(PDF) Dietary Patterns and Human Reproduction

Dietary Patterns and Human Reproduction

Voedingspatronen en de Menselijke Voortplanting

Marijana Vujković

The research presented in this dissertation was performed at the department of Obstetrics and

Gynaecology / division of Reproductive Medicine, and the department of Bioinformatics, Erasmus

University Medical Centre, Rotterdam, the Netherlands.

The printing of this thesis is financially supported by the department of Obstetrics and Gynaecology,

the department of Bioinformatics, the Molecular Medicine Post-Graduate School, Erasmus University

Medical Centre, the J.E. Jurriaanse Foundation, the Netherlands Heart Foundation, the Holy Trinity

Serbian Orthodox Church in Rotterdam, and the Erasmus University Rotterdam. Further financial

support was kindly provided by Crosslinks B.V. and Ordina N.V.

ISBN/EAN 978-90-9025675-7

Cover photo Maksim Shmeljov

Layout Marijana Vujković

Print Ipskamp Drukkers B.V., Enschede, the Netherlands

Rotterdam, the Netherlands

marijana.vujkovic@gmail.com

ously. Published manuscripts have been reproduced with explicit permission from the publishers. No

part of this thesis may be reproduced or transmitted in any form, by any means, electronic, mechanic,

photocopy, recording or otherwise, without the prior written permission of the author, or where

appropriate, of the publisher of the articles and figures.

Dietary Patterns and Human Reproduction

Voedingspatronen en de Menselijke Voortplanting

Proefschrift

ter verkrijging van de graad van doctor aan de

Erasmus Universiteit Rotterdam

op gezag van de

rector magnificus

prof.dr. H.G. Schmidt

en volgens het besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

woensdag 20 oktober 2010 om 15:30 uur

door

Marijana Vujković

geboren te Rotterdam

Promotiecommissie

Promotoren

prof.dr. R.P.M. Steegers-Theunissen

prof.dr. E.A.P. Steegers

prof.dr. P.J. van der Spek

Overige leden

prof.dr. C.H. Bangma

prof.dr. J.P. Mackenbach

prof.dr.ir. E.J.M. Feskens

Paranimfen

Drs. F. Hammiche

Drs. D. Vujković

Het verschijnen van dit proefschrift werd mede mogelijk gemaakt door de steun van de Nederlandse

Hartstichting, en de J.E. Jurriaanse Stichting.

Contents

8 CHAPTER 1

General introduction

PART I

Dietary patterns and reproductive performance

16 CHAPTER 2

Associations between dietary patterns and semen quality in men undergoing

IVF/ICSI treatment

32 CHAPTER 3

The preconceptional Mediterranean dietary pattern in couples undergoing IVF/ICSI

treatment increases the chance of pregnancy

46 CHAPTER 4

Increased preconceptional omega-3 polyunsaturated fatty acid intake improves

embryo morphology

PART II

Dietary patterns and pregnancy outcome

64 CHAPTER 5

The maternal Mediterranean dietary pattern is associated with a reduced risk of

spina bifida in the offspring

78 CHAPTER 6

A maternal dietary pattern characterized by fish and seafood is associated with a

reduced risk of congenital heart defects in the offspring

94 CHAPTER 7

Maternal Western dietary patterns and the risk of developing a cleft lip with or

without a cleft palate

108 CHAPTER 8

General discussion

123 APPENDICES

Summary | Samenvatting | List of Abbreviations | Authors and Affiliations | Publica-

tions and Awards | About the Author | PhD Portfolio | Word of Thanks

General

Introduction

‘I wondered how something so small could grow to such immense dimensions.’

Nikola Tesla (1856 – 1943)

CHAPTER

1

8 | Chapter 1

RATIONALE

An individual human life begins as a single cell at conception when an ovum and a sperm unite in the

female uterine tube.

1

As the conceptus starts travelling to the uterus it will begin to rearrange, grow,

and divide. This process is continuously repeated for 45 cell divisions until ultimately a multicellular

body of approximately 30-100 trillion cells is reached in adulthood.

2

Thirty generations of cell divisions

have already taken place when the embryo reaches its 8

th

week of existence, and by this time most

internal organs and external body structures are formed.

Till the mid of the 20

th

century the placenta was believed to effectively protect the developing embryo

in utero from all environmental influences. We know now that the developing embryo is not only

vulnerable to environmental hazards, such as viruses and radiation, but to a harmful maternal modern

lifestyle as well, such as the use of tobacco, alcohol, drugs, and unhealthy diet. Evidence is accumulat-

ing that nutritional depletions during pregnancy can permanently alter essential fetal developmental

processes in order to survive.

3-5

These fetal adaptations can result in an irreversibly altered structure

and function of some vital organs, leading ultimately to prematurity, birth defects, behavioural and

learning disabilities in childhood, and illnesses later in life.

6-10

In the three months prior to conception, sperm cells are forming in the testes, while oocytes are

maturing in the ovaries. Despite the limited knowledge about these processes, environmental agents

have shown to influence the quality of the maturation of gametes and the chances of conception.

11-13

Because nutrition is subject

to imposed environmental or, in humans, deliberate change, effects

on

reproductive health are therefore also likely.

6,14

So far, most studies have been focusing on the effects

of single nutrients with regard to health outcome. Scientific discoveries so far have allocated critical

roles for nutrient deficiencies with respect to poor reproductive outcome. The 7 best documented

nutrients are: iron, iodine, vitamin A, zinc, folate, vitamin D and calcium.

15

Nutrition is growingly investigated with regard to disease risk in terms of dietary patterns rather than

nutrients, for which three reasons are mainly responsible.

16-19

First, in everyday life people do not eat

nutrients but meals, which consist of a variety of foods. Many patterns are observed among food

intake ascribable to individual taste, neurophysiological impulses, attitude, economic resources, and

cultural habits. Second, dietary patterns are able to include complex health effects resulting from

interactions between nutrients. And third, the cumulative effect on health outcome of many foods as

part of the overall diet is often greater than an isolated nutrient effect. Dietary patterns approximate

real-world eating behaviour, and have the major advantage to provide valuable information for the

design and realization of food-based dietary guidelines.

THESIS MOTIVATION, OBJECTIVES AND ANTICIPATED RESULTS

The communication of healthy food intake via food-based dietary guidelines has shown to be more

straightforward and effective than traditional single nutrient advice.

20,21

Single nutrient analysis has

been very successful in past in targeting specific health outcomes, well-illustrated by the discovery of

General Introduction I 9

folic acid’s beneficial effects in preventing offspring with neural tube defects.

22,23

However, recent

public health programs are aiming at improving overall dietary quality in order to prevent poor

reproductive performance and adverse pregnancy outcome, and for this approach a single nutrient

focus is insufficient. Therefore, the nutrition-disease relationship needs to be understood at the level of

foods as well as dietary patterns.

The studies in the current thesis aim to delineate the impact of dietary patterns on pregnancy- and

other reproductive outcomes, and the main objectives can be summarized as follows:

1. To identify patterns in food intake from food frequency questionnaire data by using mod-

ern empirical statistical methods, such as principal components factor analysis and re-

duced rank regression.

