Milk and cardiovascular disease (CVD).

Milk and total cardiovascular disease.

Because little data was available, results about the relation with CVD risk and CVD mortality were combined.
Data about the relation between total milk consumption and total CVD was provided by 2 cohorts, including 688 cases (Elwood PC 2004; Bonthuis M 2010). In addition, 3 articles provided information about dairy fat.

Results: No significant associations were found of high vs low total milk consumption.
Modifying effect of milk fat: In one cohort the association between both skim/low-fat milk, and whole milk with CVD death was examined (Anderson LF 2006). Risks of high vs low consumption were identical (HR = 0.98).
In another cohort, the effects on CVD risk from total milk and whole milk were identical (Elwood PC 2005).
In a third cohort the ratio of full fat vs reduced fat milk was examined (Jamrozik K 2000). Drinking less reduced fat milk and more full fat milk significantly reduced CVD risk.
Effect modification: No data was found.

Conclusion: No significant associations were found between milk consumption and total CVD. No evidence was found for a modifying effect of milk fat.

**Prospective studies of total milk and total cardiovascular disease (CVD):**
Author	Cohort name	Cases	End point	Relative Risk (RR)
Bonthuis M (2010)	No cohort name defined	60	Mortality	HR = 0.81 (0.35-1.87; P = 0.84).
Elwood PC (2004)	The Caerphilly Cohort	628	Risk	HR = 0.64 (0.39-1.06; P = 0.16).
		Total number of cases: 688		Average HR = 0.65

Milk and coronary heart disease (CHD).

Milk and CHD risk.

Data about the relation between total milk and CHD risk was provided by 5 cohorts, incuding 910+ cases (no data about the amount of cases was available from 2 cohorts).

Results: No (non)significant associations were found.
Modifying effect of milk fat: Data about whole milk was provided by 2 cohorts. A significantly increased risk of whole milk was found in one cohort, while a protective effect was found of skim milk consumption (Hu FB 1999). A nonsignificantly increased risk of whole milk was found in another cohort (Fraser GE 1994).
In contrast, in a third cohort a significantly increased risk was found for consumption of low fat milk instead of high fat milk (Holmberg S 2009).
And in a fourth cohort, the effects from total milk and whole milk were identical (Elwood PC 2005).
Effect modification: No data was found.

Conclusion: No associations were found between total milk consumption and CHD risk. The average RR = 0.74, but this could be based on data from 2 cohorts only, because of missing RR's or missing amounts of cases. Findings about the associations for whole milk vs skim/low fat milk were inconsistent.
No evidence was found for an association between milk consumption and CHD risk, and no evidence was found for a modifying effect of milk fat.

Milk and CHD mortality.

Data about the relation between total milk and CHD mortality was provided by 6 cohorts, incuding 7,437 cases.

Results: A significant protective effect was found among men in one cohort (Snowdon DA). No other associatons were found.
Modifying effect of milk fat: No significant difference in effect was found between consumption of semiskimmed, or skimmed milk and whole milk in one cohort (Whiteman D 1999).
Effect modification: No data was found.

Conclusion: Few associations were found. No evidence was found for an association between milk consumption and CHD mortality (Average RR = 0.96).

**Prospective studies of total milk and CHD risk:**
Author	Cohort name	Cases	Relative Risk (RR)
28) Holmberg S (2009)	The NSHDS	291	No significant association.
18) Elwood PC (2004)	The Caerphilly Study	493	HR = 0.71 (0.40-1.26; P = 0.48).
7) Abbott RD (1996)	The Honolulu Heart Program	Not defined	No significant association.
4) Shaper AG (1991)	The British Regional Heart Study	126	RR = 1.13 for low vs high consumption.
2) Tanaka H (1987)	The Shibata Study	Not defined	RR = 0.97
		Total number of cases: 910 + X	Average RR = 0.74

