent CVD excluded4	1.00 (Ref)	1.07 (0.85, 1.34)	1.06 (0.86, 1.29)			0.675	1.02 (0.96, 1.09)
Fermented full-fat milk	1.00 (Ref)	0.77 (0.64, 0.92)					0.75 (0.61, 0.92)
Prevalent CVD excluded4	1.00 (Ref)	0.85 (0.69, 1.05)					0.81 (0.64, 1.01)
Fermented low-fat milk	1.00 (Ref)	0.80 (0.67, 0.95)	0.93 (0.78, 1.11)			0.961	0.99 (0.92, 1.06)
Prevalent CVD excluded4	1.00 (Ref)	0.79 (0.63, 0.98)	0.93 (0.75, 1.15			0.927	0.98 (0.90, 1.07)
Cheese	1.00 (Ref)	0.86 (0.68, 1.10)	0.97 (0.77, 1.20)	1.02 (0.82, 1.27)	0.97 (0.74, 1.27)	0.639	1.01 (0.97, 1.05)
Prevalent CVD excluded4	1.00 (Ref)	0.86 (0.64, 1.15)	0.97 (0.74, 1.26)	0.92 (0.70, 1.20)	0.87 (0.62, 1.19)	0.487	0.97 (0.93, 1.02)
Fat cheese	1.00 (Ref)	0.76 (0.61, 0.96)	0.85 (0.68, 1.06)	0.88 (0.71, 1.09)	0.82 (0.62, 1.09)	0.584	0.99 (0.95, 1.03)
Prevalent CVD excluded4	1.00 (Ref)	0.75 (0.56, 0.99)	0.82 (0.62, 1.08)	0.82 (0.63, 1.07)	0.75 (0.53, 1.06)	0.313	0.97 (0.92, 1.02)
Low-fat cheese	1.00 (Ref)	1.21 (0.97, 1.51)					1.08 (0.98, 1.19)
Prevalent CVD excluded4	1.00 (Ref)	0.99 (0.74, 1.33)					1.00 (0.89, 1.13)
Butter	1.00 (Ref)	0.80 (0.67, 0.95)					0.95 (0.90, 1.01)
Prevalent CVD excluded4	1.00 (Ref)	0.86 (0.70, 1.06)					0.98 (0.91, 1.04)
Low-fat dairy	1.00 (Ref)	1.04 (0.83, 1.31)	1.24 (0.99, 1.55)	0.92 (0.73, 1.16)	1.18 (0.94, 1.49)	0.361	1.03 (0.99, 1.07)
Prevalent CVD excluded4	1.00 (Ref)	1.11 (0.84, 1.46)	1.24 (0.95, 1.62)	0.82 (0.62, 1.10)	1.15 (0.87, 1.51)	0.868	1.01 (0.96, 1.06)
Fat from dairy	1.00 (Ref)	0.96 (0.77, 1.20)	0.86 (0.68, 1.08)	0.79 (0.62, 1.00)	0.77 (0.58, 1.01)	0.029	0.94 (0.89, 1.00)
Prevalent CVD excluded4	1.00 (Ref)	1.02 (0.76, 1.35)	0.85 (0.64, 1.14)	0.77 (0.57, 1.04)	0.87 (0.62, 1.23)	0.272	0.96 (0.90, 1.03)
Calcium	1.00 (Ref)	0.90 (0.72, 1.13)	0.87 (0.69, 1.10)	1.00 (0.78, 1.27)	0.92 (0.71, 1.19)	0.787	1.01 (0.98, 1.04)
Prevalent CVD excluded4	1.00 (Ref)	0.87 (0.66, 1.15)	0.88 (0.66, 1.17)	1.00 (0.74, 1.35)	0.81 (0.59, 1.10)	0.339	1.00 (0.97, 1.03)
Women (692 deaths)
Milk products	1.00 (Ref)	0.85 (0.63, 1.15)	0.99 (0.74, 1.33)	1.02 (0.74, 1.39)	1.30 (0.95, 1.79)	0.050	1.07 (1.01, 1.13)
Prevalent CVD excluded4	1.00 (Ref)	0.82 (0.59, 1.14)	0.92 (0.67, 1.28)	0.84 (0.59, 1.19)	1.09 (0.77, 1.55)	0.551	1.03 (0.96, 1.09)