2. To examine the associations between dietary patterns and several biochemical markers,

e.g. folate, vitamin B12, vitamin B6, homocysteine, S-adenosylmethionine, and S-adenosyl-

homocysteine.

3. To assess the relationship between preconceptional dietary patterns and reproductive

performance in men and women, e.g. semen quality, fertility parameters, and chance of

pregnancy.

4. To study the effects of periconceptional maternal dietary patterns on the risk of several

congenital malformations in the offspring, e.g. spina bifida, congenital heart defects, and

orofacial clefts.

With the findings described in this thesis we aim to provide a better understanding of specific dietary

patterns that are associated with reproductive performance and chances of pregnancy in subfertile

couples. Furthermore, new insights will be provided on to the degree in which maternal dietary

patterns contribute to poor pregnancy outcomes, such as congenital malformations.

OUTLINE OF THE THESIS

Part 1 of the thesis focuses on dietary patterns and fatty acid intake in couples undergoing IVF/ICSI

fertility treatment. The studies described in Chapter 2, 3 and 4 are based on the FOod Lifestyle and

Fertility Outcome study (FOLFO), a prospective cohort study examining the influence of preconcep-

tion lifestyle exposures in subfertile couples on fertility parameters and pregnancy outcome. This study

was conducted between 2004 and 2007 in the Erasmus University Medical Centre, Rotterdam, the

Netherlands.

24

Part 2 focuses on the role of the maternal dietary patterns periconception as a risk factor for congeni-

tal malformations in the offspring. The studies on risk factors for spina bifida (Chapter 5) and orofacial

cleft (Chapter 7) offspring were conducted in a nationwide large-scale case-control triad study carried

out between 1998 and 2004 at the Radboud University Nijmegen Medical Centre.

25, 26

The findings on

associations with congenital heart malformations in Chapter 6 are based on the HAVEN study (Hart

10 | Chapter 1

Afwijkingen, Vasculaire status, Erfelijkheid en Nutriënten).

7

The HAVEN study is an ongoing case-control

triad study conducted in the Western part of the Netherlands. The study emphasizes nutrition, lifestyle

and genes in the pathogenesis and prevention of congenital heart malformations and children and

both parents were included. Data collected between March 2004 and August 2008 were used for

analysis.

Finally, Chapter 8 provides a general discussion of the thesis, a summary of the key findings, reflec-

tions on the strengths and limitations of the applied methods, implication for clinical practice and

public health, and suggestions for future research.

REFERENCES

1 Larsen WJ. Human Embryology 2nd Ed. New York: Churchill Livingstone; 1997.

2 Guyton AC, Hall, J.E. Textbook of medical physiology. 10th ed. Philadelphia: W.B. Saunders;

2000.

3 Ashworth CJ, Antipatis C. Micronutrient programming of development throughout gestation.

Reproduction 2001;122:527-35.

4 Hanson MA, Gluckman PD. Developmental origins of health and disease: new insights. Basic

Clin Pharmacol Toxicol 2008;102:90-3.

5 Jackson AA. Nutrients, growth, and the development of programmed metabolic function.

Adv Exper Med Biol 2000;478:41-55.

6 Steegers-Theunissen BP. Maternal nutrition and obstetric outcome. Baillieres Clin Obstet

Gynaecol 1995;9:431-43.

7 Verkleij-Hagoort AC, de Vries JH, Ursem NT, de Jonge R, Hop WC, Steegers-Theunissen RP.

Dietary intake of B-vitamins in mothers born a child with a congenital heart defect. Eur J Nutr

2006;45:478-86.

8 Timmermans S, Jaddoe VW, Hofman A, Steegers-Theunissen RP, Steegers EA. Periconception

folic acid supplementation, fetal growth and the risks of low birth weight and preterm birth:

the Generation R Study. Br J Nutr 2009;102:777-85.

9 Case A, Paxson C. Children's health and social mobility. Future Child 2006;16:151-73.

10 Barker DJ, Clark PM. Fetal undernutrition and disease in later life. Rev Reprod 1997;2:105-12.

11 Olsen J, Rachootin P, Schiodt AV. Alcohol use, conception time, and birth weight. J Epidemiol

Community Health 1983;37:63-5.

12 Olsen J, Rachootin P, Schiodt AV, Damsbo N. Tobacco use, alcohol consumption and infertil-

ity. Int J Epidemiol 1983;12:179-84.

13 Homan GF, Davies M, Norman R. The impact of lifestyle factors on reproductive performance

in the general population and those undergoing infertility treatment: a review. Hum Reprod

Update 2007;13:209-23.

14 Godfrey KM, Barker DJ. Fetal nutrition and adult disease. Am J Clin Nutr 2000;71:1344S-52S.

15 Cetin I, Berti C, Calabrese S. Role of micronutrients in the periconception period. Human

Reprod Update 2010;16:80-95.

General Introduction I 11

16 Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol

2002;13:3-9.

17 Hu FB, Rimm E, Smith-Warner SA, et al. Reproducibility and validity of dietary patterns as-

sessed with a food-frequency questionnaire. Am J Clin Nutr 1999;69:243-9.

18 van Dam RM, Grievink L, Ocké MC, Feskens EJ. Patterns of food consumption and risk factors

for cardiovascular disease in the general Dutch population. Am J Clin Nutr. 2003 May;77:1156-

63.

19 Willett WC. Nutritional Epidemiology. New York: Oxford University Press; 1998.

20 Hegsted DM. Dietary standards--guidelines for prevention of deficiency or prescription for

total health? J Nutr 1986;116:478-81.

21 Lowik MR, Hulshof KF, Brussaard JH. Food-based dietary guidelines: some assumptions tested

for the Netherlands. Br J Nutr 1999;81:S143-9.

22 MRC. Prevention of neural tube defects: results of the Medical Research Council Vitamin

Study. Lancet 1991;338:131-7.

23 Netherlands HCot. Dietary standards: vitamin B6, folic acid and vitamin B12 [in Dutch]. The

Hague, the Netherlands: Health Council of the Netherlands; 2003. Report No.: publication nr.

2003/04.

24 Boxmeer JC, Macklon NS, Lindemans J, et al. IVF outcomes are associated with biomarkers of

the homocysteine pathway in monofollicular fluid. Hum Reprod (Oxford, England)

2009;24:1059-66.

25 Groenen PM, van Rooij IA, Peer PG, Ocké MC, Zielhuis GA, Steegers-Theunissen RP. Low

maternal dietary intakes of iron, magnesium, and niacin are associated with spina bifida in

the offspring.J Nutr. 2004 Jun;134:1516-22.

26 Krapels IP, van Rooij IA, Ocke MC, West CE, van der Horst CM, Steegers-Theunissen RP.

Maternal nutritional status and the risk for orofacial cleft offspring in humans. J Nutr

2004;134:3106-13.

