Association between Family History and Male Androgenetic Alopecia with Female Pattern Hair Loss

Abstract

Background

Male androgenetic alopecia (MAGA) is often accompanied by female pattern hair loss (FPHL). However, the risk factors related to MAGA with FPHL are unclear.

Objective

To investigate demographic and laboratory factors related to MAGA with FPHL.

Methods

This retrospective case-control study was performed in a single tertiary care center for MAGA with FPHL between March 2012 and September 2021. Eligible patients were males >12 years old diagnosed with androgenetic alopecia by a dermatologist. The patients were subdivided into MAGA with FPHL and MAGA without FPHL groups. Comorbidities as well as demographic, laboratory, and disease-specific variables were compared between the two groups. Data analysis was conducted between October 2021 and February 2022. The independent samples t-test, Mann–Whitney U test, and chi-squared test were used to assess the factors that contributed to MAGA with FPHL.

Results

Of 469 patients with MAGA, 309 (65.9%) had FPHL, which was a much higher rate than previously reported. Among the variables, only matrilineal (odds ratio, 1.605; 95% confidence interval, 1.014~2.541) and maternal history (odds ratio, 4.705; confidence interval, 1.632~13.559) of androgenetic alopecia were significantly associated with MAGA with FPHL. In the MAGA with FPHL group, a significant positive correlation was noted between body mass index and the type F score (r=0.114, p=0.025).

Conclusion

In this case-control study, patients with MAGA and a maternal history of androgenetic alopecia were at risk of FPHL. Therefore, early screening may benefit these patients.

INTRODUCTION

Androgenetic alopecia (AGA) is the most common hair loss disorder; there are racial differences in its presentation, with a prevalence of approximately 15% to 20% in Asians¹,²,³,⁴,⁵. Androgens play a major role in AGA pathogenesis⁶ and the miniaturization of hair follicles is a hallmark feature of the condition⁷,⁸. AGA is also called pattern hair loss because it shows a characteristic hair loss pattern⁵. Various classification modalities such as Hamilton, Norwood–Hamilton, and Olsen classification have been proposed to distinguish pattern hair loss¹,⁹,¹⁰. However, Ludwig, Olsen, and Sinclair classification methods are used for AGA in females¹. For AGA in females, the hair density in the mid-frontal area decreases; however, for AGA in males, the hair density in the frontal hairline is maintained¹¹. These classification methods had a limitation in that they could not classify cases of female pattern hair loss (FPHL) in males or of male pattern hair loss (MPHL) in females¹²,¹³,¹⁴,¹⁵. Therefore, the basic and specific (BASP) classification method was proposed to address this limitation¹⁶,¹⁷. In this classification method, male AGA (MAGA) is generally characterized as type V (specific type), along with the basic types and FPHL is classified as types F, T, and O (specific types), with each type divided into three stages¹⁶,¹⁸.

FPHL in males was first described in 1993¹⁹; the prevalence of FPHL in males has been reported to be higher in Koreans than in Caucasians³. Based on the BASP classification, MAGA with FPHL can be defined as MAGA with BASP type F and is considered a distinct clinical variant. In a 2014 epidemiological study, the proportion of Korean males with type F was 24.2%, which was a non-negligible proportion²⁰. Recently, Kerkemeyer et al.²¹ reported a high rate of hypotestosteronemia and hypovitaminosis D in patients with FPHL during a retrospective review of 84 males with FPHL; however, the factors that contribute to MAGA with FPHL are not yet clear. The main objective of this study was to determine the factors related to MAGA with FPHL. We hypothesized that a maternal family history of AGA, differences in male sex hormone levels including serum testosterone, and factors that could affect peripheral microcirculation were related to MAGA with FPHL.

