Comparison of aspects of smoking among the four histological types of lung cancer

BACKGROUND

The magnitude of the link between cigarette smoking and lung cancer may vary by histological type.

METHODS

We used polytomous logistic regression to evaluate whether aspects of smoking have different effects across four histological types in the Nurses' Health Study.

RESULTS

From 1976 to 2002, we identified 1062 cases of lung cancer: squamous cell (n = 201), small cell (n = 236), adenocarcinoma (n = 543) and large cell carcinoma (n = 82), among 65 560 current or former smokers. Risk reduction after quitting ranged from an 8% reduction (relative risk (RR): 0.92, 95% CI 0.91 to 0.94) to a 17% reduction (RR: 0.83, 95% CI 0.80 to 0.86) per year for adenocarcinoma and small cell carcinoma, respectively, with a 9% reduction observed for large cell carcinoma and an 11% reduction observed for squamous cell carcinoma. The association of age at smoking initiation and former cigarette smoking was similar across types, while the association of smoking duration differed. The risk of adenocarcinoma increased by 6% per year of smoking, compared to 7% for large cell, 10% for squamous cell and 12% for small cell. The 6% difference between adenocarcinoma and small cell carcinoma is equivalent to a 3.2 to 9.7-fold increase in risk for 20 years of smoking.

CONCLUSIONS

The effects of the number of cigarettes smoked per day and years since quitting smoking are different across the major types of lung cancer, which are fully appreciated at long durations of smoking and smoking cessation. Smoking prevention and cessation should continue to be the focus of public health efforts to reduce lung cancer incidence and mortality.

Validity of a food frequency questionnaire in assessing nutrient intakes of low-income pregnant women

OBJECTIVE

In 1989, a validation study of eight nutrients was performed on a modified food frequency questionnaire (FFQ) specifically designed for low-income pregnant women. The purpose of this study was to broaden the scope of the previous study by assessing the validity of the FFQ for 17 additional nutrients.

METHODS

The Pregnancy Food Frequency Questionnaire (PFFQ) was administered to a sample of 295 low-income, pregnant women aged 14-43 years living in Massachusetts. A randomly selected subsample of 101 women who provided at least one diet recall and reported intake of less than 4,500 calories were included in this analysis.

RESULTS

Mean intake of 25 nutrients as assessed by one administration of the PFFQ and up to three diet recalls collected over 1 month were similar. Unadjusted correlation coefficients between nutrient intake measured by diet recalls and the questionnaire ranged from .28 (carotene) to .61 (folate). After adjusting for energy intake the correlations ranged from .03 (B12) to .46 (folate). The correlations corrected for day-to-day variation were higher, ranging from .07 (B12) to .90 (zinc). The mean correlation was .47 and there were 54% over .40.

CONCLUSIONS

A food frequency questionnaire for English-speaking, low-income, pregnant women can provide maternal and child health practitioners and researchers a valid estimate of diet across a wide range of nutrients.

Validation of a food frequency questionnaire in Native American and Caucasian children 1 to 5 years of age

OBJECTIVE

To assess the validity of the Harvard Service Food Frequency Questionnaire (HFFQ) in the diet assessment of Native American and Caucasian children 1 to 5 years of age participating in the North Dakota WIC program.

METHODS

The 84-item HFFQ was administered twice to the parent or guardian of 131 Native American and 102 Caucasian children ages 1 to 5 years (total n = 233), first at the child's routine WIC visit and then following the completion of three 24-hr dietary recalls taken over approximately 1 month. Average nutrient intakes from the three 24-hr dietary recalls were compared to average nutrient intakes from the HFFQs by calculating Pearson correlation coefficients and adjusting for energy intake and within person variation.

RESULTS

Correlation coefficients ranged from 0.26 for dietary fiber to 0.63 for magnesium. The average correlation was 0.52, similar to that found in validation studies among adolescents and adults. The following nutrients had correlations of 0.50 or greater: carbohydrate, sucrose, total fat, vitamin C, vitamin E, vitamin B1, vitamin B2, niacin, folate, vitamin B6, calcium, magnesium, and iron.

CONCLUSIONS

The HFFQ is a simple self-administered questionnaire completed by the child's parent or guardian and is useful in assessing the diets of Native American and Caucasian children. It may also provide important nutritional information about this age group for future program planning, research, education, and intervention purposes.