In our study, 64 percent of melasma patients had a positive family history. Positive family history of melasma is reported in several studies [5
]. Although, genetic influence is suggested by a twin study but it is not generally accepted [10
Melanocytic nevi existed among 95% of patients in both groups of our study. Studying 432 healthy Caucasian subjects, MacKie found the mean body nevus count during first decade of life to be three for females and two for males, rising rapidly up to a mean of 33 for females and 22 for males in the third decade. Thereafter, numbers of moles slowly dropped until in the eighth decade; the count fell to levels similar to those seen in pre-pubertal children [12
]. In an Australian study, the mean number of melanocytic nevi was as high as 43 among men and 27 among women in second and third decades of [13
]. Mean number of melanocytic nevi was 13.2 in melasma cases compared to 2.8 in control group in our study. The mean in community can be something in between 2.8 percent and 13.2 percent which is different from other studies. The lower mean number of nevi in control group who lack melasma, seems to be mainly due to coincidence or association between melasma and melanocytic nevi, but the general difference of mean number of nevi compared to other studies can be due to other factors like different sun exposure level, different age distribution among study populations and different genetic or hormonal status among studies.
We found that having freckles, lentigines and having more than three melanocytic nevi were positively related to developing melasma in women increasing the chance of melasma up to 23 times in case of melanocytic nevi.
In spite of our vast literature review, no other study was found to focus on coincidence, association or predictive role of different types of nevi for melasma and only one study was found to report association between nevi and melasma. In a cross-sectional study carried out on 400 pregnant women in Tehran, a correlation was found between number of freckles and nevi on face with melasma [9
]. Although the study design was not a suitable one to draw the conclusions as the authors have done, but at least their findings are generally in line with ours.
Freckles are shown to have an autosomal transmission and coincidence with melanocytic nevi [4
]. It is shown also that there is no genetic influence in simple lentigo. Our main objective of study was to check for possible association between nevi (mainly melanocytic nevi) and melasma to be used for prediction purposes and not to declare a causal relation and we used a method of regression analysis. As we know the concept of regression doesn't generally imply any causal relation between regressor and regressand [15
]. Further research may be needed to clarify the validity and specificities of observed association. However if findings of this study are confirmed by large scale prospective studies, existence of some types of nevi specially the lentigines and melanocytic nevi can serve as predictors of melasma. This can be of help in selection of target groups for educational programs regarding melasma prevention.
Limitations, possible biases and strengths of the study
As like other case-control studies, selection bias is an inevitable part of study. We tried to decrease it by selecting controls on a basis that similar source population is assumed. An acceptable temporality detection is existing in this study due to natural process of nevi and melasma development. That is to say melsma develops later in life than nevi. Although this should be considered as limitation of this retrospective study, but findings of this research encourages researchers to design and conduct cohort studies that can give more reliable findings than case-control studies. Although a precise and much reliable prediction can not be expected for such a case-control study but the strong association and confidence interval function contrary to minimal systematic errors guarantees usefulness of its scientific findings.