Nonfermented full-fat milk	1.00 (Ref)	0.91 (0.67, 1.23)	1.01 (0.76, 1.35)	0.92 (0.68, 1.25)	0.88 (0.64, 1.20)	0.455	1.02 (0.92, 1.13)
Prevalent CVD excluded4	1.00 (Ref)	0.90 (0.64, 1.27)	1.02 (0.74, 1.41)	0.96 (0.68, 1.35)	0.83 (0.59, 1.18)	0.344	0.98 (0.88, 1.10)
Nonfermented low-fat milk	1.00 (Ref)	0.97 (0.76, 1.24)	1.05 (0.84, 1.30			0.594	1.04 (0.97, 1.12)
Prevalent CVD excluded4	1.00 (Ref)	1.05 (0.80, 1.39)	1.06 (0.83, 1.35)			0.679	1.03 (0.95, 1.11)
Fermented full-fat milk	1.00 (Ref)	0.99 (0.79, 1.24)					0.93 (0.72, 1.19)
Prevalent CVD excluded4	1.00 (Ref)	1.01 (0.79, 1.29)					0.88 (0.66, 1.16)
Fermented low-fat milk	1.00 (Ref)	0.92 (0.72, 1.17)	1.19 (0.94, 1.51)			0.056	1.08 (1.01, 1.16)
Prevalent CVD excluded4	1.00 (Ref)	0.94 (0.72, 1.21)	1.08 (0.83, 1.41)			0.391	1.04 (0.96, 1.13)
Cheese	1.00 (Ref)	1.01 (0.72, 1.40)	1.04 (0.77, 1.40)	1.06 (0.78, 1.44)	1.01 (0.68, 1.50)	0.832	1.01 (0.95, 1.07)
Prevalent CVD excluded4	1.00 (Ref)	0.92 (0.64, 1.32)	0.94 (0.67, 1.30)	0.87 (0.62, 1.22)	0.76 (0.49, 1.20)	0.233	0.97 (0.90, 1.04)
Fat cheese	1.00 (Ref)	0.79 (0.59, 1.07	0.88 (0.66, 1.15)	0.90 (0.67, 1.20)	0.74 (0.49, 1.13)	0.423	0.97 (0.92, 1.04)
Prevalent CVD excluded4	1.00 (Ref)	0.77 (0.55, 1.08)	0.94 (0.69, 1.28)	0.80 (0.58, 1.11)	0.64 (0.40, 1.04)	0.132	0.95 (0.89, 1.02)
Low-fat cheese	1.00 (Ref)	1.33 (1.04, 1.71)					1.15 (1.03, 1.29)
Prevalent CVD excluded4	1.00 (Ref)	1.26 (0.95, 1.68)					1.07 (0.92, 1.24)
Butter	1.00 (Ref)	1.14 (0.91, 1.43)					1.11 (1.01, 1.21)
Prevalent CVD excluded4	1.00 (Ref)	1.22 (0.95, 1.57)					1.13 (1.02, 1.25)
Low-fat dairy	1.00 (Ref)	0.93 (0.69, 1.24)	0.92 (0.69, 1.24)	0.91 (0.68, 1.23)	1.37 (1.00, 1.87)	0.040	1.07 (1.01, 1.12)
Prevalent CVD excluded4	1.00 (Ref)	0.97 (0.70, 1.36)	0.87 (0.63, 1.20)	1.01 (0.72, 1.40)	1.20 (0.85, 1.70)	0.246	1.04 (0.98, 1.10)
Fat from dairy	1.00 (Ref)	0.94 (0.71, 1.27)	0.88 (0.65, 1.20)	1.08 (0.79, 1.47)	1.26 (0.88, 1.81)	0.106	1.11 (1.01, 1.22)
Prevalent CVD excluded4	1.00 (Ref)	0.84 (0.60, 1.17)	0.83 (0.59, 1.18)	0.97 (0.68, 1.37)	1.16 (0.78, 1.74)	0.218	1.08 (0.97, 1.20)
Calcium	1.00 (Ref)	0.87 (0.65, 1.17)	0.90 (0.66, 1.23)	1.15 (0.82, 1.60)	1.18 (0.84, 1.68)	0.168	1.05 (1.01, 1.09)
Prevalent CVD excluded4	1.00 (Ref)	0.73 (0.53, 1.02)	0.78 (0.56, 1.10)	0.90 (0.63, 1.30)	0.87 (0.59, 1.28)	0.764	1.00 (0.96, 1.05)

