Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Acne has long been thought to be associated with the consumption of certain foods. During the early 1900s, dermatology textbooks commonly noted that diets high in carbohydrates and sweets tended to make acne worse, with chocolate thought to be the most offending factor (Wise and Sulzberger 1933). Elimination diets were reported to have moderate success (Cormia 1940; White 1934). However, there was little agreement among dermatologists about what foods should be avoided. Allergenic skin tests were unable to identify the culprit foods (Cormia 1940; White 1934), suggesting that foods may aggravate acne through an indeterminate mechanism. Consequently, dermatologists had to rely on their own clinical experience, and there was no unified approach to deal with the condition. The diet and acne connection finally fell from favor in 1969 when a study found no difference in acne after the daily consumption of a chocolate versus a placebo bar containing equivalent amounts of fat and sugar (Fulton et al. 1969). Although this study was criticized at the time for several design flaws (Mackie and Mackie 1974), it brought a long awaited end to the trial-and-error dietary practices in the management of acne.

It is important to consider that at the time of the chocolate study the main nutritional challenge was the prevention of dietary deficiencies. Consequently, few could comprehend the notion of food causing disease in the absence of any real metabolic or nutritional deficiency. From 1958 to 1979, dietary advice was given in terms of minimum daily requirements, and the recommendation was to eat more than the dietary guidelines suggested to satisfy energy and nutrient needs (Welsh et al. 1993). The notion to “eat more” to prevent deficiency was in clear opposition to the idea of avoiding certain foods to prevent acne. The authors of the chocolate study summarized the consideration of the time that “if a food can really alter a disease … [this] finding would set into motion a wholesale attack on the effects of foods on normal physiologic functions.” However, in 1969 scientists were just beginning to become aware of the role of diet in the etiology of chronic disease states such as cardiovascular disease and diabetes. By the late 1970s, dietary advice had shifted from preventing dietary deficiencies to recommendations aimed at avoiding excessive consumption of food components (e.g., fat, saturated fat, cholesterol, sugar, salt, alcohol). Public awareness of the cause-and-effect relationship of the role of food in general health has since caused an explosion in scientific and mainstream literature as well as the development of new nutritional concepts such as the glycemic index. This mega-trend has fueled an explosion in nutritional research and a new understanding of how the food we eat relates to health or particular pathological processes.

Since the chocolate study, there has been a logarithmic progression in the science of nutrition that has generally gone unnoticed in the dermatology field. Over the past three decades, considerable focus has been given to the role of dietary carbohydrate in the rising tide of obesity and the cluster of metabolic disturbances (hyperinsulinemia, insulin resistance, hyperglycemia, hypertension, dyslipidemia) now termed the metabolic syndrome. Because the metabolic syndrome is considered to be a condition of disordered insulin metabolism, there is merit in evaluating foods based on their rate of absorption and their effect on blood glucose and insulin concentrations. Consequently, the glycemic index (GI) was introduced to quantify the blood glucose-raising potential for a given sample of food containing 50 g of available carbohydrate (Jenkins et al. 1981). By definition, high-GI carbohydrates are rapidly digested, producing rapid elevations in blood glucose and increasing insulin demand. In contrast, low-GI carbohydrates are slowly digested and absorbed, and they elicit a low insulin response during the postprandial period. It is thought that the abundance of low-GI foods and the absence of refined, high-GI carbohydrates in traditional cultures together may play a protective role against Western diseases (Colagiuri and Brand-Miller 2002). The evidence to date suggests that low-GI diets are associated with higher high-density lipoprotein (HDL)-cholesterol levels and with a lower risk of developing cardiovascular disease and diabetes (Liu 2002). Furthermore, dietary intervention trials suggest that low-GI diets may increase satiety and facilitate weight loss when compared to high-GI diets (Ludwig 2000). Therefore, the glycemic index has been a useful nutritional concept, providing new insights into the relation of food and Western diseases.

