Modifiable prognostic factors in uterine fibroid development: a systematic review of literature

Abstract

Background

Many risk factors in uterine fibroid development have been identified, but women and their physicians are less aware of the influence of lifestyle on uterine fibroid development. The objective of this systematic review is to investigate and summarize modifiable prognostic factors associated with uterine fibroid development.

Methods

Pubmed and Embase were searched for relevant articles according to PRISMA guidelines. References from included articles were screened and when relevant also included. Human in vivo studies on modifiable factors in fibroid development were included. Studies on non-modifiable factors and treatment, in vitro studies and animal studies were excluded. 607 articles were screened and 33 articles were included. Two independent investigators collected data from the report.

Results

The strongest risk factor for fibroid development was a high BMI, while the strongest protective factors were a high fruit and vegetable intake and high vitamin D intake.

Conclusion

More high-quality studies are necessary to better understand the impact of the abovementioned factors as well as the role they play in the growth of already existing fibroids.

PLAIN LANGUAGE SUMMARY

Uterine fibroid development is multifactorial. Various non-modifiable and modifiable factors have been linked to uterine fibroid development. Modifiable factors are controllable by patients themselves. We performed a systematic review to investigate these modifiable factors. We screened 607 articles from 2 databases (PubMed and Embase) of which 33 were included in the review. We only included clinical studies on humans. A high body mass index (BMI) is found to be a modifiable risk factor for uterine fibroid development. Protective of uterine fibroid development are high fruit and vegetable intake and high vitamin D intake or sun exposure. More research is needed to investigate the applicability of these findings in clinical practice and to investigate the influence on the growth of already existing uterine fibroids.

Keywords:

• Uterine fibroid
• modifiable
• prognostic factor
• risk factor

Introduction

Uterine fibroids are a common gynaecological disorder, occurring in about 70% of women (Stewart et al. 2017). Fibroids result in substantial burdens, causing heavy menstrual bleeding, bulk symptoms and fertility problems, reducing quality of life and resulting in high healthcare costs (Al-Hendy et al. 2017). Fibroid development can be categorised into fibroid incidence and fibroid growth. Several non-modifiable risk factors of fibroid development have been described, such as age, age of menarche and ethnicity. The chance of developing fibroids increases with age and a positive family history (Stewart et al. 2017, Sharami et al. 2019). Additionally, earlier menarche is associated with increased risk of fibroid development, while a history of pregnancy and lactation appear to offer a protective effect against fibroid development (Peddada et al. 2008, Bidgoli et al. 2012, Velez Edwards et al. 2013). Moreover, research has demonstrated that black women have a higher prevalence of multiple fibroids and a higher fibroid volume (Bray et al. 2018).

Fibroids grow under the influence of gonadal hormones oestrogen and progesterone (Stewart et al. 2017). The median growth rate (for Black and white women combined) is 9% per 6 months (range −89% to 138%) (Peddada et al. 2008). Growth rate of fibroids decreases after the age of 35 years for white women, but not for Black women (Peddada et al. 2008). A higher growth rate is found in smaller fibroids, submucous fibroids and intramural fundal fibroids (Armbrust et al. 2018, Baird et al. 2020). Interestingly, spontaneous shrinkage also occurs in about 7–11% of fibroids (Peddada et al. 2008, Baird et al. 2020). Even fibroids within the same woman can show different growth rate patterns (Peddada et al. 2008). While short growth spurts of individual fibroids are not related to fibroid size, location, age, ethnicity, BMI or parity (Peddada et al. 2008, Baird et al. 2011), growth may be related to vascularity of fibroids (Nieuwenhuis et al. 2018). Additionally, women with a co-existing vitamin D deficiency tend to have larger fibroids (Baird et al. 2013).

No medical treatments have been described that prevent the development of fibroids. However, treatment of hypertension with an ACE inhibitor and treatment of hypercholesterolaemia with a statin seem to decrease the risk of fibroid development (Borahay et al. 2016, Fischer et al. 2021). Targeted medical therapy on fibroid growth has been subject to many studies. These medical therapies can (temporarily) reduce fibroid size. But after cessation of therapy, fibroid regrowth occurs in all premenopausal patients.

