New Adventist Health Study

The mortality studies of SDAs in the U.S. and elsewhere were useful in comparing risk of cancer death with non-SDA populations and for examining the relationship between diet and risk of several fatal cancers. However, the mortality studies were hindered by certain limitations. First, by restricting attention only to fatal outcomes, differences in stage at diagnosis, or differences in survival may have partly explained the differences between SDA and non-SDA populations. Second, in comparing cancer mortality with the general population, the possibility of selection bias may have obscured the true relationship between diet and cancer risk. Therefore, in the early 1970s, a new study of cancer incidence in California SDAs was designed in which risk of newly diagnosed cancers was to be measured and relationships between dietary intake and cancer incidence within the SDA population was to be evaluated.

The new Adventist Health Study was launched in 1974, when a census questionnaire was mailed to every known SDA household in the state of California. In 1976, a lifestyle questionnaire was mailed to all respondents to the first questionnaire who were 25 years of age and older. The second questionnaire included a 65-item food frequency recall component designed to measure three aspects of diet: (1) current use of specific foods, (2) past intake of specific foods, and (3) current intake of major nutrients.26

Table 4.1 Age- and Sex-Adjusted Relative Risks for Fatal Cancer Among SDA Age >35 Years, by Selected Dietary Variables, 1960-1980

Cancer

No. of Deaths

Sex

Foods

Frequency of Consumption

Relative Risk (95%CI)

Reference

Prostate

96

M

Meat

<1/day vs. >3/wk

1.3 (p>0.1)

Snowdon,

198425

99

Milk

<1 glass/day vs.

1.5 (p<0.1)*

>3 glasses/day

97

Cheese

<1/wk vs. >3/wk

1.4 (p>0.1)

97

Eggs

<1/day vs. >3/wk

1.3 (p>0.1)

Ovary

50

Eggs

<1/wk vs. >3/wk

3.0 (1.2-7.3)

Snowdon,

198524

Colon

52

M

Meat

<1/wk vs. 4/wk

1.5 (.7-3.3)

Phillips,

198522

54

Eggs

<2/wk vs. >5/wk

1.6 (.8-3.4)

55

Milk

<1/day vs.

0.5 (.2-1.1)

>3/day

54

Cheese

<1/wk vs. >3/wk

1.9 (1.0-3.6)

Colon

87

F

Meat

<1/wk vs. >4/wk

0.7 (.3-1.4)

88

Eggs

<2/wk vs. >5/wk

1.7 (.9-3.0)

91

Milk

<1/day vs.

1.1 (.5-2.2)

>3/day

88

Cheese

<1/wk vs. >3/wk

0.8 (.5-1.4)

Rectum

33

M/F

Meat

<1/wk vs. >4/wk

0.8 (.3-2.4)

33

Eggs

<2/wk vs. >5/wk

1.1 (.5-2.9)

33

Milk

<1/day vs.

1.2 (.6-2.7)

>3/day

33

Cheese

<1/wk vs. >3/wk

1.0 (.5-2.2)

Breast

142

F

Meat

None vs. >4/wk

1.15

Mills, 198823

(.53-2.53)**

Cheese

None vs. >3/wk

1.25 (.60-2.61

Milk

None vs. >3/wk

0.89

(.34-2.35)

Eggs

None vs. >3/wk

0.67

(.31-1.45)

  • Adjusted for age, education, weight and several foods.
  • Adjusted for age at menarche, age at first pregnancy, age at menopause, weight, education, and for several foods.
  • Adjusted for age, education, weight and several foods.
  • Adjusted for age at menarche, age at first pregnancy, age at menopause, weight, education, and for several foods.

Between 1976 and 1982, study of cancer incidence in the cohort of about 34,200 non-Hispanic white respondents to the lifestyle questionnaire was conducted by annual mailings to the study participants. Members of the cohort were asked to report any hospitalization in the previous year and study staff subsequently reviewed medical records for any evidence of a cancer diagnosis.27

Using an external reference population (i.e., Connecticut Tumor Registry) to generate expected numbers of cancer cases (adjusting for age, sex, and calendar year), the California SDA population was found to experience decreased cancer risk, although the reduction was more apparent in males than females. For all cancer sites combined, the standardized incidence ratio in males was 0.73, and for females it was 0.92; the result in females was of borderline statistical significance.28 (See Figure 4.10.) For most of the major cancer sites, the SDAs experienced low cancer risk (e.g., colon, lung) although for other cancer sites, risk was not substantially different in the SDAs (e.g., breast) or even somewhat elevated (e.g., prostate, corpus uteri).

Cancor Sito

Figure 4.10 Standardized incidence ratios for selected cancer sites in California SDA, 1976-1982 (Mills, 1994).

Cancor Sito

Figure 4.10 Standardized incidence ratios for selected cancer sites in California SDA, 1976-1982 (Mills, 1994).

Relationships between several foods and food groups have been examined in relation to cancer within the Adventist cohort for the years 1976 and 1982. Currently, results have been presented for breast, prostate, lung, colon, bladder, and pancreas cancer. (See Table 4.2.)

Table 4.2 Adjusted Relative Risks for Newly Diagnosed Cancers Among SDA >25 Years, By Selected Dietary Variables, 1977-1982

Cancer Site

Cases

Risk Enhanced

Rei. Risk

Risk Reduced

Rei. Risk

Comment

Ref.

Breast

215

Meat (never

1.33(.9-1.95)

Adjusted for several

Mills29

vs.>3/wk

covariates*

Cheese (<2/mo.