1

All values are rate ratios; 95% CIs in parentheses. CVD, cardiovascular disease; Ref, reference; Q1–Q5, quintiles 1–5 based on distribution in the subcohort; C1–C5, categories; C1, nonusers. Rate ratios (95% CIS) were adjusted for age (y); education (4 categories); cigarette, cigar (men only), and pipe (men only) smoking (status, number, duration); nonoccupational physical activity (4 categories); occupational physical activity (3 categories plus 1 missing, men only); BMI (4 categories); multivitamin use (yes or no); alcohol (5 categories plus former drinkers); energy (kJ); energy-adjusted mono- and polyunsaturated fat intakes (continuous); and vegetable and fruit consumption (continuous).

2

Wald test of linear estimate of variable containing median values of intake categories.

3

SeeTable 3 or 4 for increments.

4

Subjects with preexisting CVD at baseline (self-reported) were excluded.

TABLE 6

Risk of death due to stroke for dairy product variables: Netherlands Cohort Study 1986–19961

	Q1/C1	Q2/C2	Q3/C3	Q4/C4	Q5/C5	P for trend2	Continuous3
Men (520 deaths)
Milk products	1.00 (Ref)	1.01 (0.72, 1.43)	1.18 (0.85, 1.66)	1.14 (0.80, 1.61)	0.85 (0.58, 1.24)	0.551	0.98 (0.93, 1.03)
Prevalent CVD excluded4	1.00 (Ref)	0.92 (0.61, 1.38)	1.00 (0.67, 1.49)	0.90 (0.60, 1.37)	0.80 (0.52, 1.25)	0.343	0.97 (0.90, 1.04)
Nonfermented full-fat milk	1.00 (Ref)	1.01 (0.72, 1.40)	1.15 (0.79, 1.66)	1.21 (0.85, 1.73)	0.92 (0.61, 1.38)	0.667	0.96 (0.89, 1.03)
Prevalent CVD excluded4	1.00 (Ref)	0.99 (0.67, 1.46)	1.09 (0.70, 1.69)	1.05 (0.69, 1.62)	0.96 (0.60, 1.54)	0.840	1.01 (0.91, 1.11)
Nonfermented low-fat milk	1.00 (Ref)	0.97 (0.72, 1.30)	1.16 (0.90, 1.49)			0.199	1.05 (0.98, 1.13)
Prevalent CVD excluded4	1.00 (Ref)	0.91 (0.63, 1.29)	1.17 (0.87, 1.57)			0.206	1.03 (0.94, 1.12)
Fermented full-fat milk	1.00 (Ref)	0.86 (0.65, 1.15)					0.82 (0.60, 1.12)
Prevalent CVD excluded4	1.00 (Ref)	0.82 (0.59, 1.15)					0.75 (0.51, 1.09)
Fermented low-fat milk	1.00 (Ref)	0.82 (0.62, 1.08)	0.84 (0.64, 1.11)			0.382	0.93 (0.82, 1.06)