This chapter reconsiders the diet and acne connection and opens a new era for understanding the effects of nutrition on skin health. According to early scientific principles of medicine, to test the hypothesis that a dietary component is implicated in disease, one should demonstrate that (1) the diet of persons with disease is significantly different from those without the disease; (2) the signs and symptoms should be known to be or plausibly suspected of being caused by the dietary imbalance; and (3) correction of the dietary imbalance should result in alleviation of the signs and symptoms. Although these fundamental principles should be simple to apply, the major limitation in 1969 was poor understanding of the cause and exacerbation of acne. A better understanding of the disease at the biochemical level has helped elucidate a possible role of nutritionally related factors in acne etiology.

2 Acne Pathophysiology: Could Acne Be a Metabolic Disease?

Acne is considered as a disease of adolescence, affecting 80%–90% of individuals aged 12–15 years (Lucky et al. 1991). Clinical observation indicates that this condition can also affect prepubescent children and adults. Acne often begins during adrenarche (8–12 years), and incidence rates increase with pubertal maturation. During adolescence, the disease is more common and severe in boys, possibly reflecting an effect of androgens on sebum production (Stathakis et al. 1997). Acne incidence declines after 18 years of age, but a considerable number of men and women aged 20–40 years continue to be affected. A review of the data over recent years suggests an increasing prevalence of acne in people over the age of 25, particularly women (Goulden et al. 1997). It is unknown why acne tends to be chronic for a subset of adult women, but severity is reported to be influenced by factors that reflect hormonal fluctuations, including menstrual cycles, pregnancy, and menopause (Shaw and White 2001; Thiboutot and Lookingbill 1995).

Examination of the life course of acne may provide a physiological framework on which we can examine the role of diet-related factors in acne development. Acne generally begins when androgens of either adrenal or gonadal origin increase, stimulating sebum production. The most important androgen is testosterone, which may be locally converted to the more active dihydrotestosterone by 5α-reductase. However, acne severity and incidence does not correlate well with testosterone levels, suggesting that the hormonal control of acne is complex and may involve interplay of other factors. Acne has shown to correlate better with the proportion of testosterone to sex hormone binding globulin (SHBG), an indicator of testosterone bioavailability. Other biological factors, such as insulin and insulin-like growth factor (IGF)-I, can also stimulate sebum production and growth of keratinocytes in cell cultures (Deplewski and Rosenfield 2000; Eming et al. 1996). Clinically, elevated insulin levels have been described in women with persistent adult acne (Aizawa and Niimura 1996), and significantly higher IGF-I levels have been described in women with acne compared with controls (Deplewski and Rosenfield 1999; Aizawa and Niimura 1995; Cappel et al. 2005). These physiological traits may influence one or more of the pathogenic processes involved in acne development, including (1) increased sebum production, (2) hyperproliferation and differentiation of follicular keratinocytes, (3) microbial colonization by Propionbacterium acnes, and (4) inflammation.

Among the factors associated with the clinical presentation of acne, insulin may play a key role in activating a hormonal milieu conducive for acne development (Fig. 10.1). Insulin has been shown to augment the growth-promoting effects of IGF-I by decreasing levels of its binding protein, IGF-binding protein 1 (IGFBP-1) (Powell et al. 1991). Both insulin and IGF-I can stimulate adrenal androgen synthesis and gonadal testosterone production through effects on steroidogenic enzymes and gonadotrophin-releasing hormone secretion (Willis et al. 1996). Insulin and IGF-I can also inhibit hepatic SHBG production (Singh et al. 1990), increasing the concentration of androgens that are bioavailable. Therefore, in addition to the direct effects of insulin on pathogenic processes (i.e., sebum production, keratinocyte growth), insulin has the potential to indirectly affect acne through shifts in endocrine systems.