PART I

Dietary patterns and reproductive performance

Associations between dietary patterns

and s

e

men

quality in men undergoing IVF/ICSI treatment

Marijana Vujković, Jeanne H. de Vries, Gerard R. Dohle, Gouke J. Bonsel, Jan Lindemans, Nick

S. Macklon, Peter J. van der Spek, Eric A. Steegers, and Régine P. Steegers-Theunissen

based on Hum Reprod 2009; 24: 1304 – 1312

CHAPTER 2

16 | Chapter 2

ABSTRACT

Objective

This study investigates whether dietary patterns, substantiated by biomarkers, are associated with

semen quality.

Methods

In 161 men of subfertile couples undergoing in vitro fertilization

treatment in a tertiary referral clinic in

Rotterdam, the Netherlands,

we assessed nutrient intakes and performed principal component

factor

analysis to identify dietary patterns. Total homocysteine

(tHcy), folate, vitamin B12, and vitamin B6 were

measured in blood and

seminal plasma. Semen quality was assessed by sperm volume,

concentration,

motility, morphology, and DNA fragmentation index

(DFI). Linear regression models analyzed associa-

tions between

dietary patterns, biomarkers and sperm parameters, adjusted

for age, body mass index

(BMI), smoking, vitamins and varicocele.

Results

The Health Conscious dietary pattern shows high

intakes of fruit, vegetables, fish, and whole grains.

The Traditional

Dutch dietary pattern is characterized by high intakes

of meat, potatoes and whole

grains and low intakes of no-alcoholic drinks

and sweets. The Health Conscious diet was inversely

correlated with tHcy in blood (β –0.07, p 0.02)

and seminal plasma (β –1.34, p 0.02) and positively

with

vitamin B6 in blood (β 0.217, p 0.01). An inverse

association was demonstrated between the Health

Conscious

diet and DFI (β –2.81, p 0.05). The Traditional

Dutch diet was positively correlated with red

blood cell

folate (β 0.06, p 0.04) and sperm concentration (β

13.25, p 0.01).

Conclusion

The Health Conscious and Traditional Dutch

dietary pattern seem to be associated with semen quality

in

men of subfertile couples.

Dietary Patterns and Semen Quality | 17

INTRODUCTION

Over the past decades, human fertility rates have declined dramatically in industrialized countries.

1

As

a result of abnormal semen characteristics, ~30% of all subfertile couples today require artificial fertility

treatment to address failed reproductive attempts.

2

The impact of risk factors like smoking and alcohol

use is undisputed, and also manifest malnutrition is known to affect semen.

3

As a reflection of global changes in dietary behaviour, the prevalence

of unhealthy diets, characterized

by low intakes of fruit and

vegetables and high intakes of foods rich in saturated fats,

has increased in

women and men within the reproductive age range.

4

However, so far nutritional studies

on semen

quality in men have rather focused on investigating

the role of zinc and the B-vitamin folate. Zinc is a

trace element

and acts as essential cofactor in metalloenzymes involved in

many processes of sper-

matogenesis and a shortage can lead to

oligospermia.

5

A folate deficiency results in

homocysteine

abundance and induces oxidative stress and apoptosis.

Recent findings show associations between

folate deficiency-dependent

and sperm aneuploidy,

6

sperm DNA damage,

7

and low sperm count.

8

As

expected, total homocysteine (tHcy) concentrations

in seminal plasma appear to be associated with

male subfertility

and low embryo quality.

9

The human diet contains a wide range of nutrients that act in

concert on many biological pathways.

Because nutrition is subject

to imposed environmental or, in humans, deliberate change, effects

on

reproductive health can be expected. Our aim of the present

study was to: 1) identify dietary patterns

in men

of subfertile couples undergoing in vitro fertilization (IVF)

or intracytoplasmic sperm injection

(ICSI) treatment; 2) relate

dietary patterns to biomarker concentrations of the homocysteine

pathway in

blood and seminal plasma; and 3) relate dietary

patterns as defined in 2) to semen quality, controlling

for

the risk factors of age, body mass index (BMI), smoking, vitamin

supplement use and presence of

varicocele. In our paper, the intake of alcohol is conceptually regarded as a nutriental exposure.

METHODS

Study Population

Between September 2004 and January 2007, subfertile couples

undergoing IVF/ICSI treatment at the

Erasmus

University Medical Centre in Rotterdam, the Netherlands,

were included in the prospective

FOod, Lifestyle and Fertility

Outcome-study (FOLFO-study). This study was designed to study

the

influence of preconception nutrition and lifestyle on fertility

and pregnancy outcome. Fertile and

subfertile men were eligible

for enrolment unless semen was cryopreserved or obtained by

microsur-

gical or percutaneous epididymal sperm aspiration. Of

the eligible IVF/ICSI population, 66% of the

fertile and subfertile

men participated in the FOLFO Study (n 251). Because of the

influence of ethnicity

on dietary habits and lifestyle, we included

in the current analysis male participants of European origin

only. This resulted in the evaluation of 161 male participants. The study protocol was approved by the

Dutch Central Committee for Human Research and the Medical Ethical and Institutional Review Board

of Erasmus University Medical Centre in Rotterdam, the Netherlands. All participants gave their written

informed consent.

18 | Chapter 2

General Questionnaire

At the IVF intake visit, the couples were invited to participate in the study, and only the men were

included in the current analysis. All participants filled out a general questionnaire at home comprising

information on current lifestyle and demographic factors and returned it on the next hospital ap-

pointment. The extracted data comprised of age, weight, height, medical history, education, use of

medication and lifestyle factors, such as smoking and the use of vitamin supplements. Between 2

weeks before and 2 weeks after oocyte retrieval, men visited the andrology outpatient clinic for fertility

evaluation comprising semen analysis, blood sampling, and physical examination including varicocele

detection. During physical examination, scrotal ultrasonography was performed using a Toshiba

Nemio 20 with a 12 Hz transducer. A varicocele was diagnosed if at least two venous vessels with a

diameter of 3 mm or more were present, in addition to reflux or diameter increase during Valsalva's

manoeuvre. All laboratory analyses were performed by personnel blinded to

the clinical diagnosis.

Male subfertility was defined by a sperm

concentration of ≤20 x 10

6

cells/ml.

Food Frequency Questionnaire

All participants filled out a food frequency questionnaire (FFQ) to estimate habitual food

and alcohol

intake of the previous 4 weeks. This semi-quantitative

FFQ was originally developed at the division of

Human Nutrition,

Wageningen University, Wageningen, the Netherlands and validated for intake

of

energy, fatty acids and B-vitamins.

19,20

The FFQs were provided on the

day of sperm sample collection

and returned on the day

of embryo transfer. A checklist was used to verify and check

the completeness

of the FFQ. Additional questions were asked

by telephone interview. The FFQ consists of 195 food

items and

is structured according to meal pattern. Questions in the FFQ

included frequency of con-

sumption, portion size and preparation

methods. Portion sizes were estimated according to Dutch

household

measures.

21

Nutritional values

were determined with the use of the Dutch food composi-

tion table

(NEVO) of 2001.

22

From the original 248

questionnaires filled out by the men, 2 out of 23 (8.7%) with

no embryo transfer

did not respond. Five of 225 men (2.2%) with

embryo transfer did not respond. Owing to these very

small numbers,

it is very unlikely that selection bias has occurred. To evaluate

the existence of under-

reporting the ratio of energy intake

divided by basal metabolic rate (BMR) was calculated. This value

is

an estimation of the physical activity level (PAL) of a sedentary

lifestyle. The equations of Schofield

were used for estimating

BMR.