Prevalent CVD excluded4	1.00 (Ref)	0.97 (0.70, 1.33)	0.79 (0.56, 1.10)			0.130	0.85 (0.74, 0.99)
Cheese	1.00 (Ref)	0.79 (0.55, 1.15)	0.90 (0.64, 1.26)	0.95 (0.67, 1.33)	1.11 (0.75, 1.64)	0.403	1.02 (0.96, 1.08)
Prevalent CVD excluded4	1.00 (Ref)	0.76 (0.48, 1.19)	0.86 (0.58, 1.27)	0.92 (0.61, 1.37)	1.12 (0.70, 1.77)	0.407	1.02 (0.95, 1.09)
Fat cheese	1.00 (Ref)	0.68 (0.48, 0.97)	0.76 (0.54, 1.07)	0.81 (0.58, 1.14)	0.90 (0.60, 1.36)	0.993	1.00 (0.94, 1.06)
Prevalent CVD excluded4	1.00 (Ref)	0.57 (0.37, 0.87)	0.65 (0.43, 0.98)	0.74 (0.50, 1.10)	0.86 (0.53, 1.39)	0.888	0.99 (0.92, 1.06)
Low-fat cheese	1.00 (Ref)	1.56 (1.12, 2.16)					1.14 (0.99, 1.31)
Prevalent CVD excluded4	1.00 (Ref)	1.48 (0.99, 2.22)					1.15 (0.98, 1.35)
Butter	1.00 (Ref)	0.87 (0.67, 1.12)					0.97 (0.89, 1.05)
Prevalent CVD excluded4	1.00 (Ref)	0.87 (0.64, 1.18)					0.97 (0.88, 1.07)
Low-fat dairy	1.00 (Ref)	0.98 (0.70, 1.39)	1.10 (0.79, 1.55)	0.89 (0.62, 1.28)	1.07 (0.75, 1.51)	0.848	1.01 (0.95, 1.08)
Prevalent CVD excluded4	1.00 (Ref)	1.15 (0.77, 1.73)	1.20 (0.80, 1.79)	0.90 (0.59, 1.39)	1.03 (0.68, 1.55)	0.673	0.97 (0.90, 1.05)
Fat from dairy	1.00 (Ref)	0.99 (0.70, 1.40)	0.83 (0.58, 1.19)	1.04 (0.73, 1.49)	0.76 (0.49, 1.17)	0.262)	0.96 (0.88, 1.05)
Prevalent CVD excluded4	1.00 (Ref)	0.88 (0.57, 1.35)	0.75 (0.49, 1.16)	0.94 (0.61, 1.43)	0.70 (0.42, 1.17)	0.268)	0.96 (0.87, 1.07)
Calcium	1.00 (Ref)	1.13 (0.80, 1.59)	0.95 (0.67, 1.36)	0.99 (0.68, 1.44)	1.07 (0.72, 1.57)	0.944)	1.00 (0.96, 1.04)
Prevalent CVD excluded4	1.00 (Ref)	0.98 (0.65, 1.48)	0.72 (0.46, 1.11)	0.95 (0.60, 1.50)	0.91 (0.58, 1.43)	0.697)	0.99 (0.95, 1.04)
Women (322 deaths)
Milk products	1.00 (Ref)	1.13 (0.77, 1.65)	0.97 (0.65, 1.43)	0.77 (0.51, 1.18)	0.99 (0.65, 1.50)	0.557	0.97 (0.91, 1.04)
Prevalent CVD excluded4	1.00 (Ref)	1.05 (0.69, 1.60)	0.99 (0.65, 1.52)	0.72 (0.45, 1.14)	0.95 (0.60, 1.49)	0.498	0.97 (0.90, 1.05)
Nonfermented full-fat milk	1.00 (Ref)	0.79 (0.51, 1.23)	1.47 (1.01, 2.15)	1.06 (0.70, 1.59)	1.26 (0.85, 1.85)	0.215	1.10 (0.99, 1.23)
Prevalent CVD excluded4	1.00 (Ref)	0.75 (0.46, 1.23)	1.33 (0.87, 2.02)	1.02 (0.65, 1.59)	1.33 (0.87, 2.01)	0.094	1.11 (0.99, 1.24)
