This study suggested that glycemic load has a significant positive relationship with acne vulgaris occurrence after considering factors like BMI and gender. The risk of acne vulgaris decreased with the decreasing percentile of glycemic load, with a significant association noted at the 50th and 75th percentile. The association remained significant after adjustment for other confounding factors. Among the western population, high glycemic load diet was reported as a significant contributor to high acne vulgaris prevalence . The result of this study provides further support to a randomized controlled trial study among male acne patients aged between 15 – 25 years which demonstrated that low glycemic load diet was effective in improving acne vulgaris symptoms . Another study found improvements in biochemical parameters associated with acne vulgaris due to a high-protein, low glycemic-load diet intervention . A study also reported that a low glycemic load diet known as the South Beach Diet was effective to ameliorate the acne conditions of 86.7% of 2995 acne vulgaris respondents in a three months period and reduced the use of conventional systemic as well as topical treatments . In contrast to our study, these studies [12, 24, 25] found that the effect of the low glycemic load diet was lost when the data was statistically adjusted for changes in BMI.
The underlying mechanism of dietary effect on acne vulgaris formation might be the role of insulin-like growth factor-1 (IGF-1) in facilitating cell proliferation involved in acne vulgaris pathogenesis . Acute hyperinsulinemia due to consumption of high glycemic load diet would cause an increase in IGF-1/insulin-like growth factor binding protein-3 (IGFBP-3) ratio, thus enhancing the effects of IGF-1 . Hyperinsulinemia resulting from high glycemic load diet would also increase circulating androgens and decrease sex hormone binding protein, leading to increased sebum synthesis, which was crucial in acne development.
This study found that frequency of milk and ice cream intake was positively associated with acne vulgaris occurrence. The result is in accordance to a cross-sectional study in South Korea which reported that milk and dairy products intake was associated with acne vulgaris development . They reported that the intake of milk and dairy products was higher among acne vulgaris subjects compared to non-acne subjects . Our data also confirmed epidemiological studies performed in the United States [8–10]. Women, who consumed two or more servings of skimmed milk everyday, were 22% more likely to suffer from severe acne and 44% more likely to develop cystic or nodular acne than those who consumed only one glass of skimmed milk a day . Endocrine factors involved in acne may be affected by milk consumption because milk is an insulinotropic nutrient and has a high insulinemic index  which would increase serum insulin and IGF-1 levels [29–32]. Milk produced persistently by pregnant cows contains substantial amounts of steroids and androgen-precursors, which have been suggested to play another role in acne pathogenesis [33, 34]. Moreover, another group has proposed a hypothesis for the diet-induced impact of insulin/IGF-1 signaling in acne, as both high glycemic load and dairy proteins increase the serum levels of insulin and IGF-1, important promoters of sebaceous glands and sebaceous lipogenesis [35, 36]. In contrast to our study, a positive association between acne with cheese intake was found . The difference may be partly attributed to lower frequency of cheese intake among subjects in this study.
This study confirmed earlier findings that acne vulgaris patients were more likely to have family history of acne vulgaris compared to the control group. In Jordan , it was reported that 69.3% of acne vulgaris patients had family history of acne vulgaris while in the United Kingdom, a study involving 458 pairs of monozygotic and 1099 dizygotic female twins  found that genetic factors attributed to 81% of acne vulgaris variant and only 19% involved environmental factors. A higher prevalence of acne vulgaris was also found among students with parents having a history of acne vulgaris .
Our study has found that acne vulgaris was neither related to BMI nor body fat percentage. This provides further support to a study which discovered no correlation between BMI and leptin levels secreted by adipose tissue with the presence of acne vulgaris or its degree of severity . A retrospective cohort in Turkey  also found no significant difference in serum leptin levels between cases and controls. Other studies among female twins  and female acne patients [41, 42] also found no significant correlation between BMI and acne vulgaris severity among female acne vulgaris patients.
Our study found that yoghurt consumption was not correlated with acne vulgaris occurrence and is consistent with the findings by several studies [8–10]. When added to milk during fermentation process, probiotic bacteria (specifically Lactobacilli) utilize IGF-1 and lowered IGF-1 level in fermented milk by four-fold compared to skim milk . It was hypothesized that increased intestinal permeability occurred in acne vulgaris patients, thus enhancing IGF-1 absorption in the gut . Hence, milk consumption would cause an increased IGF-1 absorption rather than fermented milk products, which have lower levels of IGF-1. This might explain the association of milk with acne vulgaris occurrence in contrast to fermented dairy products such as yoghurt. However, it is proposed that the most important mechanism of milk signalling is the postprandial fast upregulation of insulin secretion and the long-lasting increase in serum IGF-1 levels [35, 36]. Recently, lactoferrin-enriched fermented milk was discovered to be effective in decreasing triglycerides in skin surface lipids, resulting in improvement of acne vulgaris symptoms through reduction in sebum production .
In contrast to common food beliefs, the present study found no statistically significant association between chocolate and nuts with acne vulgaris occurrences. Chocolate  and nuts  are commonly believed to cause or aggravate acne condition but previous studies regarding the effects of chocolate and nuts on acne vulgaris condition were inconsistent. An experiment on the effects of milk chocolate bars consumption in subjects with acne vulgaris found no exacerbation in their condition . Subsequently, it was reported that acne lesion counts as well as composition and synthesis of sebum were also not affected after consumption of chocolates bar containing 10 times higher than normal cocoa solids . Another case series study also concluded that chocolate and roasted nuts did not aggravate acne vulgaris condition . Contrary to our findings, a cross sectional study among Koreans involving 783 acne vulgaris patients and 502 controls showed that consumption of nuts among cases were significantly higher than acne-free subjects .Thus, there is a clear need to conduct a randomized clinical trial to have a better understanding of the association between these food items with acne occurrence.
A few limitations of the study should be mentioned. Firstly, due to the nature of the study design used, this study was only able to determine the association, but not the cause and effect of diet on acne vulgaris. The severity of acne vulgaris condition should also be taken into account through acne lesion counts on patients in order to relate specific types of food and the frequency of their consumption. Furthermore, a retrospective food frequency questionnaire would be the most appropriate tool to determine the association between dietary intake such as dairy products and acne development. Repeated 3 day food records would have been a better method to calculate glycemic load and determine its association with acne. However, a 3 day food record was used in this study to measure the glycemic load due to the limited study duration. Additionally, this study did not measure the amount of consumed dairy protein, which may be the most important determinant for the acne-promoting effects of milk. Other confounding factors such as stress, inadequate sleep, smoking, alcohol consumption and facial hygiene care should also been taken into account in future studies. Despite these limitations, the findings from our study have provided further support to the existing evidence for the role of diet in acne vulgaris occurrence.