MATERIALS AND METHODS

Patients and candidate factors

In this retrospective study, we reviewed males diagnosed with AGA at Wonju Severance Christian Hospital between March 2012 and September 2021. We defined MAGA with FPHL as MAGA with a frontal type (type F) pattern of hair loss in the BASP classification. Patients were subdivided into MAGA without FPHL and MAGA with FPHL. We excluded patients for whom FPHL could not be determined. The presence or absence of FPHL was confirmed based on the BASP type recorded in the medical records; if the BASP type was not recorded, two dermatologists would review the patient's clinical picture and determine the type. Each medical record was reviewed for demographic data, comorbidities, and disease-specific factors, including age, body mass index (BMI), smoking history, alcohol use, hypertension, diabetes mellitus, dyslipidemia, and family history of AGA. In the case of family history, we assessed the patients to identify those eligible for a family history of AGA; we divided family history into total family history, patrilineal history (father and/or grandfather), matrilineal history (mother and/or grandmother), and maternal history (mother only). Additionally, laboratory variables including total cholesterol, triglyceride, free thyroxine (T4), testosterone, dehydroepiandrosterone sulfate (DHEA-S), and prostate-specific antigen (in those over 45 years of age) were compared between the two groups. This study was approved by the institutional review board of Yonsei University Wonju College of Medicine (approval no: CR321158).

Statistical analysis

Continuous data such as age, BMI, smoking history, and laboratory variables were presented as means and standard deviations. Categorical data including alcohol use, comorbidities, and familial history of AGA were reported as percentages. The independent samples t-test and Mann–Whitney U test were used to compare continuous data and the chi-squared test was used to compare categorical data. Analyses were performed using SPSS version 23.0 (IBM Corp.) and GraphPad Prism software version 9.1.1 (GraphPad). p-values of <0.05 were considered statistically significant.

RESULTS

A total of 648 patients with MAGA visited our department during the study period and 469 were included in this study, after excluding patients whose FPHL status could not be determined. Table 1 summarizes the baseline characteristics of the 469 patients included in this retrospective study; 309 of the 469 patients (65.9%) had F-type hair loss. There were no significant differences in demographic variables, comorbidities, and laboratory variables between the FPHL and non-FPHL groups. Table 2 summarizes the BASP classification of the 469 patients.

Table 1. Baseline characteristics of the patients (n=469).

Characteristic		MAGA with FPHL (n=309)	MAGA without FPHL (n=160)	p-value
Demographic variable
	Age (yr)	36.1±13.0	33.9±12.2	0.102
	BMI (kg/m²)	24.6±3.2	23.9±2.8	0.053
	Smoking Hx (pack-years)	4.1±8.5	4.0±8.8	0.777
	Alcohol use	156 (50.5)	84 (52.5)	0.721
Comorbidity
	Hypertension	26 (8.4)	10 (6.3)	0.648
	Diabetes mellitus	16 (5.2)	8 (5.0)	>0.99
	Dyslipidemia	28 (9.1)	8 (5.0)	0.217
Disease-specific factor
	Familial Hx of AGA	209 (67.6)	109 (68.1)	0.810
	Patrilineal Hx of AGA	155 (50.2)	92 (57.5)	0.093
	Matrilineal Hx of AGA	93 (30.1)	34 (21.3)	0.044*
	Maternal Hx of AGA	33 (10.7)	4 (2.5)	0.002*
Laboratory variable
	Total cholesterol (mg/dl)	178.99±33.08	178.51±36.53	0.893
	Triglyceride (mg/dl)	153.00±100.05	146.60±126.37	0.142
	Free T4 (ng/dl)	1.33±0.24	1.36±0.23	0.275
	Testosterone (ng/ml)	4.77±1.60	4.98±1.60	0.219
	DHEA-S (μg/dl)	226.60±108.45	239.21±104.00	0.275
	PSA (ng/ml)†	0.82±0.72	0.84±0.43	0.251

Values are presented as mean±standard deviation or number (%). AGA: androgenetic alopecia, BMI: body mass index, DHEA-S: dehydroepiandrosterone sulfate, FPHL: female pattern hair loss, Hx: history, MAGA: male androgenetic alopecia, PSA: prostate-specific antigen. ^*Significant values (p<0.05). ^†PSA was measured only in patients over 45 years of age (n=82).

Table 2. Summary of the BASP classification of the 469 patients.