Nonfermented low-fat milk	1.00 (Ref)	0.96 (0.70, 1.32)	0.96 (0.72, 1.28)			0.806	0.96 (0.87, 1.05)
Prevalent CVD excluded4	1.00 (Ref)	0.96 (0.68, 1.38)	0.90 (0.65, 1.24)			0.528	0.96 (0.86, 1.06)
Fermented full-fat milk	1.00 (Ref)	0.81 (0.60, 1.10)					0.83 (0.60, 1.14)
Prevalent CVD excluded4	1.00 (Ref)	0.85 (0.62, 1.18)					0.82 (0.58, 1.16)
Fermented low-fat milk	1.00 (Ref)	1.13 (0.84, 1.51)	0.76 (0.55, 1.05)			0.032	0.92 (0.82, 1.03)
Prevalent CVD excluded4	1.00 (Ref)	1.12 (0.81, 1.54)	0.68 (0.47, 0.98)			0.012	0.91 (0.79, 1.04)
Cheese	1.00 (Ref)	0.93 (0.61, 1.41)	0.75 (0.51, 1.11)	0.84 (0.57, 1.26)	0.65 (0.39, 1.10)	0.169	0.94 (0.87, 1.02)
Prevalent CVD excluded4	1.00 (Ref)	0.88 (0.56, 1.41)	0.79 (0.51, 1.21)	0.83 (0.53, 1.29)	0.61 (0.34, 1.10)	0.163	0.94 (0.87, 1.03)
Fat cheese	1.00 (Ref)	0.93 (0.63, 1.36)	0.80 (0.55, 1.16)	0.88 (0.60, 1.30)	0.60 (0.35, 1.05)	0.137	0.94 (0.87, 1.01)
Prevalent CVD excluded4	1.00 (Ref)	0.99 (0.64, 1.51)	0.88 (0.58, 1.33)	0.93 (0.60, 1.43)	0.58 (0.31, 1.09)	0.164	0.94 (0.86, 1.02)
Low-fat cheese	1.00 (Ref)	1.11 (0.79, 1.55)					1.03 (0.88, 1.21)
Prevalent CVD excluded4	1.00 (Ref)	1.16 (0.79, 1.69)					1.03 (0.85, 1.25)
Butter	1.00 (Ref)	0.90 (0.66, 1.22)					1.05 (0.92, 1.19)
Prevalent CVD excluded4	1.00 (Ref)	0.90 (0.94, 1.27)					1.03 (0.89, 1.18)
Low-fat dairy	1.00 (Ref)	0.89 (0.61, 1.29)	1.04 (0.73, 1.50)	0.74 (0.50, 1.09)	0.77 (0.50, 1.17)	0.154	0.94 (0.88, 1.01)
Prevalent CVD excluded4	1.00 (Ref)	0.96 (0.64, 1.43)	0.95 (0.64, 1.43)	0.64 (0.41, 1.00)	0.77 (0.48, 1.22)	0.113	0.93 (0.85, 1.02)
Fat from dairy	1.00 (Ref)	0.73 (0.49, 1.07)	0.75 (0.50, 1.13)	0.77 (0.51, 1.17)	0.70 (0.44, 1.11)	0.286	1.01 (0.89, 1.15)
Prevalent CVD excluded4	1.00 (Ref)	0.72 (0.46, 1.13)	0.79 (0.50, 1.23)	0.76 (0.48, 1.22)	0.68 (0.40, 1.15)	0.286	0.99 (0.86, 1.14)
Calcium	1.00 (Ref)	1.04 (0.71, 1.52)	0.81 (0.54, 1.23)	0.91 (0.59, 1.40)	0.75 (0.47, 1.19)	0.186	0.96 (0.91, 1.01)
Prevalent CVD excluded4	1.00 (Ref)	0.99 (0.65, 1.52)	0.73 (0.46,1.16)	0.92 (0.58, 1.48)	0.73 (0.44, 1.22)	0.238	0.95 (0.90, 1.01)