While uterine fibroids may cause severe and chronic symptoms, with a high burden on healthcare costs, little is known about fibroid development prevention. Existing research has established that age, onset of menarche, a positive family history and ethnicity are non-modifiable and strong prognostic risk factors of fibroid development. This prompts the question of whether factors exist that are modifiable and reduce the risk of fibroid development. The purpose of this systematic review is to identify and assess modifiable prognostic factors associated with the development of uterine fibroids. This overview will provide insight for healthcare practitioners and patients to mitigate risks of fibroid development and reduce fibroid burden.

Methods

A systematic PubMed and Embase search was performed according to PRISMA Guidelines (Page et al. 2021). For search strings see Appendix A (Supplementary). Reference lists from relevant articles were screened for additional articles. Data collection took place until the 15 June 2022. Ethical approval was not applicable, because this study will collect and analyse data from already published studies. For results see Figure 1 (Flow Chart).

Figure 1. Flow Chart.

We aimed to answer the following research question: What are modifiable prognostic factors for uterine fibroid development?

Modifiable factors were defined as factors that can be influenced by individual patients. These factors include directly modifiable factors such as supplement intake, diet, smoking as well as indirectly modifiable factors such as pregnancy or (to a certain extent) health conditions that can be treated by medication, like hypertension, hypercholesterolaemia and diabetes. In this review we focus only on the directly modifiable factors. Fibroid development was defined as the incidence of new fibroids. Factors influencing fibroid growth were not included in the scope of this review.

Human in vivo studies were included while animal and in vitro studies were excluded. Studies on non-modifiable factors like ethnicity, reproductive factors, surgical treatment, embolisation, (MR-)HIFU and reviews were excluded. Studies assessing targeted (medicinal) therapies such as hormones, phyto-oestrogens, traditional (plant) medicine for the purpose of control of blood loss were excluded. Use of contraceptives for contraceptive purposes were included because they were not targeted as a treatment for fibroids. Reviews, editorials and case studies as well as publications for which a full text article could not be retrieved were also excluded.

For PICO see Table 1. Data was extracted for each study on study design, study setting, country the study took place in, study date, number of participants, information on participants and the control group, the influencing factor being assessed, and the outcome that was analysed. If outcomes were given in odds ratio (OR), relative risks (RR), incidence rate ratios (IRR), hazard ratio (HR), the data was extracted with the 95% confidence intervals (CI) and summarised in Table 2. Two investigators independently extracted the abovementioned data from the study report (A.L.K., A.S), processing the data into a table before interpreting the results. When univariate and multivariate analyses were reported, multivariate results were included in the review. To provide an impression of the degree to which the studies can be used in clinical practice as a guidance for clinical advice, a level of evidence was provided. Studies were rated according to the Levels of Evidence developed by the Oxford Centre for Evidence-Based Medicine (OCEBM) (Howick et al. 2011). Baseline characteristics of the included studies are provided in Table 3 (Supplementary).

Table 1. PICO.

Patients	Women with uterine fibroids
Intervention	Personally modifiable prognostic factors like diet and lifestyle
Control	Women without uterine fibroids
Outcome	Uterine fibroid development

Table 2. Modifiable factors that influence fibroid occurrence.