**Prospective studies of total milk and CHD mortality:**
Author	Cohort name	Cases	Relative Risk (RR)
25) van der Pols JC (2009)	The Boyd Orr Cohort	376	HR = 0.80 (0.49-1.31; P = 0.83).
19) Iso H (2007)	The JACC Study	576 men, and 383 women	Men: HR = 0.96 (0.79-1.17). Women: HR = 0.83 (0.65-1.06).
15) Ness AR (2001)	No cohort name defined	892	RR = 0.68 (0.40-1.13; P = 0.11).
10) Mann JI (1997)	The Oxford Vegetarian Study	63	DRR = 150 (81-278; P = NS).
3) Hirayama T (1990)	No cohort name defined	3,548	RR = 1.03 (0.96-1.10).
1) Snowdon DA (1984)	The Adventist Health Study	758 men, and 841 women	Men: RR = 0.94 (P = < 0.05). Women: RR = 1.11 (P = NS)
		Total number of cases: 7,437	Average RR = 0.96

Milk and stroke.

Milk and stroke risk.

Total stroke risk: Data about total stroke risk was provided by 1 cohort, including 53 cases (See extended version of table: Elwood PC 2005). A significant protective effect was found at above median consumption of total milk.
Ischemic stroke risk: Data about ischemic stroke risk in relation with total milk was provided by 3 cohorts. A significant protective effect was found in one cohort (Abbott RD). No other associations were found (Excluding incomplete data from Abbott RD [7]: Average RR = 0.71).

Modifying effect of milk fat: Whole milk consumption showed a significant trend of an increased risk of cerebral infarction, and a significantly increased risk of intracerebral hemorrhage in one cohort. While no associations were found with low-fat and sour milk (Larsson SC 2009).
In contrast, no difference in effect on total stroke incidence was found between total- and full fat milk in another cohort (Elwood PC 2005). And no associations between skim or whole milk and intracerebral hemorrhage were found in a third cohort (Iso H 2001). Effect modification: No modifying effects of physical activity and smoking were found on the relation between milk and ischemic stroke in one cohort (Abbott RD 1996).

Conclusion: Some protective effects were found across different stroke outcomes, but these findings were done among cohorts of very small size. Inconsistent effects were found of milk fat.
No evidence was found for an association between milk consumption and stroke risk. And no evidence was found for a modifying effect of milk fat.

Milk and stroke mortality.

7 articles, providing information about 6 different cohorts were found, including 14,819 cases.

Results: Significant protective effects were found in 2 cohorts among both men and women, including 13,163 cases (Kinjo Y; Iso H). No other associations were found. Both cohorts existed of Japanese subjects only. Protective effects were found at the level of consumption of ≥ 4-5 times/week.
Modifying effect of milk fat: No data was found.
Effect modification: Effects did not differ stratified by smoking status and rural/urban areas in one cohort, but protective effects were stronger among high meat consumers for all stroke subtypes (Kinjo Y).

Conclusion: Significant protective effects of milk consumption were found in 2 cohorts of large-very large size (including 89% of all cases).
High consumption of milk possibly protects against total stroke mortality (- 20%). This effect was found at milk consumption ≥ 4-5 times/week. Though all RRs were below 1, the evidence was restricted to findings from Japanese cohorts (Average RR = 0.80, also).

**Prospective studies of total milk and ischemic stroke risk:**
Author	Cohort name	Cases	Relative Risk (RR)
18) Elwood PC (2004)	The Caerphilly Study	185	HR = 0.66 (0.24-1.81; P = 0.23).
7) Abbott RD (1996)	The Honolulu Heart Program	229	A significant protective effect.
5) Iso H (1999)	The Nurses' Health Study	386	RR = 0.74 (0.51-1.06; P = 0.44)
		Total number of cases: 800	Average RR = 0.71