1

All values are rate ratios; 95% CIs in parentheses. CVD, cardiovascular disease; Ref, reference; Q1–Q5, quintiles 1–5 based on distribution in the subcohort; C1–C5, categories; C1, nonusers. Rate ratios (95% CI) were adjusted for age (y); education (4 categories); cigarette, cigar (men only), and pipe (men only) smoking (status, number, and duration); nonoccupational physical activity (4 categories); occupational physical activity (3 categories plus 1 missing, men only); BMI (4 categories); multivitamin use (yes or no); alcohol (5 categories plus former drinkers); energy (kJ); energy-adjusted mono- and polyunsaturated fat intakes (continuous); and vegetable and fruit consumption (continuous).

2

Wald test of linear estimate of variable containing median values of intake categories.

3

SeeTable 3 or 4 for increments.

4

Subjects with preexisting CVD at baseline (self-reported) were excluded.

A total of 520 (in men) and 322 (in women) deaths due to primary stroke were included in the multivariate analysis (Table 6). In men and women, the associations between almost all dairy product variables and risk of stroke were virtually flat. In both sexes, a tendency for an inverse association of both fermented milk groups with risk of stroke was observed. Investigation of the yogurt variable showed a consistent inverse association with death due to stroke, which became stronger after the exclusion of cases with prevalent cardiovascular disease in men (P for trend = 0.007; RR_continuous: 0.68; 95% CI: 0.51, 0.91 per 100 mL/d) and in women (P for trend = 0.088; RR_continuous: 0.70; 95% CI: 0.54, 0.92 per 100 mL/d). Consumption of low-fat cheese increased (NS) the risk of death due to stroke in men, but the risk did not decrease after exclusion of prevalent cardiovascular cases.

In Table 7, the risk of death (all causes, IHD, and stroke) is compared with dairy product intake classified according to the prevailing dietary recommendation for dairy intake in the Netherlands. With the exception of a slightly increased risk of death in men with a dairy product intake higher than that recommended, no departures from null were observed.

TABLE 7

Risk of death according to compliance with recommendations for dairy product consumption: Netherlands Cohort Study 1986–19961

	Men			Women
	Too low	In compliance	Too high	Too low	In compliance	Too high
All-cause mortality
Age-adjusted RR	1.00 (Ref)	0.95 (0.91, 0.99)	0.99 (0.94, 1.05)	1.00 (Ref)	0.95 (0.90, 1.01)	1.03 (0.95, 1.11)
Multivariate RR2	1.00 (Ref)	1.00 (0.95, 1.05)	1.06 (1.00, 1.13)	1.00 (Ref)	0.99 (0.93, 1.06)	1.04 (0.95, 1.14)
Ischemic heart disease
Multivariate RR2	1.00 (Ref)	0.88 (0.68, 1.13)	1.02 (0.75, 1.39)	1.00 (Ref)	1.00 (0.80, 1.25)	1.17 (0.87, 1.57)
Prevalent CVD excluded3	1.00 (Ref)	0.91 (0.67, 1.23)	1.02 (0.72, 1.45)	1.00 (Ref)	0.84 (0.66, 1.28)	0.91 (0.65, 1.29)
Stroke
Multivariate RR2	1.00 (Ref)	0.88 (0.68, 1.13)	1.02 (0.75, 1.39)	1.00 (Ref)	0.93 (0.71, 1.24	0.81 (0.55, 1.21)
Prevalent CVD excluded3	1.00 (Ref)	0.91 (0.67, 1.23)	1.02 (0.72, 1.45)	1.00 (Ref)	0.90 (0.65, 1.23)	0.83 (0.53, 1.28)

1

All values are rate ratios (RRs); 95% CIs in parentheses. Values were calculated by using the following algorithm: 600 < milk/0.86 + 10 × cheese/0.64 < 1000 g/d. CVD, cardiovascular disease; Ref, reference.