Factor	Author (year)	Effect^†	N	Effect size		95% CI	Clarification	Level of Evidence‡
Contraceptives
OCP use	Chiaffarino et al. (1999a,1999b)		2.400					4
Current OCP use		–		aOR	0.3	0.2 − 0.6	Current users vs. never users
Long term OCP use		–		aOR	0.5	0.3 − 0.9	≥7 years of use vs never users
Long term OCP use	Bidgoli et al. (2012)	–	276	OR	2.69	1.30 − 5.55	use ≥5 years more frequent in control group	4
Long term OCP use	Wise (2004a, 2004b)	*	22.895	IRR	1.1	0.9 − 1.3	≥10 years vs. no use	2
DMPA use	Wise (2004a, 2004b)	–	22.895	IRR	0.6	0.4 − 0.9		2
History of DMPA use	Harmon and Baird, (2015)	–	722	aRR	0.8	0.6 − 0.9		4
DMPA use in the previous 2 years	Harmon et al. (2022)	–	1.610	aHR	0.6	0.4 − 1.0		2
Metabolic factors
BMI	Terry et al. (2007)		116.609				Compared to women with a BMI of 20	2
BMI 20 − 21.9		*		HR	0.83	0.67 − 1.01
BMI 22 − 23.9		*		HR	1.07	0.91–1.26
BMI 24 − 24.9		+		HR	1.34	1.09 − 1.65
BMI 25 − 26.9		+		HR	1.35	1.12 – 1.62
BMI 27 − 29.9		+		HR	1.77	1.48 − 2.12
BMI 30		+		HR	1.98	1.69 − 2.33
BMI	Ciebiera et al. (2016)	+	188	OR	1.097	1.01–1.19	For one unit increase of BMI	4
BMI	Wise et al. (2005)		21.506				Compared to women with a BMI of < 20	2
BMI 20.0 − 22.4		+		IRR	1.34	1.02 − 1.75
BMI 22.5 − 24.9		+		IRR	1.39	1.07 − 1.81
BMI 25.0–27.4		+		IRR	1.45	1.12 − 1.89
BMI 27.5–29.9		+		IRR	1.47	1.11 − 1.93
BMI 30.0 − 32.4		+		IRR	1.36	1.02 − 1.80
BMI 32.5		*		IRR	1.21	0.93 − 1.58
Current body weight	Lee et al. (2018)						In Korean women, compared to women weighing 50 – 60 kg	4
women weighing 60–70 kg		+	5.062	OR	1.44	1.05 − 1.99
women weighing 70 kg or more		+	5.062	OR	1.18	0.67 − 2.05
Weight gain since the age of 18	Terry et al. (2007)		116.609				Compared to stable weight	2
Lost ≥5 kg		*		aRR	1.10	0.96–1.26
Weight gain 3–4.9 kg		+		aRR	1.15	1.05–1.26
Weight gain 5–9.9 kg		+		aRR	1.14	1.05–1.23
Weight gain 10–14.9 kg		+		aRR	1.25	1.15–1.35
Weight gain 15–19.9 kg		+		aRR	1.25	1.14–1.37
Weight gain ≥20 kg		+		aRR	1.31	1.22–1.42
Weight gain since the age of 18	Lee et al. (2018)		5.062				In Korean women, even in women with normal body weight	4
weight gain 5 – 10 kg		+		aOR	1.84	1.22 − 2.76
weight gain > 10 kg		+		aOR	2.00	1.25 − 3.20
Physical activity	Baird et al. (2007)		1.189				Comparing lowest category of < 2 hours exercise per week to the highest category of 7 hours exercise per week	4
without chores		–		aOR	0.61	0.41 − 0.90
with chores		–		aOR	0.64	0.43 − 0.95
Nutrition and diet
Fruit and vegetable intake	Wise et al. (2011)	–	22.583	IRR	0.90	0.82 − 0.98	4 compared with 1 serving per day	2
Fruit		–		IRR	0.89	0.81 − 0.98	2 servings per day compared with 2 -servings per week
Vegetables		*		IRR	0.97	0.89 − 1.05	2 servings per day compared with 4 servings per week
Citrus fruit		–		IRR	0.92	0.86 − 1.00	3 servings per week compared with 1 serving per month
Dietary vitamin A	Wise et al. (2011)	–	22.583	IRR	0.89	0.83 − 0.97	Upper quintile (≥ 1051 mcg/day of dietary vitamin A intake) compared with lower quintiles	2
Dietary vitamin A	Wise et al. (2021)	*	1.230	aHR	0.99	0.71–1.40	Upper quintile (≥955 mcg/day of dietary vitamin A intake) compared with lower quintile	2
Serum vitamin A levels	Martin (2011)		887					4
Middle tertile vs low tertile		+		aOR	2.43	1.35 − 4.37
High tertile vs low tertile		+		aOR	2.66	1.16 − 6.10
Serum vitamin C levels	Martin (2011)		887					4
Middle tertile vs low tertile		*		aOR	1.17	0.55 − 2.48
High tertile vs low tertile		*		aOR	1.62	0.80 − 3.28
Beef and red meat	Chiaffarino et al. (1999a,1999b)	+	2.400	aOR	1.7	1.4–2.2	High consumption vs low consumption	4
Caffeine	Wise et al. (2004a, 2004b)		21.885					2
All ages		*		aOR	0.93	0.75–1.15	≥3 cups a day vs almost never
≤35 years		+		aOR	1.53	1.06–2.22
Caffeine	Chiaffarino et al. (1999a,1999b)	*	2.400	aOR	0.8	0.6–1.0	High intake vs low intake	4
Soy intake	Shen et al. (2013)	+	1.200	aOR	2.52	1.90–3.35	Chinese population	3
Soy intake	Wise et al. (2010)		22.120					2
Total soy foods		*		IRR	0.95	0.73–1.24	≥7 servings/week vs <1 serving/week U.S population
Soy milk		*		IRR	1.00	0.73–1.37
Tofu		*		IRR	1.09	0.78–1.53
Soy/veggie burger		*		IRR	0.81	0.54–1.23
Soy intake	Nagata et al. (2009)	*	285	aOR	1.82	0.79–4.17	High vs low soya isoflavones	4
Vitamin D
Serum vitamin D levels	Ciebiera et al. (2016)	–	188	OR	0.96	0.93 − 0.99	For a 1 unit increase of serum vitamin D concentration	4
Serum vitamin D levels	Baird et al. (2013)	–	1036	aOR	0.68	0.48 − 0.96	Sufficient vitamin D levels [>20 ng/ml] compared to insufficient levels	4
Self-reported sun exposure	Baird et al. (2013)	–	1036	aOR	0.6	0.4 − 0.9	reported sun exposure ≥1 hour/day	4
Self-reported sun exposure	Kumari et al. (2022)	–	310	OR	9.08	4.53 − 18.21	Sun exposure of <1 hour a day compared to high sun-exposure of ≥1 hour/day	4
Toxins
Alcohol consumption	Wise et al (2004a, 2004b)		21.885					2
Alcohol consumption 14+ drinks/week		*		IRR	1.23	0.921.66	14+ drinks vs. none
Beer consumption of <1 drink/week		+		IRR	1.11	0.98 − 1.27	Compared to non-drinkers
Beer consumption of 1–6 drinks/week		+		IRR	1.18	1.00 − 1.40
Beer consumption of 7+ drinks/week		+		IRR	1.57	1.17 − 2.11
Alcohol consumption	Nagata et al. (2009)		852					4
Overall alcohol consumption		+		aOR	2.78	1.25–6.20	Highest tertile vs lowest tertile, Japanese cohort
Beer consumption		+		aOR	2.75	1.30–5.83
Alcohol consumption	Chiaffarino et al. (1999a,1999b)	*	2.400	aOR	1.0	0.1–1.3	High vs low alcohol consumption. 90% of participants drank wine	4
Smoking	Wise et al. (2004a, 2004b)		21.885				Compared to never smoking	2
Former smoking		*		IRR	0.95	0.84 − 1.08
Current smoking		*		IRR	0.88	0.77 − 1.01
Current smoking	Parazzini et al. (1996)	–	1759	RR	0.5	0.4 − 0.7	Compared to never smoking	4
Smoking	Oskovi Kaplan (2018)	–	124	OR	0.35	0.16 − 0.76		4
Exposure to smoking	Bidgoli et al. (2012)	+	276	OR	2.09	1.29– 3.37		4