23

According to Goldberg et al.

24

a cut-off point for underreporting for a

sedentary lifestyle

is a ratio of ≤1.35.

Semen Analysis

Semen samples were produced via masturbation into polypropylene

containers. Within half an hour,

the samples were liquefied,

and the semen parameters of volume, sperm concentration, sperm

count,

percentage progressive motility, and percentage normal morphology were assessed according to

World Health Organization guidelines.

25

The 1999 World Health Organization reference values for

normal sperm are fulfilled when sperm concentration exceeds 20x10

6

cells/ml, >50% of spermatozoa

have forward progression and >30% a normal morphology. Subsequently an aliquot of semen was

Dietary Patterns and Semen Quality | 19

centrifuged at 2500g for 10 minutes. The supernatant seminal plasma was frozen without preserva-

tives and stored at –20°C until assayed.

DNA Fragmentation Index

The principles and procedures of measuring sperm DNA damage

by a FACScan flow cytometry SCSA

have been described previously.

26

In short, semen samples were diluted with

TNE buffer [0.01 M Tris–

HCl, 0.15 M NaCl, 1 mM ethylenediamine

tetraacetate (EDTA), pH 7.4] to a concentration of 1-2

x 10

6

sperm cells/ml in a volume of 0.20 ml. This cell suspension

was mixed with 0.40 ml of acid detergent

solution (0.08 N HCl,

0.15 M NaCl, 0.1% Triton-X 100, pH 1.2) and then stained with

1.2 ml Acridine

Orange (AO) staining solution (0.1 M citric

acid, 0.2 M Na

2

PO

4

, 1 mM EDTA, 0.15 M NaCl, pH 6.0, contain-

ing

0.6 µl/L AO). A reference sample treated in the same way

was run prior to the actual measurements

and used to adjust

the voltage gains of the flow cytometer FL3 and FL1 photomultipliers

that detected

red and green fluorescence, respectively. An aliquot

of reference sample was stained and run again

after every 5-10

samples. Data collection of the fluorescent pattern in 5000

cells was performed at 3

minutes after acid treatment. Each sperm

sample was analyzed twice.

The extent of DNA damage was

expressed as the DNA fragmentation

index (DFI), reflecting the red to total

fluorescence ratio. Cell

Quest Pro and WinList software (Becton Dickinson,

San Jose, CA, US) were used to calculate the DFI of

each sample.

Biomarker Assay

Venous blood samples were drawn into dry vacutainer tubes and

allowed to clot. After centrifugation

at 2000g, the blood serum

was collected before being assayed for the concentrations of

folate, vitamin

B12 and testosterone. For the determination

of red blood cell (RBC) folate and tHcy, venous blood

samples

were drawn into EDTA containing vacutainer tubes. The EDTA-blood

samples were kept on

ice, and plasma was separated by centrifugation

within 1 hour for determination of tHcy. For the

determination

of vitamin B6, blood was drawn into lithium-heparin containing

vacutainers. Blood

serum and seminal plasma samples from each patient were

analyzed during routine laboratory

procedures for folate and

vitamin B12 using an immuno-electro-chemoluminescence assay (E170;

Roche Diagnostics GmbH, Mannheim, Germany). Directly after blood

sampling, 0.1 ml of blood out of

an EDTA tube was hemolyzed

with 0.9 ml of freshly prepared 1.0% ascorbic acid. Subsequently,

the

hematocrit of the EDTA-blood was determined on an ADVIA

120 Hematology Analyzer (Bayer Diagnos-

tics, Leverkusen, Germany).

The hemolysate was centrifuged for 5 minutes at 1000 g shortly before

the

folate measurement.

The folate concentration in the hemolysate

was recalculated in RBC folate using the following formula:

(nM hemolysate folate x 10 / hematocrit) – (nM serum folate

x [1 – hematocrit] / hematocrit) = nM RBC

folate. Vitamin

B6 levels in whole blood and seminal plasma and tHcy levels

in EDTA plasma and

seminal plasma were determined during routine

laboratory procedures using high-performance liquid

chromatography

with reversed-phase separation and fluorescence detection.

27,28

We determined

whole blood

vitamin B6 with pyridoxal-5'-phosphate (PLP) being the most

common form. Testoster-

one was measured using the Coat-a-Count radioimmunoassay (Diagnostic Products Corp, Los Ange-

20 | Chapter 2

les, CA, US). Sex hormone binding globulin (SHBG) was determined, using an immunometric tech-

nique on an Immulite Analyzer obtained from the same supplier. Serum inhibin B was measured by

immunoenzymometric assay (OBI, Oxford Bio-Innovation, Oxford, UK).

The between-run coefficient of variation for serum vitamin B12 was 5.1% at 125 pmol/L and 2.9% at

753 pmol/L; the coefficients of variation for serum folate were 9.5% at 8.3 nmol/L and 3.2% at 20.2

nmol/L; 3.3% at 14.55 µmol/L and 2.3% at 34.23 µmol/L for tHcy; 1.8% at 40 nmol/L and 1.3% at 115

nmol/L for PLP; for testosterone, these coefficients of variation were ≤7.5%, 6.1% at 11.6 nmol/L and

6.9% at 93 nmol/L for SHBG. For inhibin B, the coefficients of variation were ≤15%. The detection limit

was 1.36 nmol/L for folate, 22 pmol/L for vitamin B12, 5 nmol/L for PLP, 4 µmol/L for tHcy, 0.1 nmol/L

for testosterone, 5 nmol/L for SHBG and 10 ng/L for inhibin B.

Statistics

Because of the prior knowledge of high correlations among several

food constituents, we used

principal component factor analysis (PCA)

to summarize dietary patterns from food consumption

data.

4,29

In summary, all 195 food items

from the FFQ data of men undergoing IVF/ICSI treatment were

first reduced to 22 predefined food groups based on a nutrient

content grouping, conform to other

reported schemes.

30

Hereafter food groups were adjusted

for the total intake of energy,

31

and subse-

quently

PCA was performed. To keep rotated factors uncorrelated, the

solution was rotated with the

varimax method by maximizing the

sum of the variance of the loading vectors.

32

The

first two factors

were extracted, and each represents a distinct

dietary pattern. A factor consists of a selection of the

initial

variables, each with its own coefficient (here called loading),

which defines the observed correla-

tion of that variable with

the latent constructed factor. As a weighted ‘mix’

of the initial variables, a

factor explains a substantial amount

of variation in the data set under study. In the preferable

case, the

statistically derived factor represents a recognizable

pattern in the observable world. The Eigenvalue

was used as

indicator of the amount of variation explained by each factor.

Each participant was

assigned two personalized scores for the

two factors, i.e. which represents a quantification of the

similarity

of the individual's diet with each of the two extracted factors. The factor loadings, i.e. the

association of a factor with all

measured food components, are presented for each factor separately;

the association is calculated by Pearson's r correlation coefficient

(Table 1).