Type			MAGA with FPHL (n=309)	MAGA without FPHL (n=160)	Total (n=469)
Basic type
	L		84 (27.2)	25 (15.6)	109 (23.2)
	M
		M0	3 (1.0)	2 (1.3)	5 (1.1)
		M1	123 (39.8)	69 (43.1)	192 (40.9)
		M2	76 (24.6)	44 (27.5)	120 (25.6)
		M3	14 (4.5)	6 (3.8)	20 (4.3)
	C
		C0	2 (0.6)	0 (0.0)	2 (0.4)
		C1	4 (1.3)	2 (1.3)	6 (1.3)
		C2	3 (1.0)	1 (0.6)	4 (0.9)
		C3	0 (0.0)	1 (0.6)	1 (0.2)
	U
		U1	0 (0.0)	8 (5.0)	8 (1.7)
		U2	0 (0.0)	2 (1.3)	2 (0.4)
		U3	0 (0.0)	0 (0.0)	0 (0.0)
Specific type
	V
		V1	151 (48.9)	63 (39.4)	214 (45.6)
		V2	52 (16.8)	17 (10.6)	69 (14.7)
		V3	17 (5.5)	12 (7.5)	29 (6.2)
	F
		F1	161 (52.1)	0 (0.0)	161 (34.3)
		F2	125 (40.5)	0 (0.0)	125 (26.7)
		F3	23 (7.4)	0 (0.0)	23 (4.9)

Values are presented as number (%). BASP: basic and specific, MAGA: male androgenetic alopecia, FPHL: female pattern hair loss.

Regarding disease-specific variables, although there were very few patients with maternal history in both groups that could affect these results, there was a significant difference in matrilineal history and maternal history between the two groups, and the rate of FPHL was higher if there was a maternal history of AGA. Patients with MAGA who had a patrilineal history of AGA had a decreased risk of FPHL (odds ratio [OR], 0.690; 95% confidence interval [CI], 0.455~1.049). In contrast, patients with MAGA who had a matrilineal history of AGA had an increased risk of FPHL (OR, 1.605; 95% CI, 1.014~2.541). In the presence of a maternal history of AGA, the OR significantly increased to 4.705 (95% CI, 1.632~13.559) (Fig. 1).

Fig. 1. Factors related to male androgenetic alopecia (MAGA) with female pattern hair loss (FPHL). The forest plot shows the OR of FPHL in patients with MAGA versus in the controls. AGA: androgenetic alopecia, OR: odds ratio, CI: confidence interval.

Correlation analysis was performed to determine factors that were correlated with the type F score of the BASP classification method in the FPHL group. The type F score showed a positive correlation with age and BMI; a significant negative correlation with free T4, testosterone, and DHEA-S. In partial correlation analysis for age, only BMI retained its positive correlation with the type F score (r=0.114, p=0.025) (Table 3).

Table 3. Correlation between type F scores and variables.

		Type F scores
		r	p-value	r†	p-value†
Demographic variables
	Age	0.256*	<0.001*	0.114*	0.025*
	BMI	0.136*	0.007*
	Smoking Hx	0.044	0.393
Laboratory variables
	Total cholesterol	0.014	0.768
	Triglyceride	0.060	0.222
	Free T4	–0.122	0.042*	–0.090	0.138
	Testosterone	–0.117	0.017*	–0.082	0.094
	DHEA-S	–0.186	<0.001*	–0.075	0.126
	PSA	–0.113	0.311

BMI: body mass index, Hx: history, DHEA-S: dehydroepiandrosterone sulfate, PSA: prostate-specific antigen. ^*Significant values. ^†Ageadjusted correlation coefficient and p-value.

DISCUSSION

Traditionally, there is a clinical difference between MPHL and FPHL. A previous study reported that females had fewer androgen receptors, 5α-reductase types 1 and 2, and a higher amount of cytochrome P-450 aromatase than males⁸. Such androgen-independent features are considered the reason for the appearance of FPHL clinical features. However, it is common for males to have FPHL, which is a variant of MAGA and the factors related to MAGA with FPHL have not been identified. MAGA with FPHL is known to have a higher prevalence in Asians than in Caucasians. In 2014, Yeo et al.²⁰ reported that 24.2% of Korean patients with MAGA had F-type hair loss. According to recently reported Australian data, the rate of MAGA with FPHL was very low at 3.9%²¹. However, our data showed a considerably higher rate (65.9%) compared to previous results. Although if there was no record of BASP classification, the authenticity of our diagnosis of MAGA with FPHL would seem questionable because of the much higher rate compared to previous reports. However, with the BASP classification, the result was reasonable because it was based on clinical photographs being determined by two dermatologists at the time of the initial consultation. Another explanation for higher prevalence is that patients without confirmed FPHL, who may not have FPHL, were not included in our study. A representative patient is presented in Fig. 2.