^aEffect on fibroid occurrence (positive +, negative - or not significant *).

^bLevel of evidence were rated according to the Oxford Centre for Evidence-Based Medicine (OCEBM) (Howick et al. 2011).

Results

With the initial search described in Appendix A (Supplementary) and following reference screening, 607 articles were collected. After title and abstract screening, 134 publications remained. 33 eligible publications were included in the systematic review after full-text screening.

Table 2 provides a summary of the extracted data on individual prognostic factors associated with fibroid development in terms of OR, RR, IRR, or HR. Baseline characteristics of the publications used for the systematic review can be found in Table 3 (Supplementary). There were a wide variety of study designs amongst the selected studies. Of the 33 publications included, there were 11 case-control studies, 7 cross-sectional studies, 12 prospective cohort studies (4 of which with a study population (n ≤ 102)), 2 retrospective cohort studies and 1 small pilot study. The populations analysed also varied, with studies examining different nationalities (e.g., Italian, Chinese, American), women with varying racial backgrounds, as well as different cohort settings (e.g., women screened in a healthcare setting vs. survey respondents). Moreover, diverse diagnostic methods were used amongst the different studies including imaging- or histologically confirmed fibroids, as well as self-reported fibroid diagnoses.

Contraceptives

Oralcontraceptives

In several studies, use of oral contraceptives (OCP) for contraceptive purposes appear to have a protective effect against fibroids. A strong protective effect was found amongst Italian and Iranian women who were currently using OCPs, as compared to women who had never used them. Moreover, a further reduction of fibroid development was found with longer OCP use (Chiaffarino et al. 1999a, 1999b, Bidgoli et al. 2012). Although ever use of OCP’s amongst Italian women compared to never use showed no significant effect (Parazzini et al. 2004). The Black Women’s Health Study did not find a decreased risk of fibroid development for current or ever OCP use (Wise et al. 2004a, 2004b).