2

Adjusted for age (y); education (4 categories); cigarette, cigar (men only), and pipe (men only) smoking (status, number, and duration); nonoccupational physical activity (4 categories); occupational physical activity (3 categories plus 1 missing, men only); BMI (4 categories); multivitamin use (yes or no); alcohol (5 categories plus former drinkers); energy (kJ); energy-adjusted mono- and polyunsaturated fat intakes (continuous); and vegetable and fruit consumption (continuous).

3

Subjects with preexisting CVD at baseline (self-reported) were excluded.

DISCUSSION

In the present study of the association between dairy product intake and mortality, only a few statistically significant, but mostly weak, associations were found. A slightly increased risk of all-cause mortality and mortality due to IHD was found for butter intake and fat intake from dairy, but only in women. In men, dairy product consumption above the recommended amount and a high calcium intake were associated with a slightly increased all-cause mortality. A small protective effect on all-cause mortality was found for fermented full-fat milk intake in both men and women. An additional analysis showed that this was not attributable to a high yogurt intake. Low-fat cheese intake was associated with a small increased risk of all-cause mortality and of stroke-specific mortality, which may be partly explained by men with cardiovascular disease at baseline having switched to low-fat cheese. Yogurt consumption was associated with a decrease in stroke mortality in both sexes.

These findings provide consistent evidence that dairy products are neither very harmful nor very beneficial to the risk of dying at a relatively young age, ie, well before the life expectancy of the cohort members at baseline. The results in men were in line with those of other cohort studies (16–20), although some studies found increased (21) or decreased (22) mortality associated with milk consumption. We found no cohort studies on dairy fat and all-cause mortality in women.

Our results on IHD only partly agree with the available evidence from cohort studies. The meta-analyses found summary RR estimates to be below one in men, but not in women (5, 23). The most recent review concluded that the evidence for a causal association between milk and coronary heart disease (CHD) was weak according to the Bradford Hill criteria (23). The studies in women found no association between dairy calcium intake and CHD mortality (24), total dairy product intake and IHD mortality (25), or high- and low-fat dairy product intake and CHD incidence (26), although high-fat dairy product intake tended to increase the risk slightly in both studies that looked at this (25, 26). The overall evidence for sex-specific associations between saturated fat intake and CHD is limited (27); therefore, we cannot readily explain why we observed a slightly positive association between dairy fat intake and IHD mortality in women, but not in men. An intervention study on the effects of trans fatty acids from ruminant sources on blood lipids, however, supports the hypothesis that the response in women may differ from that in men (28).

We observed virtually no associations between dairy product consumption and stroke mortality, with exception of fermented milk or yogurt. Several cohort studies observed a slightly decreased risk of stroke (death) associated with a high milk or dairy product intake or a high dairy calcium intake (24, 29–32); some of the results were not statistically significant (17, 22, 33). However, the Finnish Alpha-Tocopherol Beta-Carotene Study Cohort of men, which used a detailed dairy consumption assessment and had large statistical power, observed an increased risk of cerebral hemorrhage and ischemic stroke with whole milk consumption, a nonsignificant decreased risk of cerebral hemorrhage with sour milk and yogurt consumption, and a decreased risk of ischemic stroke with cheese and cream consumption (34). The decreased risks associated with dairy calcium consumption observed in 2 Japanese cohorts were similar for cerebral hemorrhage and ischemic stroke (24, 32) and for men and women (24).

More evidence is needed on the inverse relation of fermented full-fat milk and/or yogurt intake with all-cause mortality and what might be the underlying mechanism before a protective effect can be concluded. The same applies to the association between yogurt and risk of stroke. The presumed protective effect of dairy product consumption on blood pressure and hypertension might explain the association between low-fat dairy product consumption and stroke (7, 35), although this effect has not been shown consistently (36, 37). Whether and how fermented dairy product consumption affects this pathway is not yet clear. A meta-analysis of 12 randomized controlled trials conducted in Japan and Finland showed that specific tripeptides originating from fermentation of milk decreased blood pressure (38), but this effect was not confirmed in a more recent Dutch study (39).

