To review systematically the association between female reproductive factors of parity, female hormones, female hormonal factors, age at menopause, age at menarche, age at first birth, parity, menstrual cycle length, and use of exogenous female hormones (oral contraceptive pills and hormone replacement therapy) and lung cancer risk, taking into account factors such as smoking status, age, and lung tumour histology, looking at data from observational studies using a case-control, cohort or nested case-control, case-cohort approach and where available randomized controlled trials.

Sources searched: PubMed; Web of Science; EMBASE.

No language or publication year restriction.

Include: case-control, cohort or nested case-control, case-cohort approach and randomized controlled trials.

Exclude: cross sectional studies, ecological studies, case series, studies without control group or comparison group; studies without informative effect estimates and unable to contact authors; studies looking at survival or prognosis of a lung cancer cohort; studies conducted in special cohorts .

Many studies have evaluated association between female reproductive factors of parity, age at menopause, age at menarche, age at first birth, menstrual cycle length and use of exogenous female hormones (oral contraceptive pills and hormone replacement therapy) and lung cancer, but results have been inconsistent and difficult to understand. It is possible that differences in study population in terms of other factors (e.g., smoking status and age) may have contributed to the inconsistent results

General female population living within the community.

Level of exposure to female hormones, as measured by the following proxies:

a) age at menarche;

b) age at menopause;

c) menstrual cycle lengths;

d) age at first birth;

e) parity; and

f) exogenous female hormone exposure (contraceptive or replacement therapy).

Non–exposed groups (exogenous exposures); lowest category of exposure for female reproductive factors.

General population setting, women at average risk of lung cancer.

Primary outcomes

Lung cancer incidence or mortality.

Most factors studied are natural biological parameters; with regard to exogenous hormone use – where possible, we will obtain timing and duration of exposure.

Effect measures studied will be odds ratios, hazards ratios or relative risk ratios, depending on the study design.

Secondary outcomes

None.

Extracted articles from the different databases will be combined, with duplicates removed. A title scan will be conducted to remove articles that were unrelated to the objective of this meta-analysis.

Exclusion criteria for the title review are:

the title indicates that the subject matter of the article is:

a) not about lung cancer;

b) about metastatic lesions to the lung;

c) about treatment of lung cancer, or prognosis or survival of lung cancer;

d) about studies conducted on cells, including human cell-lines, rather than humans;

e) about studies looking at ecological associations.

This will be performed by a single investigator.

All articles selected from title review will then undergo a full text review to check that the study fulfils the criteria. In the event that the full text is unavailable, a review of the abstract will be performed. Articles where neither the full text nor abstract are avaialble will not be included if we are unable to contact the author for the paper.

i) Two authors will evaluate articles to determine eligibility; in event of uncertainty, a 3rd author will be consulted to arrive at a consensus on which articles will be included;

ii) data extraction performed independently by 2 authors; discrepancies resolved by repeating review and reaching a consensus;

iii) where there are more than 1 published article of the same study population, data from the latest article will be extracted

Publication bias will be assessed using Egger's test and funnel plot.

Aggregate data will be used with quantitative synthesis planned. Assessment and analysis, will be performed for the following exposures: parity, age at first birth, age at menarche, age at menopause, menstrual cycle length and duration of hormonal contraception or replacement therapy use. For every study, the mean or median value of the exposure of interest for each exposure category will be assigned to each corresponding effect estimates (OR/RR). If the median or mean per category is not reported in the article, the midpoint of the upper and lower boundaries of each exposure category will be used. If the lower or upper boundary for the lowest and highest category is not reported, a dose value which is at least as high as the lower bound plus the width of the second-highest category will be assigned to the uppermost, open-ended category. For the lower open-ended category, a value of its upper bound minus half the width of the next (second-to-lowest) category will be assigned. To assess the shape of the dose-response relationship, all available data points from each study will be plotted; further, appropriate statistical methods will be employed to estimate dose-response associations. For linear dose-response relationship, Generalized Least Squares for Trend estimation will be used to estimate the log-linear dose-response associations. This will be a two stage approach that estimates a slope/ beta-coefficient for each study at first stage, and then derives an overall estimate of dose-response effect by weighted average of the individual slopes. For non-linear dose-response relationship, a restricted cubic spline model will be estimated using generalized least square regression taking into account the correlation within each set of reported effect estimates (OR/RR). Further, the study-specific estimates will be combined using the restricted maximum likelihood method in a multivariate random-effects meta-analysis. A probability value for nonlinearity will be calculated by testing the null hypothesis that the coefficient of the second spline is equal to 0.

Smoking status and the different histologic types of lung cancer.

Publication in an international refereed journal.

Vincent Tan

Saw Swee Hock School of Public Health

National University of Singapore

MD3, 16 Medical Drive

SINGAPORE 117597

ephtyv@nus.edu.sg

National University of Singapore, Saw Swee Hock School of Public Health

www.sph.nus.edu.sg

Dr Wei Yen Lim, Saw Swee Hock School of Public Health
Dr Agus Salim, Saw Swee Hock School of Public Health
Mr Vincent Tan, Saw Swee Hock School of Public Health

06 August 2012

31 March 2013

None

None known

English

Singapore

Subject indexing assigned by CRD

Female; Humans; Lung Neoplasms; Parity

None

23 August 2012

23 August 2012