1.43 (.99-2.06)

vs.>3/wk.

Whole Milk (never

0.94(.66-1.33)

vs. daily)

Eggs (<1/wk vs.

1.07(.73-1.56)

>2/wk)

Lung

61

Meat (never

1.31 (.52-3.28)

Fruit (<3/wk vs.>2/day)

0.26(.1 -.7)

Adjusted for age,

Fraser30

vs.>2/wk)

sex, smoking

Poultry (never

2.20 (.84-5.77)

Green Salads (<3/wk

0.65 (.29-1.47)

vs.>1/wk)

vs.>7/wk)

Prostate

180

Current meat

1.41 (.79-2.51)

Beans, lentils, peas

0.53(.31-90)

Adjusted for several

Mills31

(never vs.>daily)

(<1/mo.vs.>3/wk)

covariates**

Current Fish

1.57 (.88-2.78)

Current Tomatoes

0.60(.37-97)

(never vs. >1/wk)

(<1/wk vs.>5/wk)

Raisins, dates, dryfruit

0.62(.36-1.06)

(<1/wk vs.>5/wk)

Soy milk (>1/day vs.

0.30(.1-1.0)

Jacobsen

<1/day)

46

Colon

145

Total meat (never

1.85

Legumes (never vs.

0.53(.33-86)

Adjusted for several

Singh32

  1. >1/wk) (1.16-2.87) >2/wk) covariates***
  2. >1/wk) (1.16-2.87) >2/wk) covariates***

Bladder

Pancreas

52 Meat, poultry, fish 2.38

Red meat (never 1.41 (.9-2.21) Nuts (never vs. >4/wk) 0.68 (.45-1.04) vs. >1/wk)

Sweetened Fruit Juice 0.34 (.11-1.11) Adjusted for Mills33

Cooked Green 0.77(.41-1.43) Vegetables (<2/wkvs. >1/day)

Vegetarian Protein 0.15(.03-89) Adjusted for several Mills34

Beans, lentils, peas ,03(.003-.24) (<1/wk vs.>3/wk)

Raisins, dates, dryfruit 0.19 9.04-.86) (<1/mo. Vs. >3/wk)

*Adjusted for age at entry, age at first pregnancy, age at menarche, menopausal status, history of benign breast disease, maternal history of breast cancer, education and weight.

"Adjusted for age, education and several other foods.

  • Adjusted for age, sex, weight, physical activity, parental history of colon cancer, smoking, alcohol and aspirin use.
  • Adjusted for age, sex, smoking, coffee, alcohol, urban vs. rural residence, and several other foods.
  • Adusted for age, sex, smoking and several other foods.

As with the analysis of fatal breast cancer, no strong relationship between consumption of animal products and breast cancer risk emerged from the new incidence study. After controlling for the effects of several covariates known to be associated with breast cancer risk (e.g., age at first menstrual period, age at first full-term pregnancy, and body weight), risks associated with meat intake were negligible. Comparing current use of meat, poultry and fish in those who consumed these products more than three times per week with never users, a non-significant relative risk of 1.33 was found, thus indicating that meat was only weakly associated with breast cancer in this study, if at all.29

When meat and poultry intake were evaluated in regard to lung cancer risk, the relative risks were somewhat elevated, though not significantly so. More noteworthy was the protective association between fruit and green salad intake and lung cancer risk. After taking into account smoking history, lung cancer risk was decreased 74% in those who frequently consumed fruit and a significant dose-response relationship was found.30 Prostate cancer risk bore a similar relationship to meat intake in that risk appeared to be somewhat elevated in the highest consumption categories (daily intake), yet the elevated risks were not as substantial (or statistically significant) as the protective associations seen with consumption of certain vegetables and fruits. In the prostate cancer analysis, a relative risk of 0.53 was noted for frequent consumption of beans, lentils, or peas and a relative risk of 0.60 was noted for frequent tomato con-sumption.31

For colon cancer, however, meat intake and, in particular, intake of both red meat and white meat (i.e. poultry, fish) bore significant associations with elevated risk. In addition, legumes (i.e., beans, lentils, split peas) were associated with a nearly 50% reduction in colon cancer risk. The increases in risk appeared to act synergistically in that men with high meat intake, heavy body weight, and low legume consumption experienced a threefold increase in colon cancer risk.32

Bladder cancer risk was also found to be positively associated with meat consumption (relative risk = 2.38) and inversely associated with fruit juice and cooked green vegetable consumption. These associations persisted after simultaneously controlling for smoking history (a risk factor for bladder cancer) as well as for all dietary variables.33

Pancreas cancer death was also evaluated in the incidence study. Because survival with this form of cancer is extremely poor, mortality is tantamount to incidence. When patterns of dietary intake were evaluated, risk of pancreas cancer was observed to decrease with increasing consumption of several foods commonly found in the vegetarian diet. Consumption of vegetarian protein products, legumes, and dried fruit all bore substantial reductions in risk of pancreas cancer. When meat consumption was included in the multivariate model, it was not associated with cancer risk although the vegetarian foods were all associated with decreased risk.34

My Life My Diet

My Life My Diet

I lost over 60 pounds and 4+ inches off my waist without pills, strenuous exercise, or any of the things that the diet experts tell you to do...and I did it in less than 4 months! If you have the desire, and can read through my e-book , then this is for you! I could have easily made it a lot more difficult, with stacks of information that people will never read, but why?

Get My Free Ebook


Post a comment