**Prospective studies of total milk and total stroke mortality:**
Author	Cohort name	Cases	Relative Risk (RR)
25) van der Pols JC (2009)	The Boyd Orr Cohort	121	HR = 0.60 (0.28-1.33; P = 0.26).
19) Iso H (2007)	The JACC Study	1,148 men,and 985 women	Men: HR = 0.77 (0.67-0.89; P = < 0.01). Women: HR = 0.83 (0.72-0.97; P = < 0.05).
17) Sauvaget C (2003)	The Hiroshima/Nagasaki Life Span Study	1,094	HR = 0.94 (0.79-1.12; P = 0.232).
15) Ness AR (2001)	No cohort name defined	196	RR = 0.84 (0.31-2.30; P = 0.58).
9) Ross RK (1997)	No cohort name defined	245	No significant association.
3) Kinjo Y (1999)	No cohort name defined	11,030	RR = 0.79 (0.75-0.83).
		Total number of cases: 14,819	Average RR = 0.80

All men
(628 cases)

Men with prior IHD or stroke
( cases)

HR = 0.64 (0.39-1.06; P = 0.19) for the highest vs lowest quartile of consumption.

Amount specific data (L per day):
No milk: HR = 1.
≤ 0.285: HR = 0.90 (0.64-1.26).
> 0.285-< 0.57: HR = 0.89 (0.63-1.26).
≥ 0.57: HR = 0.64 (0.39-1.06).

HR = 0.37 (0.15-0.90; P = 0.11) for the highest vs lowest quartile of consumption.

Amount specific data (L per day):
No milk: HR = 1.
≤ 0.285: HR = 0.76 (0.48-1.22).
> 0.285-< 0.57: HR = 0.77 (0.47-1.27).
≥ 0.57: HR = 0.37 (0.15-0.90; P = 0.03).

Skim and low-fat milk

whole milk

HR = 0.98 (0.84-1.14; P = 0.87) for the highest vs lowest quintile of consumption.

Amount specific data (servings/wk):
0: RR = 1.
0.5-1: RR = 0.95 (0.78-1.15).
3-5.5: RR = 0.91 (0.78-1.07).
7: RR = 0.91 (0.78-1.06).
≥ 17.5: RR = 0.98 (0.84-1.14).

HR = 0.98 (0.83-1.16; P = 0.58) for the highest vs lowest tertile of consumption.

Amount specific data (servings/wk):
0: RR = 1.
0.5-1: RR = 1.06 (0.88-1.27).
≥ 3: RR = 0.98 (0.83-1.16).

Men:
(236 cases)

Women:
(55 cases)

Cases consumed the same amount (200 g/d) as controls (196 g/d).

Cases consumed the same amount (160 g/d) as controls (160 g/d).

All men
(493 cases)

Men with prior IHD or stroke
(202 cases)

HR = 0.71 (0.40-1.26; P = 0.48) for the highest vs lowest quartile of consumption.

Amount specific data (L per day):
No milk: HR = 1.
≤ 0.285: HR = 0.94 (0.65-1.37).
> 0.285-< 0.57: HR = 0.96 (0.65-1.41).
≥ 0.57: HR = 0.71 (0.40-1.26).

HR = 0.40 (0.14-1.13; P = 0.44) for the highest vs lowest quartile of consumption.

Amount specific data (L per day):
No milk: HR = 1.
≤ 0.285: HR = 0.83 (0.49-1.39).
> 0.285-< 0.57: HR = 0.93 (0.54-1.62).
≥ 0.57: HR = 0.40 (0.14-1.13).

Whole milk

Skim milk

RR = 1.67 (1.14-1.90; P = < 0.0001) for the highest vs lowest quintile of consumption.

Amount specific data (servings/day):
Almost never: RR = 1.0.
< 244 g/wk (1 glass/wk): RR = 1.20.
488-1464 g/wk (2-6 glasses/wk): RR = 1.17.
244 g/d (1 glass/d): RR = 1.48.
≥ 488 g/d (2 glasses/d): RR = 1.67.

RR = 0.78 (0.63-0.96; P = 0.09) for the highest vs lowest quintile of consumption.

Amount specific data (servings/day):
Almost never: RR = 1.0.
< 244 g/wk (1 glass/wk): RR = 0.83.
488-1464 g/wk (2-6 glasses/wk): RR = 0.77.
244 g/d (1 glass/d): RR = 0.89.
≥ 488 g/d (2 glasses/d): RR = 0.78.

