Our data confirm a clear relationship between BMI and age at thelarche and menarche, as is reported in the literature [2, 5,6,7,8]. They also suggest that the Z-score change between BW and BMI is able to influence the timing of puberty and age at menarche, confirming the relationship between earlier pubertal maturation and increased BMI or adiposity in girls  and indicating the importance of the early weight gain from birth in the timing of the appearance of these phenomena.
Interestingly, the variation in BMI between age at thelarche and menarche and the Z-score change between BW and BMI at thelarche also has a relationship with the ‘tempo’ of puberty, suggesting that early weight gain between BW and BMI at thelarche also influences the duration of puberty. Among the numerous factors that could influence the timing and tempo of pubertal development are foetal nutrition, birth weight, childhood dietary habits, physical activity, psychological factors, exposure to electromagnetic fields (EMF) and/or endocrine disrupting, the activating effect of leptin on GnRH—gonadotropin axis and hyperinsulinemia related to obesity .
Many data have demonstrated an association between adiposity and early puberty, particularly in girls; for example, a Pediatric Research in Office Settings (PROS) study showed that 6–9-year-old girls with breast development had higher BMI z-scores compared to prepubertal girls , and that the prevalence of excess weight is significantly higher in girls with early puberty than in controls . The National Health Examination Survey (NHES 1963–1970) and National Health and Nutrition Examination Survey III (NHANES 1988–1994), examining trends in menarche age over 25 years, reported that age at menarche decreased from 12.7 to 12.5 years whereas the percentage of girls classed as overweight increased from 16 to 27% . Early menarche is more prevalent in overweight girls than in normal weight girls , suggesting a causality between increasing obesity and lower age at menarche.
Unfortunately, many of these studies are cross-sectional. Longitudinal studies investigating the association between changes in body composition and pubertal timing and tempo in girls are very poor , because they relate to only one pubertal parameter  or use self-reported data .
Our data do not show a significant correlation between age at thelarche and adult height, supporting previous data in the literature , although there is a negative relationship between the change in BMI Z-score from thelarche to menarche and adult height. An increase in BMI during puberty is related to a reduced adult height, probably due to the reduced duration of puberty in these girls.
Some data clearly show that higher BMI Z-scores in infancy and childhood are associated with faster length/height velocity in early life, while higher BMI z-scores during mid-childhood are associated with slower length/height velocity during adolescence . This may be particularly important for girls with fast puberty, considering that puberty may follow a nonlinear trajectory with significant interindividual differences of timing and tempo [2, 28]. Further studies will be necessary to better evaluate this finding.
Our study is limited by the fact that we did not have data on weight change for patients with precocious puberty. Future studies would ideally include reliable measurements of body fat mass as well as BMI, hormonal markers for pubertal development and be longitudinal to overcome genetic, nutritional and environmental variables.
In conclusion, our data confirm a significant relationship between BMI and age at B2 and menarche in Italian girls, as well as a clear relationship among weight changes during the first years of life, age at thelarche and menarche and the duration of puberty. There is increasing evidence that environmental factors explain consistent and continuous variations in the onset and tempo of puberty.