Fig. 10.1
figure 1_10

Mechanism of a high glycemic load diet promoting acne in insulin-resistant states. SHBG sex hormone binding globulin, IGFBP-1 insulin-like growth factor binding protein-1, IGF-I insulin-like growth factor-I

Full size image

Support for a role of insulin in acne development can be found in the pattern of acne prevalence. Acne clinically presents during puberty, which is a transient period of normal insulin resistance (Bloch et al. 1987; Moran et al. 1999). Even before puberty, hyperinsulinemia is found to be predictive of acne incidence in later years (Miller et al. 1996; Ibáñez et al. 1997). Acne incidence follows the rise and fall of pubertal insulin resistance more closely than the change in androgen levels, as androgen concentrations remain elevated following acne regression during the late teenage years. Fluctuations in the incidence of acne throughout the normal life cycle appears to coincide with changes in insulin sensitivity, with insulin resistance observed also during pregnancy, menses, and menopause (Homko et al. 2001; Pulido and Salazar 1999; Godsland et al. 1995). Perhaps the strongest evidence for an association of acne and insulin resistance comes from the fact that acne is a common feature of women with polycystic ovary syndrome (PCOS), a condition characterized by hyperinsulinemia and hyperandrogenism. Clinical observation suggests that insulin resistance is the underlying feature in PCOS, as it generally precedes and gives rise to hyperandrogenism (Dunaif et al. 1989). Furthermore, reducing insulin secretion and/or increasing insulin sensitivity with pharmacological interventions (i.e., acarbose and metformin, respectively) decreases the severity of acne symptoms in individuals with PCOS (Ciotta et al. 2001; Kazerooni and Dehghan-Kooshkghazi 2003; Kolodziejczyk et al. 2000).

If insulin resistance and hyperinsulinemia are suspected to play a role in acne development, it suggests that obese hyperinsulinemic adults should also exhibit signs of acne. The absence of acne in older, obese adults may be because, during puberty, both insulin resistance and acne are intrinsically linked to the growth hormone/IGF-I axis (Deplewski and Rosenfield 1999; Moran et al. 2002a). Acne typically presents prior to puberty when growth hormone (GH) and IGF-I levels begin to rise; and usually it resolves by the third decade as GH and IGF-I levels decline. In obese hyperinsulinemic adults, however, there is a negative association between adiposity and GH levels (Pijl et al. 2001; Luque and Kineman 2006). Obese individuals not only have a low basal GH output, they exhibit blunted responses to all recognized GH stimuli, including fasting, acute exercise, GH-releasing hormone, and insulin tolerance tests (Luque and Kineman 2006; Williams et al. 1984; Qu et al. 2004). Some individuals continue to exhibit high IGF-I levels beyond puberty, and elevated IGF-I levels have been found to be a precipitating factor in persistent adult acne (Aizawa and Niimura 1995). This suggests that the status of the GH/IGF-I axis may be an important underlying factor in acne pathogenesis.

3 Secular Trends of Advancing Pubertal Age and Acne: Evidence for a Role of Diet?

The reasons for the change in insulin sensitivity at the various hormonal stages of life (i.e., pregnancy, puberty, menses, menopause) are unknown. It has been suggested that insulin resistance of puberty may relate more to changes in growth hormone release than changes in body mass (Moran et al. 2002b). Interestingly, in Western countries, the mean age of menarche has fallen from 16 to 13 years since the beginning of the century, which suggests an earlier onset of pubertal insulin resistance and hyperinsulinemia. Although the insulin resistance of puberty is thought to be relatively benign, exposure to hyperinsulinemia at a younger age is predicative of the later development of states of androgen excess: PCOS, acne, hirsutism (Miller et al. 1996). Western nutrition is generally assumed to be responsible for the secular trend to an ever-earlier onset of puberty. Although the nutritional factors responsible remain to be identified, possible metabolic cues include degree of fatness (“critical fatness”), glucose availability, and IGF-I and leptin levels (Foster and Nagatani 1999).