Use of milk:

Tea or coffee only
(243 cases)

Drink
(156 cases)

In cereal
(155 cases)

In cereal plus drink
(126 cases)

RR = 1.12 for use of milk in tea or coffee only vs use in cereal plus drink.

RR = 1.06 for use of milk as drink vs use in cereal plus drink.

RR = 1.00 for use of milk in cereal vs use in cereal plus drink.

RR = 1.13 for no use of milk vs use in cereal plus drink.

Men:

Women:

HR = 0.96 (0.79-1.17) for the highest vs lowest tertile of consumption.

Amount specific data:
< 3/w: HR = 1.
1-4/w: HR = 0.96 (0.77-1.20).
≥ 5/w: HR = 0.96 (0.79-1.17).

HR = 0.83 (0.65-1.06) for the highest vs lowest tertile of consumption.

Amount specific data:
< 3/w: HR = 1.
1-4/w: HR = 0.97 (0.74-1.27).
≥ 5/w: HR = 0.83 (0.65-1.06).

Men:

Women:

RR = 0.94 (No 95% CI; P = < 0.05) for 2 glasses/day vs none.

RR = 1.11 (No 95% CI; P = nonsignificant) for for 2 glasses/day vs none.

Stratified by stroke subtype:

Cerebral infarction
(2,702 cases)

Intracerebral hemorrhage
(383 cases)

Subarachnoid hemorrhage
(196 cases)

RR = 1.04 (0.92-1.18; P = 0.60) for the highest vs lowest quintile of consumption.

Amount specific data (g/d):
64: RR = 1.
148: RR = 0.97 (0.86-1.10).
244: RR = 1.09 (0.96-1.24).
418: RR = 1.00 (0.88-1.13).
783: RR = 1.04 (0.92-1.18).

RR = 0.91 (0.64-1.30; P = 0.62) for the highest vs lowest quintile of consumption.

Amount specific data (g/d):
64: RR = 1.
148: RR = 1.06 (0.77-1.47).
244: RR = 1.06 (0.76-1.47).
418: RR = 1.23 (0.89-1.70).
783: RR = 0.91 (0.64-1.30).

RR = 1.31 (0.81-2.12; P = 0.48) for the highest vs lowest quintile of consumption.

Amount specific data (g/d):
64: RR = 1.
148: RR = 1.17 (0.72-1.91).
244: RR = 1.31 (0.81-2.13).
418: RR = 1.03 (0.62-1.71).
783: RR = 1.31 (0.81-2.12).

Stratified by stroke subtype:

Cerebral infarction
(2,702 cases)

Intracerebral hemorrhage
(383 cases)

Subarachnoid hemorrhage
(196 cases)

RR = 1.08 (0.95-1.23; P = 0.04) for the highest vs lowest quintile of consumption.

Amount specific data (g/d):
0: RR = 1.
170: RR = 0.98 (0.86-1.12).
340: RR = 1.17 (1.03-1.33).
510: RR = 1.20 (1.05-1.37).
850: RR = 1.08 (0.95-1.23).

RR = 1.41 (1.02-1.96; P = 0.05) for the highest vs lowest quintile of consumption.

Amount specific data (g/d):
0: RR = 1.
170: RR = 1.21 (0.89-1.66).
340: RR = 1.03 (0.72-1.46).
510: RR = 1.17 (0.80-1.69).
850: RR = 1.41 (1.02-1.96).

RR = 1.24 (0.80-1.92; P = 0.50) for the highest vs lowest quintile of consumption.

Amount specific data (g/d):
0: RR = 1.
170: RR = 1.01 (0.63-1.63).
340: RR = 1.21 (0.76-1.94).
510: RR = 0.71 (0.39-1.32).
850: RR = 1.24 (0.80-1.92).

Stratified by stroke subtype:

Cerebral infarction
(2,702 cases)

Intracerebral hemorrhage
(383 cases)

Subarachnoid hemorrhage
(196 cases)

RR = 0.99 (0.89-1.12; P = 0.92) for the highest vs lowest quintile of consumption.