Depot medroxyprogesterone acetate

A history of depot medroxyprogesterone acetate (DMPA) use is associated with a decreased risk of fibroid development of up to 40%, with also a stronger effect for longer use (Wise et al. 2004a, 2004b, Harmon and Baird 2015, Harmon et al. 2022).

Levonorgestrel intra uterine device

Although several studies investigated the effect of the levonorgestrel intra uterine device (LNG-IUD) on fibroid volume, no studies report an effect of LNG-IUD on fibroid development (Magalhaes et al. 2007, Shawki et al. 2009, Tasci et al. 2009, Socolov et al. 2011, Xie et al. 2012).

Metabolic factors

Body mass index

Several studies show that Body Mass Index (BMI) is associated with fibroid development. Amongst women who had their fibroids surgically removed, the presence of elevated BMI and waist to hip ratio (WHR) were found to be associated with a higher incidence of multiple- and larger fibroids (Pan et al. 2021). A current higher BMI is positively correlated with an increased risk of fibroid development, but not for all ethnicities (Wise et al. 2005, Terry et al. 2007, Dandolu et al. 2010, Yang et al. 2014, Ciebiera et al. 2016, Sharami et al. 2019). For Korean women current BMI was not associated with fibroid development, but weight gain since the age of 18 was, even among women with a normal BMI (Lee et al. 2018). Increased weight gain since the age of 18 is also associated with an increased risk of fibroid development in other studies (Wise et al. 2005, Terry et al. 2007). However one study found a decreased risk of fibroids for women with severe overweight (Parazzini et al. 2004).

Physical activity

One registry study found exercise to be associated with a decreased risk of fibroid development for women in their reproductive age, in a dose-dependent manner, even when adjusting for BMI (Baird et al. 2007).

Diet, nutrition and supplements

Fruit,vegetables and vitamins

Various studies suggest that diet plays a role in fibroid development. Fruit and vegetable intake were found to be inversely associated with the presence of fibroids in Black women, with a stronger association for fruit than vegetables(Wise et al. 2011). An inverse association was also found for dietary vitamin A intake, driven mostly by vitamin A from animal sources, not provitamin A derived from fruit and vegetables (Wise et al. 2011). A subsequent study found, however, no such relation with any type of vitamin A (Wise et al. 2021). Dietary vitamin C, E, folate, fibre and carotenoids were not associated with fibroids (Wise et al. 2011, 2021). In contrast, serum vitamin A levels seem to be positively associated with fibroids in a dose-response relationship (Martin et al. 2011). This was also found for vitamin C but results were not statistically significant (Martin et al. 2011).

Vitamin D

Several studies confirm that vitamin D may play a protective role in fibroid development. Vitamin D is inversely associated with fibroid development (Baird et al. 2013, Ciebiera et al. 2016, Oskovi Kaplan et al. 2018). This was also found for self-reported sun exposure (Baird et al. 2013, Kumari et al. 2022). Hypovitaminosis D is found more often in black women than in white women (Baird et al. 2013). A retrospective cohort study among white women suggests that a one unit increase in serum vitamin D can lower the risk of fibroid development by 4% (Ciebiera et al. 2016).

Meat

Meat consumption seems to be associated with an increased risk of fibroid development. An Italian study suggests that a large amount of beef and red meat consumption might increase the risk of fibroid development (Chiaffarino et al. 1999a, 1999b). No direct effect on uterine fibroid risk was found for eating barbecued and grilled foods. However, a study among Iranian women found elevated levels of the aryl hydrocarbon receptor (AhR) in women who eat grilled meat routinely (Bidgoli et al. 2012). The AhR mediates the effects of polycyclic aromatic hydrocarbons (PAH) which are found in barbecued and grilled foods. The AhR is strongly overexpressed in uterine fibroids (Bidgoli et al. 2012).

Caffeine

Heavy coffee and caffeine consumption were overall not associated with a risk of fibroid development (Chiaffarino et al. 1999a, Wise et al. 2004a, 2004b). Only amongst young women under the age of 35 years drinking 3 or more cups of coffee daily, was associated with an increased risk of fibroid development (Wise et al. 2004a, 2004b).