Epidemiological observations have provided us with some compelling evidence for a role of Western dietary factors in acne development and may provide us with some insight into the role of nutrition in advancing pubertal development. A recent observational report described the low incidence of acne in non-Westernized societies, where the mean age of menarche was 16 years (Cordain et al. 2002). The authors implicated diet—mainly the absence of high-GI foods—for the low rates of acne in these societies. This is in support of earlier observations during the 1970s of the emergence of acne in Eskimos groups following the introduction of Western foods (Schaefer 1971; Bendiner 1974). The higher rates of acne in Eskimo groups paralleled the increase in annual per-capita consumption of refined sugar and flour, and the per-capita consumption of protein from animal sources showed an inverse relation. Even though the Eskimo’s traditional hunter-gatherer diet was very low in carbohydrate, there was a relatively high intake of carbohydrate following the introduction of agriculture by Russian settlers some 70–100 years ago (Schaefer 1970). However, these carbohydrates (e.g., barley, buckwheat, cabbage, potatoes) had a relatively low glycemic index (Table 10.1) and did not replace animal protein as the main source of energy. Only since the relatively recent exposure to refined high-GI carbohydrates have the Eskimos demonstrated faster growth (increased final height), earlier puberty, and dramatic increases in the incidence of obesity, diabetes, and heart disease.

Table 10.1 Glycemic and insulinemic responses to foods commonly consumed in traditional (acne-free) cultures and Westernized societies
Full size table

The traditional diets of acne-free populations characteristically have a low glycemic load and therefore produce only modest postprandial rises in plasma glucose and insulin. As the glycemic index can only be used to compare foods of equal carbohydrate content, the glycemic load was later developed to characterize the glycemic effect of whole meals and diets on the basis of the rate of glucose appearance and the quantity of carbohydrate consumed (glycemic index  ×  carbohydrate content). At present, there are relatively few data available on the classification of foods according to their insulin response, although the correlation between glycemic and insulinemic responses is reported to be high (r  =  0.74 and 0.90, respectively) (Holt et al. 1997; Bornet et al. 1987). Table 10.1 illustrates that food staples of traditional cultures elicit lower glycemic and insulinemic responses than Western dietary staples. When applying the glycemic load concept to whole diets, the glycemic load may be reduced by decreasing total carbohydrate intake or by selecting foods using the glycemic index concept. Traditionally, the Eskimo diet would have been low in glycemic load due to low intakes of carbohydrate and the consumption of low-GI foods. Acne became a problem in these societies only when the adolescents began to consume high-GI carbohydrates (i.e., sweet biscuits, potato crisps, soft drinks, confectionary) in large quantities (Schaefer 1971; Bendiner 1974). Consequently, reducing the dietary glycemic load may represent a unique dietary strategy to alleviate acne via a reduction in hyperinsulinemia and its hormonal sequelae.

4 Clinical Evidence of a Therapeutic Effect of Low Glycemic Load Diets in Acne Vulgaris

A recent randomized controlled trial found that a low glycemic load (GL) diet that mimics the diets of acne-free populations may alleviate acne symptoms and hormonal markers of acne. In a 12-week study, 43 young male subjects (age 15–25 years) with mild to moderate acne consumed either a conventional high-GL diet or a low-GL diet and had their acne assessed every 4 weeks (Smith et al. 2007a, b). The experimental low-GL diet was achieved through a reduction in carbohydrate intake and through selection of low-GI foods. After 12 weeks, study participants on the lowGL diet demonstrated a 20% greater reduction in acne lesion counts than the participants on the high-GL diet. The lessening of acne severity can be explained by improvements in metabolic-endocrine parameters. When compared to controls, participants on the low-GL diet demonstrated significant improvements in insulin sensitivity and hormonal markers of acne (increases in SHBG and IGFBP-1, suggesting decreased bioavailability of testosterone and IGF-I) (Smith et al. 2007b). These changes may also relate to the modest weight loss (2.5 kg) that occurred with the reduction in dietary glycemic load. Post hoc analyses revealed that the effect of the low-GL diet on acne and certain endocrine parameters was lost after statistically adjusting for weight loss. This suggests that weight loss mediated the reduction in insulin resistance and its associated hyperinsulinemia, which may be important in the clinical regression of acne.