After computation of the personalized scores,

all 161 men were classified into tertiles according to

their

personal score for the respective dietary pattern. Additional explanatory variables were conven-

tionally described. However, as age, body mass index (BMI, defined as kg/m

2

) and energy intake

showed skewed

distributions even if log-transformed, these variables are described

by the median. For

the same reason, we displayed medians and

inter-quartile ranges for the biomarkers. Differences in

general characteristics between tertiles were evaluated in an unconditional linear regression model.

The prevalence of causes of subfertility, ethnic background, smoking and vitamin use were related to

each tertile of the respective dietary pattern score, and the dietary associations with this prevalence

were tested with chi-square test for linear association.

Dietary Patterns and Semen Quality | 21

Next, the diet was used to predict the logarithmically transformed

biomarkers of the homocysteine

pathway and semen quality parameters

in a multivariable linear regression model, additionally ad-

justed

for the potential confounding variables age, BMI, smoking, intake of multivitamin and/or folic

acid supplements and the presence

of a varicocele. Individual food groups were also analyzed in

a

linear regression model to predict semen parameters and investigate

which food groups of the dietary

pattern contribute most to

the observed associations with semen parameters.

Table 1 | Food group factor loadings for two identified dietary patterns from food-frequency

questionnaire data of 160 men undergoing IVF/ICSI treatment

food group Health Conscious diet Traditional Dutch diet

alcohol – 0.10 – 0.19

1

cereals – 0.09 – 0.28

2

butter 0.00 0.03

dairy

– 0.04 0.04

eggs 0.00 0.12

fish

– 0.56

2

– 0.14

fruit 0.80

2

– 0.16

1

legumes 0.16

1

0.05

margarine – 0.13 0.18

1

mayonnaise – 0.20

1

0.22

2

meat

– 0.17

1

0.47

2

non-alcoholic drinks

0.06 – 0.71

2

nuts 0.12 0.00

refined grains – 0.16

1

– 0.13

potatoes – 0.01 0.56

2

sauces 0.11 0.05

snacks – 0.08 – 0.03

soup 0.08 – 0.23

2

sweets – 0.17

1

– 0.46

2

vegetable oil 0.09 0.03

vegetables 0.74

2

0.12

whole grains 0.43

2

0.48

2

variance explained 11.7% 9.5%

1

p ≤0.05,

2

p ≤0.01

Covariates

were age, BMI, smoking, intake of multivitamin and/or folic

acid supplements. The food

groups with all β estimates

and 95% confidence intervals (95%CI) described. All analysis was performed

with SPSS software,

release 15.0.0 for Windows (SPSS Inc., Chicago, IL, US).

22 | Chapter 2

RESULTS

The characteristics of the dietary intake of 161 male participants,

in terms of the two selected factors

(dietary patterns), are shown

in Table 1. The first factor explained 11.7% of the total

variance. After

inspection of the associative pattern, it was

labelled the Health Conscious diet;

it comprises high

intakes of fruit, vegetables, fish, whole grains, and legumes and low intakes of mayonnaise

and other

fatty sauces, meat, refined grains, and sweets. The second factor explained 9.5% of the total

variance of

dietary intake, and in our interpretation represented

the Traditional Dutch diet; it is characterized

by

high intakes of potatoes, meat, whole grains, margarine,

mayonnaise and other fatty sauces and low

intakes of non-alcoholic

and alcoholic drinks, cereals, fruit, soup, and sweets. Neither subgroups of the

Health Conscious

nor the Traditional Dutch diet showed differences

in general characteristics or

endocrine parameters (Table 2).

Initial significant correlations were present between the Health

Conscious diet and serum and RBC

folate, tHcy, vitamin

B6 and vitamin B12, but after adjusting for confounders, significant

correlations

only remained for tHcy (β –0.07, p 0.02) and vitamin B6 (β 0.22, p 0.01) as shown in

Table 3. In seminal

plasma, the adjusted linear model

revealed an inverse association between tHcy and the Health

Conscious diet (β –1.34, p 0.02). The Traditional Dutch pattern was positively correlated

with RBC folate

(β 0.06, p 0.05, p

adj

0.04),

and no significant associations were observed with biomarkers

in seminal

plasma.

Table 4 reveals that the use of the Health Conscious

diet is inversely associated with sperm

DNA damage (β –2.81, p 0.05). This effect seems to be explained by

the high intakes of fruit (β –0.13,

95%CI –0.25;

–0.02) and vegetables (β –0.25, 95%CI –0.49;

–0.01) (Table 5). The Traditional Dutch

diet

showed an increase in sperm concentrations (β 13.25,

p 0.01) (Table 4). The high intake of potatoes (β

0.82, 95%CI 0.24; 1.4) and low intake of non-alcoholic drinks

(β –0.29, 95%CI –0.49; –0.1) seem to

explain

this finding (Table 5). All significance levels denoted

have been adjusted for age, BMI, smoking,

vitamin use and presence

of varicocele at the study moment.

DISCUSSION

This study demonstrates that human nutrition affects semen quality

in men undergoing IVF/ICSI

procedures. We observed that men

who consume a Health Conscious diet have lower

sperm DNA

damage. Furthermore, sperm concentrations are much

higher in men who strongly adhere to the

Traditional

Dutch dietary pattern. These findings suggest that the two

dietary patterns indeed have

beneficial effects on semen

quality. Our findings also represent important epidemiological

information

as it provides a specific explanation and empirical

evidence on the putative relation between the

increased unhealthy food availability and

the decline of semen quality in industrialized countries.

We were surprised by the observation that the Health

Conscious dietary pattern is inversely related to

tHcy

in blood and seminal plasma, but positively related to vitamin

B6 in blood. This may imply that

this dietary pattern prohibits

sperm DNA from being damaged by regulating the flow of tHcy

concen-

trations.

Dietary Patterns and Semen Quality | 23

Table 2 | Baseline characteristics of the Health Conscious and Traditional Dutch diet

Health Conscious diet

characteristic

low

(n 53)

intermediate

(n 55)

high

(n 53)

p

age

1,2

35.3 (29.1- 53.9) 35.8 (29.8 - 46.6) 37.3 (28.6 - 50.5) 0.13

BMI

1,3

25.0 (18.4 - 37.9) 26.0 (19.6 - 37.1) 24.9 (18.8 - 34.8) 0.38

cause of subfertility

4

0.69

male factor

20 (37.7) 20 (36.4) 17 (32.1)

female factor 13 (24.5) 11 (20.0) 7 (13.2)

both 3 (5.7) 4 (7.2) 4 (7.5)

unexplained 17 (32.1) 20 (36.4) 25 (47.2)