Fig. 2. A 30-year-old male with LF2 type (basic and specific classification) of androgenetic alopecia. While the anterior hairline has a normal appearance, a marked decrease in hair density can be observed in the frontal scalp.

To our knowledge, this study is the first study to investigate demographic and laboratory factors related to MAGA with FPHL. Furthermore, we subdivided the family history of AGA into total, patrilineal, and matrilineal. We confirmed that the risk of FPHL in MAGA patients significantly increased if there was a matrilineal history of AGA, especially a maternal history. To date, several susceptibility genes/loci, including the androgen receptor/ectodysplasin A2 receptor on chromosome X²² and histone deacetylase 9 on chromosomes 7p21²³, 2q35²⁴, 2q37.3 have been associated with AGA²⁵. However, genes/loci located on autosomes showed no association with FPHL²⁶. A study in the Chinese population confirmed that 22 autosomal MAGA-associated tag single-nucleotide polymorphisms were unrelated to FPHL²⁷. Considering these results, the observed impact of a maternal history of AGA on MAGA with FPHL suggests that the androgen receptor gene located on the X chromosome, inherited from the mother, plays a major role in the expression of FPHL.

However, in the case of cardiovascular disease risk factors such as hypertension, diabetes mellitus, dyslipidemia, BMI, and smoking history, which can affect peripheral microcirculation, there were no significant differences between MAGA with and without FPHL. Therefore, the microcirculation of the scalp may have an effect, but the effect is not considerable, and in all, the roles of age and genetic factors, which are the most important risk factors of AGA can be re-emphasized.

In the correlation analysis conducted within the MAGA with FPHL group, type F scores showed a significant positive correlation with age and BMI, which is a reasonable finding considering that AGA progresses with aging. Although the correlation was not significant with adjustment for age, as noted in recent reports²¹, an increase in the type F score was associated with lower testosterone levels. Testosterone levels were found to be significantly lower in males with F-type pattern hair loss²⁸, which has been linked to MAGA²⁹. We hypothesize that aromatase, prostaglandin D₂ (PGD2) synthase, which is induced by testosterone³⁰, and PGD2 are likely to be involved in FPHL, owing to the lower testosterone levels in Asian males. These result in higher aromatase levels, which play a detoxifying role against androgens in females³¹ and lower PGD2 levels, which inhibit hair growth³²,³³. Furthermore, if testosterone levels are high, the hair follicles in the frontal hairline become more sensitive to dihydrotestosterone, and if they are low, the sensitivity of the frontal area increases relatively. Genetic predispositions, such as androgen receptor polymorphism on the X chromosome, as well as differences in testosterone levels, may explain the variations in the prevalence of MAGA with FPHL observed between regions and ethnicities. Further research is needed to confirm this phenomenon.

In our study, BMI showed a significant positive correlation even with adjustment for age. Considering that the prevalence of FPHL was high in females with metabolic syndrome in a previous study of Koreans, risk factors of metabolic syndrome such as BMI appear to affect the presence of FPHL in males³⁴.

In this study, MAGA patients were classified using the unique BASP classification method that can classify both MPHL and FPHL¹⁶. In addition, this study, which aimed to reconfirm the validity of the BASP classification by showing that males with a maternal family history of AGA have a higher risk of accompanying FPHL, has implications in that it is a follow-up concept study of the 2011 study¹⁷. Further, because MAGA patients with FPHL are thought to have a poorer quality of life³⁵, for early screening of FPHL, a more detailed family history of AGA should be obtained even in males without FPHL.

This study has some limitations. First, the study considered only one ethnicity and was a single-institution retrospective study. However, sufficient sample size was obtained, and the results derived from the single ethnicity suggest there may be differences in polymorphism between races. Second, although there was a risk of patient recall bias, the family history of AGA obtained from the patients was considered reliable because the average age of the patients was not high (mid-30s). Third, there may have been issues with interrater reliability, but the BASP classification is an objective and intuitive method, and since the evaluation was performed by two trained dermatologists, their results can be considered reliable¹⁶. In all, studies seem necessary to confirm the quality of life and compliance of MAGA patients with FPHL, and larger studies are needed to precisely determine the prevalence of MAGA with FPHL.

In conclusion, our study shows that patients with MAGA and a maternal history of AGA are more likely to have FPHL. Therefore, early recognition and intervention would be necessary for these patients.

DATA SHARING STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.