Our prospective study suggests that girls with higher intake of milk, experience menarcheal age earlier compared to those with a lower intake. This association was observed independently of energy, protein and BMI Z-score at baseline. Controlling for height Z-score at baseline made the association non-significant, suggesting that height may have mediating effect on the association between milk intake and age at menarche.
Few studies have been conducted to examine the association between milk intake and age at menarche 
. In accordance to our finding, data of girls aged 9-12 years, from NHANES 1994-2004 showed that individuals in the middle tertile of milk intake, < 244 g/day, had a lower risk of early menarche than did those in the highest tertile (> 245 g/day). In addition, girls with intakes between 100-229 Kcal/day from dairy had higher risk of early menarche than those with intakes > 230 Kcal/day after controlling for height and overweight status 
. In a recently published study on a representative Hong Kong Chinese birth cohort, cow’s milk consumption was not associated with age at pubertal onset in children aged 7-14 years 
; however, in the study, the effect of milk was evaluated as frequency of milk intake and total energy intake was not controlled for. This discrepancy in findings from observational studies may be due to variation in study design or extent of adjustment for potentially confounding factors. Milk supplements in pre-menarche girl, aged 10–12 years, showed no significant effect on age at menarche 
. Since it has been proposed that diet in early to mid-childhood is more strongly related to the occurrence of menarche than diet later in childhood 
, it is likely that milk supplementation at ages of close to time of menarche did not affect age of menarche in the trials. It is also possible that the amount of milk supplementation in these studies was not high enough to influence the age of menarche in interventions as compared to controls.
One of the possible mechanisms by which milk intake can accelerate growth and development, might be due, at least in part, to increased insulin-like growth factor-1(IGF-1) concentration. IGF-1 is an important mitogen peptide that can increase estrogen production through stimulating the adrenal androgen secretion 
or enhancing the secretion of gonadotropin-releasing hormone (GnRH) by hypothalamic neurons 
. Cow’s milk contains IGF-1, which can be enhanced by administrating of recombinant growth hormone to cows in order to increase milk production 
. In addition, several identified nutrients of milk including protein, Ca, Mg, and P or some yet unknown bioactive components can stimulate IGF-1 production 
In the present study, higher intakes Ca, Mg, and P were related consistently to the risk of earlier menarche. The findings suggest that Ca, Mg, and P are some components of milk, which might be involved in the menarche-promoting effect observed. Previous observational studies regarding the effect of nutrients including Ca and Mg provide inconsistent results 
. In a randomized controlled trial, girls who received 850 mg/day as milk-Ca-P extract from age 7.9–8.9 years, experienced menarche 5.4 months earlier than the placebo groups. Also, total calcium intake during the intervention was inversely correlated with menarcheal age, even after adjusting for baseline body weight and height 
. Low calcium intake was associated with a decrease in all aromatizable adrenal androgens and an increase in sex hormones binding globulin (SHBG); leading to delayed pubertal development in girls, aged 11–14 years 
. There is also some evidence that Mg and P are involved in modulation of anabolic hormones secretion including testosterone and IGF-1 in middle-aged and elderly 
. The relationship between Mg, P, and sex-steroid hormones in children and adolescents deserve investigation to ascertain their potential effects on puberty.
We have observed an opposite effect for cheese intake and risk of age at menarche ≤ 12 years as compared to milk intake. The lower risk of age at menarche ≤ 12 years in the middle tertile of cheese intake as compared to the lowest was significant in model 1 and 2. Furthermore, yogurt intake had no association with the risk of earlier age at menarche. Dairy products are a complex group of foods and the compositions can be changed based on their processing; for instance, previous findings suggest that IGF-1 in milk is less influenced by homogenization and pasteurization than the processing of milk to dairy products 
. Therefore, it is possible that IGF-1 or substances stimulating the endogenous production of IGF-1 or even unknown components in milk, responsible for promoting age at menarche, be inactivated during processing 
Attaining age at menarche earlier has diverse impacts on health 
; earlier age of menarche increases the risk of breast cancer later in life
, on the other hand it is associated with higher bone mineral density and lower risk of osteoporosis 
. Hence, before public health recommendations can be made, long-term health consequences of reaching menarche earlier in those with higher intakes of milk need to be elucidated.
A major strength of the current study is that we could prospectively investigate the associations between mid-childhood consumption of dairy and age of menarche. The information on maternal age at menarche obtained from the TLGS study is another strength of the study, which made it possible to control for potential genetic influences on menarche. To the best of our Knowledge, this is a first time that the effects of other dairy products including cheese and yoghurt on age at menarche were examined individually. Furthermore, we followed our participants long enough for all of them to reach menarche.
This study has some limitations. First, dietary intakes were used only at baseline; however, it has been shown that dietary intakes tend to track throughout childhood into adolescence 
. Therefore, we consider dietary intakes at baseline as a usual, established pattern of intake. Second, dietary intakes were determined by using 2-day 24-hour food recalls that may have assessed dairy and mineral intakes less precisely than if measured on 7- day 24-hour food recalls or if the FFQ were used; however, this may have underestimated the true effects so it is unlikely to be an explanation for the associations observed. Third, socioeconomic status (SES) as a potential confounding factor was not taken into account. In Iran, age at menarche is related to SES 
. It is not clear whether milk and dairy intakes are also socially patterned but high cost of healthy food and inadequate knowledge/information were recognized as barriers of choosing healthy nutrition in Iran 
. Based on the above evidence and due to lack of valid data on SES in the present study, our findings might be confounded by socioeconomic status. Fourth, we could not separate high and low fat dairy products and hence did not differentiate the effects of high and low fat dairy products, whereas previous studies suggest that these may influence age of menarche differently 
. Finally, high consumption of milk and nutrient intakes may reflect overall better dietary quality and healthier life style in those with early menarche. Thus, the associations observed may be attributed to the other characteristics related to the timing of menarche. However, we cannot rule out such residual confounding.
In summary, findings of this community based study cohort provided evidence that milk intake in pre-puberty might hasten the timing of menarche. It seems that dairy products including milk, cheese and yoghurt may have different effects on puberty. Higher intakes of Ca, Mg, and P were also related to an earlier age of menarche. Further studies are needed to confirm these findings, and to clarify the possible effects of different types of milk and dairy products individually, regarding to their content of fat or technological processing. Investigating the relationship between dairy intakes and other pubertal events are also recommended in future studies.