Amount specific data (g/d):
0: RR = 1.
16: RR = 1.07 (0.95-1.20).
74: RR = 0.97 (0.86-1.09).
183: RR = 1.10 (0.98-1.24).
412: RR = 0.99 (0.89-1.12).

RR = 0.80 (0.59-1.10; P = 0.11) for the highest vs lowest quintile of consumption.

Amount specific data (g/d):
0: RR = 1.
16: RR = 1.07 (0.80-1.45).
74: RR = 0.93 (0.69-1.25).
183: RR = 0.92 (0.67-1.25).
412: RR = 0.80 (0.59-1.10).

RR = 1.25 (0.81-1.92; P = 0.39) for the highest vs lowest quintile of consumption.

Amount specific data (g/d):
0: RR = 1.
16: RR = 1.12 (0.71-1.75).
74: RR = 1.23 (0.80-1.88).
183: RR = 1.20 (0.76-1.87).
412: RR = 1.25 (0.81-1.92).

All men
(185 cases)

Men with prior IHD or stroke
(65 cases)

HR = 0.66 (0.24-1.81; P = 0.23) for the highest vs lowest quartile of consumption.

Amount specific data (L per day):
No milk: HR = 1.
≤ 0.285: HR = 1.24 (0.62-2.48).
> 0.285-< 0.57: HR = 1.07 (0.53-2.16).
≥ 0.57: HR = 0.66 (0.24-1.81).

HR = 0.37 (0.07-2.03; P = 0.08) for the highest vs lowest quartile of consumption.

Amount specific data (L per day):
No milk: HR = 1.
≤ 0.285: HR = 0.95 (0.37-2.45).
> 0.285-< 0.57: HR = 0.61 (0.22-1.71).
≥ 0.57: HR = 0.37 (0.07-2.03).

Men:

Women:

HR = 0.77 (0.67-0.89; P = < 0.01) for the highest vs lowest tertile of consumption.

Amount specific data:
< 3/w: HR = 1.
1-4/w: HR = 0.83 (0.72-0.97; P = < 0.05).
≥ 5/w: HR = 0.77 (0.67-0.89).

HR = 0.83 (0.72-0.97; P = < 0.05) for the highest vs lowest tertile of consumption.

Amount specific data:
< 3/w: HR = 1.
1-4/w: HR = 0.95 (0.80-1.12).
≥ 5/w: HR = 0.83 (0.72-0.97).

All cerebrovascular disease
(11,030 cases)

Cerebral haemorrhage
(4,773 cases)

Cerebral embolism and thrombosis
(4,084 cases)

RR = 0.79 (0.75-0.83) for the highest vs lowest tertile of consumption.

Amount specific data (Frequency of consumption/week):
< 1: RR = 1.
1-3: RR = 0.89 (0.84-0.94).
≥ 4: RR = 0.79 (0.75-0.83).

RR = 0.74 (0.68-0.80) for the highest vs lowest tertile of consumption.

Amount specific data (Frequency of consumption/week):
< 1: RR = 1.
1-3: RR = 0.85 (0.78-0.92).
≥ 4: RR = 0.74 (0.68-0.80).

RR = 0.85 (0.77-0.92) for the highest vs lowest tertile of consumption.

Amount specific data (Frequency of consumption/week):
< 1: RR = 1.
1-3: RR = 0.92 (0.84-1.00).
≥ 4: RR = 0.85 (0.77-0.92).

All cerebrovascular disease.
Milk consumption ≥ 1 vs < 1 time/week.
Joint effects with meat and fish consumption:

Low meat, low fish
(Meat < once/week, and fish < 4 times/week)

Low meat, high fish
(Meat < once/week, and fish ≥ 4 times/week)

High meat, low fish
(Meat ≥ 1 time/week, and fish < 4 times/week)

High meat, high fish
(Meat ≥ 1 time/week, and ≥ 4 times/week)

< 1 RR = 1.00 (reference).
≥ 1: RR = 0.92 (0.82-1.02).

< 1 RR = 0.89 (0.81-0.97).
≥ 1: RR = 0.84 (0.73-0.95).

