Another design factor is that the NMIHS/LF data were collected with self-administered, written questionnaires which did not contain prompts or definitions of terms, so our findings may not be generalizeable to other data collection methods, such as those that use a symptom- or consequences-based approach. However, other population-based health surveys such as the National Health Interview Survey (NHIS) or the medical history portion of the National Health and Nutrition Examination Survey (NHANES) that use a checklist to collect information on chronic conditions often do not include definitions of the terms or lists of related symptoms to provide a consistent definition across subjects. As a consequence, maternal- or self- reported health information from those studies may well be subject to similar types of biases. Health examination surveys (e.g., NHANES) will detect only current symptoms, whereas health interviews can capture both current and previous symptoms.