E Second World War, from October 1944 till April 1945, inhabitants of the occupied Western part of the Netherlands were exposed to famine. Their daily food rations dropped to less than 25 of the pre-famine rations and varied between 400?00 kcal/day [18]. After approximately 6 months of hunger the famine ended abruptly by purchase Tenapanor liberation of the Netherlands in May 1945, and food became available again through supplies of the allied forces. This short period of extreme hunger allows the study of long-term effects of famine exposure.The Prospect-EPIC cohortWe investigated the association between famine exposure and an unhealthy lifestyle in the Prospect-EPIC cohort. This is one of two Dutch cohorts of the European Prospective Investigation into GDC-0084 web Cancer and Nutrition [19, 20]. Between 1993 and 1997 17,357 women were recruited in the Prospect-EPIC cohort. They all participated in the nationwide breast cancer screening program and were living in the city of Utrecht or surroundings. At recruitment, the women completed a general questionnaire (containing among others three questions about exposure to the 1944?945 famine) and a validated food frequency questionnaire [21, 22], and underwent a physical examination. All participants provided written informed consent before study inclusion. The Prospect-EPIC study complies with the Declaration of Helsinki and was approved by the Institutional Review Board of the University Medical Center Utrecht.Exclusion criteriaWe excluded participants who answered `not applicable’ or `I don’t know’ to one or more of the three famine exposure questions (n = 4975). Furthermore, we excluded women who were born after the Dutch famine (n = 2559) or who were >18 years during the famine (N = 481), or who lived outside the Netherlands during the famine (n = 1732), or who had no dietary information available (n = 85). Our final study population consisted of 7,525 women.Individual famine scoreParticipants were asked about their experience of hunger and weight loss during the famine [3]. The questions each contained the answer categories `hardly’, `little’, and `very much’. These categories were combined into a three-point famine exposure score, as previously reported: 1) severely exposed: women who reported being `very much’ exposed to both hunger and weight loss; 2) unexposed: women who reported `hardly’ being exposed to both hunger and weight loss; and 3) moderately exposed: all others [3].Exposure age categoriesWe divided women into two age categories, using age at start of the famine (October 1st, 1944), because we wanted to investigate the effect of famine exposure during different growth periods. These categories were made according to the human life cycle as defined by Bogin [23] and have been used in the Prospect-EPIC cohort before [4]: 0? years (childhood, n = 4385), and 10?7 years (adolescence, n = 3140).Unhealthy lifestyle factorsSmoking. Information on smoking status and smoking intensity was available from the general questionnaire at recruitment (1993?). Smoking status was defined as current, formerPLOS ONE | DOI:10.1371/journal.pone.0156609 May 31,3 /Famine Exposure and Unhealthy Lifestyle Behavioror never smoker (categorical). Pack years of smoking (continuous) were calculated as packs (25 cigarettes) smoked per day multiplied by years of smoking. Pack years were available and analyzed for current and former smokers. Alcohol consumption. Information on alcohol consumption from the baseline questionnaire (being a.E Second World War, from October 1944 till April 1945, inhabitants of the occupied Western part of the Netherlands were exposed to famine. Their daily food rations dropped to less than 25 of the pre-famine rations and varied between 400?00 kcal/day [18]. After approximately 6 months of hunger the famine ended abruptly by liberation of the Netherlands in May 1945, and food became available again through supplies of the allied forces. This short period of extreme hunger allows the study of long-term effects of famine exposure.The Prospect-EPIC cohortWe investigated the association between famine exposure and an unhealthy lifestyle in the Prospect-EPIC cohort. This is one of two Dutch cohorts of the European Prospective Investigation into Cancer and Nutrition [19, 20]. Between 1993 and 1997 17,357 women were recruited in the Prospect-EPIC cohort. They all participated in the nationwide breast cancer screening program and were living in the city of Utrecht or surroundings. At recruitment, the women completed a general questionnaire (containing among others three questions about exposure to the 1944?945 famine) and a validated food frequency questionnaire [21, 22], and underwent a physical examination. All participants provided written informed consent before study inclusion. The Prospect-EPIC study complies with the Declaration of Helsinki and was approved by the Institutional Review Board of the University Medical Center Utrecht.Exclusion criteriaWe excluded participants who answered `not applicable’ or `I don’t know’ to one or more of the three famine exposure questions (n = 4975). Furthermore, we excluded women who were born after the Dutch famine (n = 2559) or who were >18 years during the famine (N = 481), or who lived outside the Netherlands during the famine (n = 1732), or who had no dietary information available (n = 85). Our final study population consisted of 7,525 women.Individual famine scoreParticipants were asked about their experience of hunger and weight loss during the famine [3]. The questions each contained the answer categories `hardly’, `little’, and `very much’. These categories were combined into a three-point famine exposure score, as previously reported: 1) severely exposed: women who reported being `very much’ exposed to both hunger and weight loss; 2) unexposed: women who reported `hardly’ being exposed to both hunger and weight loss; and 3) moderately exposed: all others [3].Exposure age categoriesWe divided women into two age categories, using age at start of the famine (October 1st, 1944), because we wanted to investigate the effect of famine exposure during different growth periods. These categories were made according to the human life cycle as defined by Bogin [23] and have been used in the Prospect-EPIC cohort before [4]: 0? years (childhood, n = 4385), and 10?7 years (adolescence, n = 3140).Unhealthy lifestyle factorsSmoking. Information on smoking status and smoking intensity was available from the general questionnaire at recruitment (1993?). Smoking status was defined as current, formerPLOS ONE | DOI:10.1371/journal.pone.0156609 May 31,3 /Famine Exposure and Unhealthy Lifestyle Behavioror never smoker (categorical). Pack years of smoking (continuous) were calculated as packs (25 cigarettes) smoked per day multiplied by years of smoking. Pack years were available and analyzed for current and former smokers. Alcohol consumption. Information on alcohol consumption from the baseline questionnaire (being a.