The findings on low-fat cheese may be surprising and do not support the hypothesis that low-fat dairy products, as seen in the DASH trial, contribute to the prevention of stroke by reducing blood pressure (8). At the time of completion of the baseline questionnaire, consumption of low-fat cheese was not common and applied to only a small group of people, including heart disease patients who were recommended to reduce their saturated fat intake. The increased risk of IHD death in men indeed disappeared after exclusion of prevalent cases of cardiovascular disease. This was, however, not the case for the increased risk of death due to stroke. Our results did not support the hypothesis that BMI is an intermediate variable in the pathway between dairy consumption and risk of IHD and stroke.

One of the strengths of our study was that it included a comprehensive and detailed quantitative assessment of all types of dairy foods in a population characterized by a relatively high average dairy consumption but also by a substantial contrast in dairy consumption. Our findings were not likely to have been confounded by the major risk factors for chronic disease (ie, smoking, physical activity, socioeconomic status, and dietary risk factors), because we adjusted the estimates for these variables. The prospective design and the completeness of follow-up of our cohort ensured that information bias due to disease status and selection bias due to loss of follow-up were unlikely. In addition, the associations did not appear to be spuriously produced by symptoms of early, but not yet diagnosed disease, because the exclusion of deaths diagnosed within the first 2 y of follow-up did not change the results. Also, a possible change in diet as a consequence of preexisting cardiovascular conditions (self-reported myocardial infarction, angina pectoris, or stroke) did not appear to explain most results.

Our study also had some limitations. We tested many associations in this study and cannot rule out that some of the statistically significant associations were entirely due to chance. However, because of the large power of the study, the results that were statistically significant, but biologically insignificant, deserve equal caution. Measurement error can be substantial in dietary assessments. The validation study of our FFQ showed that it performs relatively well (11), but measurement error may still have attenuated the associations. Also, we had no possibility to update dietary intake data during the 10 y of follow-up, which may have resulted in some attenuation of associations. Furthermore, in addition to the well-known limitations of information derived from death certificates, cases that died of IHD or stroke may not be completely representative of all incident cases with IHD or stroke.

In conclusion, given the postulated risk-enhancing effects of dairy fat and the protective effects of low-fat dairy products, the role of dairy consumption in mortality generally appears to be neutral, particularly in men. In women, dairy fat intake was associated with slightly increased all-cause and IHD mortality rates. More research is warranted on the possible protective effect of yogurt intake on the risk of death due to stroke.

We are indebted to the participants of this study and thank the Dutch Central Bureau of Genealogy and the Dutch Central Bureau of Statistics for providing data on mortality. We also thank A Volovics and A Kester for statistical advice; S Nauta, S van de Crommert, H Brants, J Nelissen, C de Zwart, M Moll, W van Dijk, and A Pisters for operational and administrative assistance; and H van Montfort, T van Moergastel, L van den Bosch, R Schmeitz, and J Berben for programming assistance.

The authors’ responsibilities were as follows—RAG and AMJC: drafted the manuscript; FGG: conducted the data analyses; LJS: contributed to the design and conduct of the study; and RAG and PAvdB: conceived, designed, and conducted the study. All authors interpreted the data and reviewed the manuscript. The authors disclosed no conflicts of interest. The Dutch Dairy Foundation did not at any time contribute to the study design, analysis, interpretation, or reporting of the data and was not involved in the formulation of the conclusions.

FOOTNOTES

2

Supported by the Dutch Dairy Foundation with an unrestricted grant. The Dutch Dairy Foundation did not at any time contribute to the study design, analysis, interpretation or reporting of the data and was not involved in formulation of the conclusions.