presence of varicocele

5

10 (20.8) 8 (17.8) 8 (17.4) 0.90

Dutch ethnicity 46 (86.8) 45 (81.8) 44 (83.0) 0.77

smoking 14 (26.4) 14 (25.5) 6 (11.3) 0.10

vitamin use 12 (22.6) 13 (23.6) 17 (32.1) 0.27

energy intake

1,6

10.8 (4.5 - 19. 4) 9.2 (3.6 - 18.1) 10.5 (5.3 - 15.5) 0.86

testosterone

7,8

16.0 (14.4 - 17.7) 14.6 (13.3 - 16.2) 14.9 (13.8 - 16.1) 0.96

SHBG

7,8

28.2 (25.0 - 31.8) 23.1 (20.7 - 25.8) 28.0 (24.6 - 31.7) 0.35

inhibin B

7,9

172 (150 - 195) 171 (151 - 191) 175 (152 - 199) 0.79

Traditional Dutch

characteristic

low

(n 54)

intermediate

(n 54)

high

(n 53)

p

age

1,2

36.3 (30.0 - 50.5) 36.3 (28.6 - 46.8) 35.6 (28.7 - 53.9) 0.91

BMI

1,3

24.8 (18.4 - 37.9) 24.7 (18.8 - 33.8) 26 (19.9 - 37.1) 0.07

cause of subfertility

4

0.11

male factor

14 (25.9) 27 (50.0) 16 (30.2)

female factor 14 (25.9) 9 (16.7) 8 (15.1)

both 5 (9.2) 2 (3.7) 4 (7.5)

unexplained 21 (38.9) 16 (29.6) 25 (47.2)

presence of varicocele

5

12 (26.1) 4 (8.2) 10 (22.7) 0.06

Dutch ethnicity 44 (81.5) 46 (85.2) 45 (84.9) 0.84

smoking 12 (22.2) 9 (16.7) 13 (24.5) 0.59

vitamin use 16 (29.6) 14 (25.9) 12 (22.6) 0.41

energy intake

1,6

10.6 (5.7 - 19.4) 9.6 (3.6 - 18.1) 10.8 (5.8 - 18.6) 0.36

testosterone

7,8

15.2 (13.9 - 16.5) 14.9 (13.5 - 16.4) 15.5 (14.1 - 17.2) 0.99

SHBG

7,8

28.0 (24.8 - 31.6) 26.2 (23.6 - 29.2) 24.4 (21.4 - 27.9) 0.19

inhibin B

7,9

174 (153 - 194) 174 (151 - 198) 170 (148 - 193) 0.86

Data are presented by

1

median (range),

2

years,

3

kg/m

2

,

4

n (%),

5

presence of varicocele was missing in 22 men,

6

mJoule/day,

7

mean (95% CI),

8

nmol/L,

9

ng/L.

24 | Chapter 2

We also found that the beneficial effects of

the Health Conscious pattern seem to be determined

in

particular by vegetable and fruit intake, as has been further

substantiated by the single food group

analyses. The high consumption

of folate- and vitamin B6-rich foods may stimulate tHcy to be

re-

methylated into methionine and trans-sulfurated into cystathionine

and cysteine.

In males, the regulation of homocysteine is mediated

partially through the testosterone-dependent

cystathionine-β-synthase

pathway.

33

Other studies support the

importance of the homocysteine

pathway on semen parameters.

A Spanish study revealed positive associations between the intake

of

folate-rich food sources, such as fruit and vegetables, and

semen quality.

34

Furthermore, Young et

al.

6

observed inverse relationships between folate, zinc

and antioxidant intake in healthy non-smoking

men in the US

and sperm aneuploidy. Suggested mechanisms by which elevated

tHcy may exert its

effects are excessive induction of oxidative

stress, defective methylation of proteins, lipids, and DNA,

altered

nitric oxide bioavailability, induction of vascular inflammation,

and activation of apoptosis.

3

However, these

mechanisms in reproductive tissues should be studied in much

more detail in future

studies, both experimentally in animals

and observationally in human cohorts.

Table 3 | Concentration of biomarkers in blood and seminal plasma in two dietary patterns

1

Health Conscious diet

biomarker

low

(n 32)

intermediate

(n 21)

high

(n 26)

p

5

blood serum folate

2

16.1 (5.6) 15.3 (6.5) 18.5 (8.9) 0.15

RBC folate

2

1006 (350) 1063 (536) 1107 (542) 0.10

vitamin B6

2

81.5 (36.8) 77.0 (35.6) 94.0 (48.3) 0.01

vitamin B12

3

288 (94) 289 (188) 361 (155) 0.07

tHcy

4

12.2 (3.2) 11.7 (4.1) 10.8 (4.4) 0.02

semen folate

2

26.5 (11.6) 25.7 (16.4) 25.0 (15.5) 0.70

vitamin B6

2

30.5 (39.0) 30.0 (23.5) 28.5 (30.2) 0.30

vitamin B12

3

467 (608) 424 (450) 493 (339) 0.51

tHcy

4

4.1 (3.1) 4.4 (3.2) 3.7 (4.5) 0.02

Traditional Dutch diet

biomarker low intermediate high p

5

blood serum folate

2

17.3 (7.7) 16.2 (6.7) 15.7 (9.1) 0.84

RBC folate

2

990 (554) 986 (382) 1110 (511) 0.04

vitamin B6

2

99.0 (48.0) 78.5 (31.0) 77.0 (36.5) 0.81

vitamin B12

3

329 (172) 293 (148) 282 (158) 0.41

tHcy

4

11.9 (4.7) 11.2 (3.1) 12.3 (3.7) 0.84

semen folate

2

28.3 (14.8) 24.8 (10.5) 22.6 (15.8) 0.69

vitamin B6

2

33.0 (32.0) 29.0 (21.3) 29.0 (33.5) 0.42

vitamin B12

3

409 (406) 438 (400) 680 (440) 0.44

tHcy

4

3.7 (3.9) 4.0 (2.5) 4.7 (5.7) 0.62

1

mean(sd),

2

nmol/L,

3

pmol/L

4

µmol/L.

5

Adjusted for age, BMI, smoking, vitamins, varicocele.

Dietary Patterns and Semen Quality | 25

Unfortunately, no data

on seminal poly-unsaturated fatty acid (PUFA) concentrations were collected,

which would have

enabled us to measure whether the increased fish intake resulted

in higher levels of

unsaturated fats that might enhance the

fluidity of the plasma membrane. A fluid membrane, how-

ever,

also makes spermatozoa more vulnerable to ROS attack, which

may explain why oxidative stress-

inducing lifestyle factors may have led to reduced human fertility in the past decades.

The second identified dietary pattern being the Traditional

Dutch comprises high intakes of meat,

potatoes,

and whole grains. This resembles a typically Dutch dietary tradition

up to the 19

th

century

when agriculture and domestic education

were widely implemented in the Netherlands. This dietary

pattern

is positively correlated with folate in RBC and shows a significant

positive association with

sperm concentrations. Because of the substantial amount of potatoes eaten in one serving, they

provide

a rich source of folate. Furthermore, the high intake of meat,

a natural source of zinc elements,

influences the bioavailability

of dietary folate. The zinc-dependent enzyme γ-glutamylhydrolase

in the

jejunum efficiently converts dietary folate as poly-glutamates

into mono-glutamates, which are the

only absorbable form of folate.

Furthermore, zinc is a cofactor of methionine synthase involved

in the

remethylation of tHcy into methionine thereby reducing

tHcy. The beneficial effects of zinc are sup-

ported by a randomized,

placebo-controlled intervention study conducted in the late

nineties, in

which administration of zinc and/or folic acid

to subfertile patients led to a significant increase in

semen

concentration ranging from 18% to 74%.

