Our study results showed that a higher intake of poultry was associated with a higher risk of early menarche among school girls in Shanghai. The relationship remained after controlling for the girls’ BMI, age, physical activity, sleep, household income and parental education level. Neither the intake of pork, beef, lamb, processes meat nor total meat was associated with menarche age, after adjusting for BMI, age, physical activity, sleep, household income and parental education.
To our knowledge, this is the first report revealing an association between poultry intake and risk of early menarche. One potential explanation of the results is hormone abuse during poultry feeding. According to media reports, oestrogen as a ‘ripening agent’ is widely used for accelerating the growth of chicken on the ‘black’ market(29). In 2009 the Guangzhou Centre for Disease Control and Prevention reported a detection rate as high as 22.30% in poultry for diethylstilboestrol, a type of synthetic oestrogen(30). In 2001, 2003, 2004 and 2006, the sex hormone residue detection rates in poultry were 26%, 25%, 37.84% and 22.73%, respectively. The mean diethylstilboestrol residue in poultry was 1.24 mg/kg, and maximum residue can be up to 4.98 mg/kg(30). Chinese poultry products are usually sold nationwide, thus the above data from Guangzhou can also be considered representative for other big cities, including Shanghai.
Epidemiological studies strongly support the hypothesis that early or prepubertal exposure to oestrogenic compounds may induce human pubertal development(31; 32; 33; 34; 35). An investigation into an alarming incidence of premature menarche and precocious puberty from 1978 to 1984 in Puerto Rico suggested that early sexual development could be caused by exogenous oestrogen contamination in meat ingested by children(31). A multicentre analysis of CPP distribution in north-west Tuscany, Italy, revealed a significantly higher annual CPP incidence in the Viareggio area than in other north-west Tuscany areas. The geographic distribution of CPP strongly suggested the involvement of environmental oestrogen exposure in CPP onset(33). That early menarche may be caused by hormone residues in poultry is also supported by our results showing that higher intake of poultry was associated with a higher risk of early menarche. Further studies are needed to confirm the finding, following which related government agencies must implement efficient measures to end hormone abuse in poultry feeding. Chinese Ministry of Agriculture Announcement number 235 was, in fact, promulgated to prohibit the use of synthetic hormones (trenbolone acetate and zeranol) in animal fattening. Thus, stronger and more efficient governmental supervision and management are required to avoid the illegal use of sex hormones in the poultry industry.
In our study, the total intake of meat was not significantly associated with early menarche. Consistently, a prospective study of 5 583 US girls and a cross-sectional study of 777 schoolgirls in Spain also concluded that the risk of early menarche was not associated with meat intake(24; 25). In the US prospective study, red meat (including pork, beef, lamb) as a whole was found not associated with age at menarche(24). Likewise, there were no statistically significant associations between age at menarche and consumption of pork, beef, lamb respectively in our study, after adjusting for covariates. However, several other studies have reported that higher intakes of meat were associated with earlier menarche(15; 16; 36; 37). A British cohort study reported that those with the highest meat intake (>8 portions/week) at 7 years had a 1.75 (95% CI: 1.25–2.44) times higher risk of early menarche, compared with girls in the lowest meat intake category (<4 portions/week)(15). In a cross-sectional study of 422 Korean children, breast development was significantly positively associated with the factor score of ‘shellfish and processed meat’ after adjusting for age, body fat, and bone mineral content(22). Additionally, a recent prospective study reported that higher red meat intake frequency during childhood was associated with an earlier age at menarche. Girls who consumed red meat twice or more a day had a 64% higher probability of earlier age at menarche, compared with girls whose intake of red meat was less than four times a week(38). There are several reasons to explain the different results. Firstly, the classification of intake frequency of meat differed with our study. The British cohort study categorized the meat intake into five levels(15), whereas total meat intake was categorized into tertiles in our study. Secondly, outcome measures differ between studies. The longitudinal Dortmund Nutritional and Anthropometric Longitudinally Designed Study used age at menarche as a continuous variable to analyse the association(36) and the outcome of the British cohort study was whether menarche was reached before or after 12 years 8 months, which was likely to be as near as possible to the median age at menarche of their study population(15). Breast development was also used as the outcome in some studies(22; 39). Thirdly, differences in dietary assessment may be another reason. A food frequency questionnaire combined with a 3-day dietary record was used in the British cohort study(15). The Korean cross-sectional study derived dietary patterns from 3-day food records. The food preference ‘shellfish and processed meat’ in the Korean cross-sectional study was vague; our study analysed the intake of specific categories of meat from a food frequency questionnaire. Notably, previous studies have yielded mixed results for different dietary assessments and different classifications of outcome. Further studies to confirm the association between age at menarche and total meat intake as well as red meat intake are required.
Our study reported that the mean age at menarche was 11.97±1.24 years and 26.10% of the girls had early menarche. Owing to the two-stage random sampling design, the result could be representative of the entire Shanghai school girl population, and even that of all large Chinese cities. The reported average age at menarche was 0.3 years earlier than that of urban Chinese girls in 2012 (12.28 years)(40), and also earlier than in some developed countries, including Denmark (13.13 years)(5) and the United Kingdom (12.9 years)(41). It may be time for China to take action to stop this worrying trend.
Some limitations are of note: the FFQ has not been formally validated. However the FFQ was modified from the validated semi-quantitative food frequency questionnaire which was used for China National Nutrition and Health Survey. The China National Nutrition and Health Survey is a nationally representative cross-sectional study on nutrition and non-communicable chronic diseases. It covered all 31 provinces, autonomous regions, and municipalities directly under the Chinese central government (except Taiwan, Hong Kong, and Macao). The FFQ has been used for decades. Many previous studies provided strong evidence for its validity(42; 43). Next the study design was cross-sectional and there was unavoidable recall bias when surveying with food frequency questionnaires. As the information of ‘age at menarche’ as well as ‘sleep time’ and ‘physical activity’ were also self-reported, they were all also subject to recall bias. And Data quality control was guaranteed by a high standard of training of the field interviewers.