The claims of exposure are not authentic.  Exposure is not measured.  The studies actually measure nothing but rely on the vague and grossly imprecise recall of queried subjects who attempt to evoke in a few minutes their individual lifetime memories of passive smoking exposure.

Errors in individual exposure recollection, most likely large, are unknown, and are unknowable.  Digitized numerical claims of exposure are, therefore, incongruous and impermissible.  Their numerical representation give an impression of reliability and precision that is demonstrably false and misleading.

A recall bias has been demonstrated to be larger in subjects who are more likely to amplify their recall of passive smoking exposure as a justification for their disease, with lung cancer or cardio vascular disease. (remember when you or your wife was pregnant-remember seeing all the pregnant women you never noticed before?)

A misclassification bias has been demonstrated to be larger in subjects with lung cancer or cardio vascular disease because they are more likely than healthy subjects to classify themselves as non smokers.

A mismatch error of cases and controls is inevitable because the groups compared are not homogeneous and differ in many characteristics other than recall of passive smoking exposure.

Confounding errors by definition are likely to be more prevalent among lung cancer and cardio vascular disease cases.  Confounders are all other known and unknown potential causes of lung cancer and cardio vascular disease that interfere with the specific attribution of risk to passive smoking.

Probable errors of disease diagnosis are seldom addressed by passive smoking studies.

Publication errors have been found to favor the publication of studies that claim associations of increased risk.

Statistical errors of sampling and statistical significance are grossly inconsistent among passive smoking studies owing to the feeble differentials of exposure recall and the small number of subjects in each study.  A majority of studies have not reached statistical significance.  In any event, significant or not, the statistical indices of all passive smoking studies are illusory because they are derived from the grossly illusory and misleading numerical renderings of vague individual exposure recalls.

 Results from different studies have not ben consistent and reproducible.

 Epidemologic criteria of causal inference are not met by passive smoking studies.