9,12,17

Table 4 | Semen parameters according to the tertiles of the two major dietary patterns

1

Health Conscious diet

semen parameter

low

(n 40)

intermediate

(n 44)

high

(n 42)

p

5

DFI

2

25.2 (21.1 - 29.2) 25.0 (20.2 - 29.8) 20.6 (17.5 - 23.6) 0.05

volume

3

3.4 (2.9 - 3.8) 2.8 (2.4 - 3.2) 3.1 (2.6 - 3.5) 0.41

concentration

4

46.2 (30.8 - 61.5) 39 (27.6 - 50.4) 48.2 (33.1 - 63.2) 0.89

motility

2

35 (30 - 40) 33 (28 - 39) 37 (32 - 42) 0.06

morphology

2

5 (4 - 6) 5 (4 - 6) 5 (4 - 6) 0.74

Traditional Dutch diet

semen parameter

low

(n 44)

intermediate

(n 43)

high

(n 39)

p

5

DFI

2

23.5 (20.1 - 26.9) 25.0 (20.0 - 30.1) 22.0 (18.5 - 25.6) 0.53

volume

3

3.1 (2.6 - 3.5) 3.1 (2.7 - 3.5) 3.0 (2.6 - 3.5) 0.89

concentration

4

36.7 (26.8 - 46.6) 36.4 (23.1 - 49.8) 61.7 (44.5 - 78.9) 0.01

motility

2

35 (31 - 39) 34 (29 - 40) 37 (31 - 43) 0.98

morphology

2

5 (4 - 6) 5 (4 - 6) 6 (5 - 7) 0.34

Data are presented by

1

median with range,

2

percentages,

3

ml,

4

x10

6

cells/ml.

5

Adjusted for age, BMI, smoking,

vitamin use and presence of a varicocele.

26 | Chapter 2

In the same FOLFO study, but in a much smaller study group of

men, we measured the biomarkers of

homocysteine metabolism in

blood and seminal plasma and investigated the associations with

semen

parameters. A positive association was observed between

seminal plasma vitamin B12 and sperm

concentration.

35

In the current analysis, we demonstrate a significant

association between the Tradi-

tional Dutch dietary

pattern and semen concentrations (increase from 36.7x10

6

to

61.7x10

6

cells/ml in

the highest tertile, p 0.01). We also

observe positive but not statistically significant associations

be-

tween the Traditional Dutch dietary pattern

and seminal vitamin B12 concentrations, which are very

likely

to be due to the high meat intake. This supports the previous

observed association between

seminal vitamin B12 and sperm concentration.

In a recent study, Mendiola et al.

34

suggested that meat

intake negatively affects semen quality. They

assumed this detrimental

effect to be due to natural and synthetic estrogens in meat,

hormones that

are used in cattle to stimulate growth and development.

36

However, their study size was rather

small (n

61) and only single food items were investigated.

There was no possibility of comparing dosage effects

because only reports of frequency were used to assess the

dietary intake, and portion sizes were not

reported. We can

only speculate that the favourable effect of the meat-rich Traditional

Dutch diet on

semen quality might be due to the relative

lower amounts of meat used in our country, or perhaps a

lower

exposure of Dutch cattle to hormones and endocrine disruptors.

Some concerns regarding the study design apply. We cannot exclude

the possibility that the dietary

patterns identified are confounded

by lifestyle factors other than the adjusted covariates age,

energy

intake, BMI, smoking and use of vitamin supplements.

We observed little or no association of these

factors with the

Health Conscious and Traditional Dutch

dietary pattern, but cannot rule out a possible

role of other

residual confounding variables, e.g. exercise, psychological

stress, and exposure to

environmental pollutants.

3

We have reduced the possibility of multiple testing by

summarizing the

overall nutrient intake with PCA into two dietary

patterns. By limiting our analysis to only two exposure

variables and

taking all nutrients into account, we have avoided the issue

of selective testing. These

two exposure variables were associated

with biomarkers of the homocysteine metabolism and semen

parameters,

whereby we have attempted to minimize the probability of chance-finding.

All methods of dietary assessment are prone to error. Nutritional

intake was assessed with an FFQ and

individually checked for

completeness and accuracy. However, self-reported dietary intake

tends to

habitually underestimate energy intake.

37

Therefore, we evaluated the presence of underreporting

by

estimating the PAL. The mean PAL for all participants in

our study was 1.28, and similar in fertile and

subfertile men

indicating an underestimation of energy intake in both groups

according to Goldberg

standard (PAL ≤1.35).

24

We judge it

therefore to be unlikely that underestimation has distorted

the

associations between semen parameters and energy.

We recognize that the findings from this Dutch sample and from

comparable European subfertile

samples may not apply to other

populations, particularly if patterns in dietary intake are dissimilar or

foods are differently produced and processed. Still our results

apply to a substantial part of the globe

Dietary Patterns and Semen Quality | 27

with westernized

dietary patterns and lifestyle. The strength of the study is its prospective

design and

the relatively large study sample size of 161 men.

This gave us the opportunity to produce reliable

dietary patterns

which were comparable to studies conducted in the Dutch general

population.

38

General Conclusion

The Health Conscious and the Traditional

Dutch dietary pattern seem to be associated with semen

quality. More studies are needed to show the causation of these

associations and their effects on

human reproduction.

Table 5 | Contribution of individual food groups to semen parameters

semen parameter

food group

DFI volume concentration motility morphology

alcohol

0.0 (-0.1; 0.1)

0.0 (-0.1; 0.2) –

0.0 (–0.1; 0.1)

0.0 (-0.1; 0.1)

cereals

–

0.4 (-1.1; 0.4) –

0.1 (-0.1; 0.1)

1.0 (-1.3; 3.2)

0.5 (0.2; 1.3)

0.0 (-0.2; 0.2)

butter

–

4.1 (-13; 4.8) –

0.4 (-1.3; 0.4)

2.8 (–24; 30)

7.6 (-1.7; 17) –

1.4 (-3.2; 0.4)

dairy

0.1 (-0.3; 0.4)

0.0 (0.0; 0.1)

0.5 (–0.6; 1.6) –

0.1 (-0.5; 0.3)

0.0 (-0.1; 0.1)

eggs

–

0.3 (-0.9; 0.3)

0.1 (0.0; 0.1)

1

–

0.2 (–2.1; 1.7) –

0.1 (-0.7; 0.5)

0.0 (-0.1; 0.1)

fish

–

0.7 (-2.0; 0.6) –

0.1 (-0.2; 0.1)

0.9 (–3.3; 5.1)

1.3 (0.0; 2.6)

1

0.1 (-0.1; 0.4)

fruit

–

0.1 (-0.3; -0.0)

1

0.0 (-0.1; 0.1) –

0.1 (–0.5; 0.3)

0.2 (0.1; 0.3)

1

0.0 (-0.1; 0.1)

legumes

–

0.5 (-1.2; 0.1) –

0.0 (-0.1; 0.1)