< 1 RR = 0.93 (0.87-0.99).
≥ 1: RR = 0.80 (0.74-0.86).

< 1 RR = 0.86 (0.80-0.92).
≥ 1: RR = 0.68 (0.63-0.74).

Cerebral haemorrhage.
Milk consumption ≥ 1 vs < 1 time/week.
Joint effects with meat and fish consumption:

Low meat, low fish
(Meat < once/week, and fish < 4 times/week)

Low meat, high fish
(Meat < once/week, and fish ≥ 4 times/week)

High meat, low fish
(Meat ≥ 1 time/week, and fish < 4 times/week)

High meat, high fish
(Meat ≥ 1 time/week, and ≥ 4 times/week)

< 1 RR = 1.00 (Reference).
≥ 1: RR = 0.89 (0.75-1.05).

< 1 RR = 0.95 (0.82-1.09).
≥ 1: RR = 0.73 (0.58-0.91).

< 1 RR = 0.94 (0.85-1.03).
≥ 1: RR = 0.79 (0.70-0.87).

< 1 RR = 0.84 (0.75-0.93).
≥ 1: RR = 0.63 (0.55-0.70).

Cerebral embolism and thrombosis.
Milk consumption ≥ 1 vs < 1 time/week.
Joint effects with meat and fish consumption:

Low meat, low fish
(Meat < once/week, and fish < 4 times/week)

Low meat, high fish
(Meat < once/week, and fish ≥ 4 times/week)

High meat, low fish
(Meat ≥ 1 time/week, and fish < 4 times/week)

High meat, high fish
(Meat ≥ 1 time/week, and ≥ 4 times/week)

< 1 RR = 1.00 (Reference).
≥ 1: RR = 0.92 (0.77-1.10).

< 1 RR = 0.90 (0.78-1.05).
≥ 1: RR = 0.98 (0.79-1.19).

< 1 RR = 0.95 (0.85-1.05).
≥ 1: RR = 0.84 (0.74-0.94).

< 1 RR = 0.94 (0.84-1.06).
≥ 1: RR = 0.79 (0.70-0.89).

Prospective studies of milk cardiovascular diseases other than CHD and stroke:

Author

Cohort name

Subjects

Years of follow-up

Cases

End point

Consumption of

Relative Risk (RR)

Adjustments

3) Hirayama T (1990)

Not defined.

265,118 adults (122,261 men, 142,857 women) aged ≥ 40 from 6 prefectures in Japan.

17
(1966-1982)

Hypertension: (Number of cases not clearly defined).
Arterial sclerosis: (169? men, and 135? women).
Aneurysm mortality: (Number of cases not clearly defined).
Arterial embolism: (14? men, and 6? women).

Cardiovascular & Cerebrovascular disease mortality

Milk & goat milk

Hypertension: RR = 0.83 (90% CI = 0.66-1.03; No P-value).

Arterial sclerosis: RR = 0.83 (90% CI = 0.65-1.06; No P-value).

Aneurysm mortality: RR = 1.11 (90% CI = 0.77-1.61; No P-value).

Arterial embolism: RR = 1.07 (90% CI = 0.58-1.98; No P-value).

RRs are for consumption ≥ 4 vs ≤ 3 times/wk.

Age and sex.

Cardiovascular disease & Diet. A systematic review.

Literature review.

Relating various aspects of our
diet to heart disease and stroke.

Milk and cardiovascular disease (CVD).

Milk and total cardiovascular disease.

Milk and coronary heart disease (CHD).

Milk and CHD risk.

Milk and CHD mortality.

Milk and stroke.

Milk and stroke risk.

Milk and stroke mortality.

Cardiovascular disease & Diet. A systematic review.

Literature review.

Relating various aspects of ourdiet to heart disease and stroke.

Milk and cardiovascular disease (CVD).

Milk and total cardiovascular disease.

Milk and coronary heart disease (CHD).

Milk and CHD risk.

Milk and CHD mortality.

Milk and stroke.

Milk and stroke risk.

Milk and stroke mortality.

Relating various aspects of our
diet to heart disease and stroke.