Soy products

Soy intake does not seem to have a clear effect on fibroid development. In two large Chinese studies, results indicate an increased incidence of uterine fibroids amongst women consuming soybean milk and milk or soybean milk (Shen et al. 2013, Gao and Wang 2018). The Black Women’s Health study in the United States, however, found no such association, also when soy milk, tofu or soy/veggie burgers were analysed separately (Wise et al. 2010). Also in a Japanese study amongst a population with generally high soy intake, these results could not be confirmed (Nagata et al. 2009).

Food additives

Few studies have assessed the effect of food additives on fibroid development. One Chinese study suggests an increased risk for uterine fibroid development with the consumption of food additives, sweeteners and preservatives (Shen et al. 2013).

Toxins

Alchohol consumption

Alcohol consumption maybe associated with an increased risk of fibroid development. In the Black Women’s Health study current alcohol consumption had a small positive association with uterine fibroids, however, particularly drinking beer was found to be associated with increased risk of fibroid development (Wise et al. 2004a, 2004b). A Japanese study found similar results for overall alcohol consumption. This effect was driven strongly by beer consumption (Nagata et al. 2009). On the contrary, a cross-sectional Italian study found no association between overall alcohol intake and the presence of fibroids. However, 90% of participants drank wine instead of other alcoholic beverages (Chiaffarino et al. 1999a, 1999b).

Smoking

Smoking seems to be associated with a decreased risk of fibroid development. In the Black Women’s Health study there was a non-significant inverse association between smoking and fibroids. No trend was found for pack-years of smoking, number of cigarettes smoked per day, years of smoking or age first smoked (Wise et al. 2004a, 2004b). For Italian women, current smoking compared to never smoking was found to decrease the risk of fibroids, but no trend was found for number of cigarettes smoked or duration of smoking (Parazzini et al. 1996, 2004). In a study amongst Turkish women, smoking was also found to decrease the risk of having fibroids (Oskovi Kaplan et al. 2018). In contrast, a study amongst Iranian women found an increased risk of fibroids for exposure to smoking (Bidgoli et al. 2012).

Discussion

Main findings

Several modifiable risk and protective factors have been identified that may play a role in fibroid development.

The studies found show that adequate vitamin D and healthy nutrition with high fruit and vegetable intake may be protective factors against uterine fibroid development.

Hypovitaminosis D or reduced sun-exposure correlate with increased fibroid occurrence. These results are in line with in vitro studies where vitamin D inhibits myometrial and leiomyoma cell growth, functioning as an antagonist of sex steroid hormone receptors, reducing proliferation and extracellular matrix formation and increasing apoptosis (Blauer et al. 2009, Sharan et al. 2011, Al-Hendy et al. 2015, Corachán et al. 2021). A reduced amount of serum vitamin D may therefore play an important role in fibroid pathogenesis. While hypovitaminosis D is prevalent among minority populations, it is possible that vitamin D deficiency could, in part, contribute to the racial disparity seen with fibroid development (Zadshir et al. 2005). Whether prophylactic vitamin D supplementation could reduce the risk of fibroid development in a high-risk population has not been assessed.

Several studies suggest a protective role of contraceptives, particularly DMPA use, reporting reduced risks of uterine fibroid development with current or previous DMPA use. The mechanism of action is unknown, but are likely due to the suppressed oestradiol and progesterone concentrations with DMPA use, where oestradiol concentrations even resemble those of postmenopausal women (Clark et al. 2001). DMPA use is associated with decreased bone mineral density over time, but bone mineral density recovers after cessation of DMPA (Curtis and Martins 2006). Data concerning OCP use are contradictory, with several case-control studies reporting a reduced risk and one prospective cohort study reporting no association.

A strong risk factor identified for fibroid development is a higher BMI. Numerous studies link weight with fibroid development. A possible explanation for increased fibroid risk with obesity, is the increased availability of active oestrogen as a result of adipose tissue mediated conversion of androgens to oestrogens, and more importantly the reduced concentration of sex hormone binding globulin (Wise et al. 2005). There is an indication that physical exercise reduces the risk of fibroid development even when adjusting for BMI, but more research is needed to assess the temporal relationship between exercise and fibroid development and whether exercise may diminish existing fibroid burden.