At present, it remains uncertain whether diet can alleviate acne without weight loss. In theory, hyperinsulinemia can be reduced through improvements in the metabolic state of insulin resistance and/or reducing postprandial hyperglycemia. Interestingly, another study found no difference in the global assessment of acne in weight-maintained individuals following the consumption of diets that had a high and a low glycemic index (Reynolds et al. 2010, personal communication). Although this suggests that weight loss may be responsible for the clinical regression in the earlier study, it should be noted that differences in the nature of the dietary interventions (reduced dietary glycemic index versus reduced dietary glycemic load) and the assessment of acne (acne lesion counts versus global assessment of acne) may also account for the different study outcomes.

Few studies have reported an association of body weight and the incidence of acne. The U.S. National Health Survey of 1966–1970 found that dietary excess (as reported by the parent) was significantly associated with a greater degree of acne prevalence, particularly in young boys (US Department of Health Education and Welfare 1976). A survey of 2,720 soldiers demonstrated that adult men with acne were significantly heavier (5.6 kg) than adult men without acne. However, this association was found to be dependent on age, as weight was not associated with age in adolescents aged 15–19 years. Interestingly, a similar age-dependent association between acne and obesity was found in the low-GL study. This study found that acne lesion counts correlated with body mass index (BMI) in men ≥18 years of age, but this association was not true for boys <18 years of age (Fig. 10.2). The reason for the age-dependent association is unknown, but it is possible that acne during adolescence may relate more to physiological changes in insulin sensitivity (via a weight-independent mechanism), whereas adult acne may be more pathological in origin (obesity-related insulin resistance).

Fig. 10.2
figure 2_10

Age-dependent association of acne severity and body mass index (BMI). Bivariate analysis was performed with a two-tailed Pearson’s correlation (n  =  43)

Full size image

These clinical observations may provide a foundation for future dietary recommendations in the management of acne. There is already scope for diet as a treatment option, as acne patients often treat themselves with over-the-counter therapies. Diet therapy may be used alone or in conjunction with conventional acne therapies in cases of mild to moderate acne. However, as severe nodulocystic acne can be painful, disfiguring, and leave permanent scars, it is recommended that patients seek optimal treatment from a specialist physician. When considering diets for adolescents, there are two eating behaviors that also require special consideration: eating that leads to obesity and the disturbed psychiatric conditions of anorexia and bulimia. Like acne, obesity can have significant negative psychological consequences, including low self-esteem, social inhibition, depression, and anxiety. Low-GL diets may present a useful strategy for individuals seeking to lose weight and at the same time prevent or lessen the severity of acne. However, diet should not be considered as a strategy for psychologically vulnerable individuals with a preexisting eating disorder, as it may increase the food-associated anxiety and the preoccupation with food.

There is now also compelling evidence from clinical and epidemiological studies to suggest a potential role of diet in acne development. In 1969, the authors of the chocolate study stated that “it would be remarkable if skin functions were easily influenced by the vagaries of the diverse diets which have evolved in human populations.” During the four-decade scientific vacuum since the chocolate study, much has been learned from the diets of non-Westernized societies. Dietary intervention trials suggest that low-GL diets can alleviate acne symptoms, possibly through improving insulin metabolism and decreasing the bioavailability of testosterone and IGF-I. These endocrine changes may influence the desquamation of follicular keratinocytes and sebum production, two primary factors involved in the development of an acne lesion. It remains to be objectively determined whether weight loss is the principal factor in the clinical alleviation of acne.