Different studies also found alcohol, particularly beer consumption, to be associated with increased risks of fibroid development. A potential explanation for these observed patterns is that chronic alcohol consumption in premenopausal women has been shown to increase plasma and urinary oestradiol hormone concentrations (Reichman et al. 1993). While soy contains isoflavones which are phytoestrogens with chemically similar structures as oestradiol, research assessing increased soy intake found no effect on fibroid development. Increased coffee consumption, which has previously been shown to increase early follicular phase oestradiol levels has contradictory results, with increased fibroid risks found only in women under the age of 35 (Lucero et al. 2001). Smoking has antiestrogenic effects but similarly has conflicting results for fibroid development (Wise et al. 2004a, 2004b).

Strengths and limitations

The present review provides an overview of the current literature assessing modifiable prognostic factors influencing fibroid development. We distinguished between fibroid incidence and fibroid growth and focus in this review on fibroid incidence. Literature was systematically searched and data systematically collected. Trends and patterns were identified, helping researchers recognise possible modifiable factors to further assess.

There are numerous studies reporting on modifiable prognostic factors influencing uterine fibroid development of which some include fibroid growth. However, most studies are small with plausibly many (non-modifiable) confounders, like ethnicity and age which can influence study results and generalisability. As mentioned in the introduction, different fibroids, even within the same woman, can display different growth patterns and even spontaneous shrinkage. In addition, the majority of the studies had a low level of evidence, with most of them having a cross-sectional or retrospective design with small sample sizes and/or wide confidence intervals and thus a high level of uncertainty in the estimate. Few studies assessed the temporal effect of the influencing factors with the use of a prospective study design. Therefore, interpretation of these studies should be with some caution. Furthermore, while uterine fibroids are believed to be a common condition affecting women in the reproductive age, only 25% of these women are symptomatic (Stewart et al. 2016). While the majority of women with fibroids are asymptomatic, the frequency of uterine fibroids within a population is likely largely underestimated, especially in self-reporting studies, resulting in a bias towards clinically more severe cases (Payson et al. 2006). Due to the heterogeneity of the studies a meta-analysis was not deemed feasible, except for vitamin D, however this has recently been done (Mohammadi et al. 2020).

Future directions

The present review’s scope entails the modifiable prognostic factors of fibroid development. Less research, however, is done on modifiable prognostic factors influencing fibroid growth. The largest amount of evidence is found for the protective role of vitamin D in fibroid development. But other dietary measures like increased fruit and vegetable intake and lower BMI are likely protective factors in fibroid development. More research is necessary in well-conducted prospective study designs, to assess these correlations and temporal relations. Future research should assess the abovementioned lifestyle intervention including physical exercise, amongst women with a high risk of uterine fibroid development or newly diagnosed fibroids. Further research is also necessary to explore the impact of these lifestyle interventions on existing fibroids and their growth, as this could empower women to take more control over the progression of their condition.

Conclusion

Even though much research has been done on risk factors that influence uterine fibroid development, only few show a strong association. Of the directly modifiable factors, a healthy lifestyle with high fruit and vegetable intake and vitamin D exposure are the strongest. Since a higher BMI is associated with a higher chance of fibroid development, a general advice on healthy diet and normal weight is a safe suggestion to make to patients, also in the context of other lifestyle associated diseases. There are some indications that exercise may be preventive even if corrected for BMI, but future studies are needed to confirm this. Large well-conducted prospective studies are needed to better understand the association between modifiable factors and fibroid development, shed light on fibroid aetiology, and therefore facilitate fibroid prevention strategies.

Authors’ contribution

A.L. Keizer contributed to conception and design, acquisition, analysis and interpretation of data, drafting the work, approving the final version and accepting accountability for all aspects of the work.

A. Semmler contributed to conception and design, acquisition, analysis and interpretation of data, drafting the work, approving the final version and accepting accountability for all aspects of the work.

H.S. Kok contributed to drafting and revising the work, approving the final version and accepting accountability for all aspects of the work.

P.J.M. van Kesteren contributed to drafting and revising the work, approving the final version and accepting accountability for all aspects of the work.

J.A.F. Huirne contributed to conception and design, drafting and revising the work, approving the final version and accepting accountability for all aspects of the work.

W.J.K. Hehenkamp contributed to conception and design, drafting and revising the work, approving the final version and accepting accountability for all aspects of the work.

Acknowledgements

None.

Data availability statement

Data availability is not applicable to this article as no new data were created or analysed in this study.

Additional information

Funding

The authors have no relevant financial or non-financial interests to disclose.