0.2 (-1.7; 2.1)

0.1(-0.5; 0.8)

0.0 (-0.1; 0.1)

margarine

4.4 (-2.3; 11)

0.5 (-0.2; 1.2) –

3.9 (-26; 18) –

9.1 (-16; -1.5)

1

0.2 (-1.3; 1.7)

mayonnaise

0.1 (-1.3; 1.6) –

0.2 (-0.3; -0.1)

1

–

0.1 (-4.3; 4.2) –

0.8 (-2.3; 0.6) –

0.1 (-0.4; 0.2)

meat

1.0 (-0.1; 2.0)

0.1 (-0.1; 0.2) –

2.2 (-5.7; 1.4) –

1.0 (-2.2; 0.2) –

0.1 (-0.4; 0.1)

non-

alc. drinks

0.0 (-0.1; 0.1)

0.0 (-0.1; 0.1) –

0.3 (-0.5; -0.1)

1

–

0.0 (-0.1; 0.1)

nuts

–

0.8 (-1.6; -0.1)

1

–

0.1 (-0.2; 0.1) –

1.0 (-3.4; 1.3)

0.2 (-0.6; 1.0)

0.1 (-0.1; 0.2)

refined grains

–

0.1 (-0.5; 0.2) –

0.0 (-0.1; 0.1)

0.3 (-0.7; 1.3) –

0.1 (-0.5; 0.2)

0.0 (-0.1; 0.1)

potatoes

–

0.1 (-0.2; 0.1)

0.0 (-0.1; 0.1)

0.8 (0.2; 1.4)

1

0.1 (-0.1; 0.3)

0.0 (-0.1; 0.1)

sauces

–

0.8 (-1.8; 1.6) –

0.1 (-0.3; 0.1)

0.2 (-5.3; 5.7) –

0.8 (-2.6; 1.0) –

0.0 (-0.4; 0.3)

snacks

0.1 (-0.4; 0.5)

0.0 (-0.1; 0.1) –

0.6 (-2.1; 1.0)

0.2 (-0.3; 0.7) –

0.1 (-0.2; 0.0)

soup

0.0 (-0.1; 0.1)

0.0 (-0.1; 0.1) –

0.2 (-0.4; -0.0)

1

–

0.0 (-0.1; 0.1)

sweets

0.4 (-0.4; 1.3) –

0.0 (-.01; 0.1)

0.3 (-2.5; 3.1) –

0.1 (-0.9; 0.8) –

0.1 (-0.2; 0.1)

vegetable oil

–

1.1 (-5.2; 3.0) –

0.3 (-0.7; 0.2) –

3.3 (-17; 10)

1.1 (-3.4; 5.6) –

0.2 (-1.1; 0.7)

vegetables

–

0.3 (-0.5; -0.0)

1

–

0.0 (-0.1; 0.1)

0.8 (0.0; 1.6)

1

0.3 (0.1; 0.6)

1

0.0 (-0.1; 0.1)

whole grains

–

0.2 (-0.3; 0.0)

0.0 (-0.1; 0.1)

0.4 (-0.2; 1.0)

0.3 (-0.0; 0.4)

0.0 (-0.1; 0.1)

Data are presented by mean with 95% CI.

1

p ≤0.05, after adjustments for age, BMI, smoking and vitamin use.

28 | Chapter 2

REFERENCES

1 The World Bank Group. url: http://www.worldbank.org, 2005.

2 Wong WY, Thomas CM, Merkus JM, Zielhuis GA, Steegers-Theunissen RP. Male factor subfertil-

ity: possible causes and the impact of nutritional factors. Fertil Steril 2000;73:435-42.

3 Homan GF, Davies M, Norman R. The impact of lifestyle factors on reproductive performance

in the general population and those undergoing infertility treatment: a review. Hum Reprod

Update 2007;13:209-23.

4 Vujkovic M, Ocke MC, van der Spek PJ, Yazdanpanah N, Steegers EA, Steegers-Theunissen RP.

Maternal Western dietary patterns and the risk of developing a cleft lip with or without a cleft

palate. Obstet Gynecol 2007;110:378-84.

5 Favier AE. The role of zinc in reproduction. Hormonal mechanisms. Biol Trace Elem Res

1992;32:363-82.

6 Young SS, Eskenazi B, Marchetti FM, Block G, Wyrobek AJ. The association of folate, zinc and

antioxidant intake with sperm aneuploidy in healthy non-smoking men. Hum Reprod 2008;

23:1014-22.

7 Boxmeer JC, Smit M, Utomo E, et al. Low folate in seminal plasma is associated with increased

sperm DNA damage. Fertil Steril 2008;92:548-56.

8 Wallock LM, Tamura T, Mayr CA, Johnston KE, Ames BN, Jacob RA. Low seminal plasma folate

concentrations are associated with low sperm density and count in male smokers and non-

smokers. Fertil Steril 2001;75:252-9.

9 Ebisch IM, Peters WH, Thomas CM, Wetzels AM, Peer PG, Steegers-Theunissen RP. Homocys-

teine, glutathione and related thiols affect fertility parameters in the (sub)fertile couple. Hum

Reprod 2006;21:1725-33.

10 Comhaire FH, Christophe AB, Zalata AA, Dhooge WS, Mahmoud AM, Depuydt CE. The effects

of combined conventional treatment, oral antioxidants and essential fatty acids on sperm

biology in subfertile men. Prostaglandins Leukot Essent Fatty Acids 2000;63:159-65.

11 Agarwal A, Said TM. Oxidative stress, DNA damage and apoptosis in male infertility: a clinical

approach. BJU Int 2005;95:503-7.

12 Wong WY, Merkus HM, Thomas CM, Menkveld R, Zielhuis GA, Steegers-Theunissen RP. Effects

of folic acid and zinc sulfate on male factor subfertility: a double-blind, randomized, placebo-

controlled trial. Fertil Steril 2002;77:491-8.

13 Rolf C, Cooper TG, Yeung CH, Nieschlag E. Antioxidant treatment of patients with asthenozoo-

spermia or moderate oligoasthenozoospermia with high-dose vitamin C and vitamin E: a

randomized, placebo-controlled, double-blind study. Hum Reprod 1999;14:1028-33.

14 Moilanen J, Hovatta O, Lindroth L. Vitamin E levels in seminal plasma can be elevated by oral

administration of vitamin E in infertile men. Int J Androl 1993;16:165-6.

15 Kessopoulou E, Powers HJ, Sharma KK, et al. A double-blind randomized placebo cross-over

controlled trial using the antioxidant vitamin E to treat reactive oxygen species associated

male infertility. Fertil Steril 1995;64:825-31.

16 Ebisch IM, Thomas CM, Peters WH, Braat DD, Steegers-Theunissen RP. The importance of

folate, zinc and antioxidants in the pathogenesis and prevention of subfertility. Hum Reprod

Update 2007;13:163-74.

17 Ebisch IM, Pierik FH, FH DEJ, Thomas CM, Steegers-Theunissen RP. Does folic acid and zinc

sulphate intervention affect endocrine parameters and sperm characteristics in men? Int J

Androl 2006;29:339-45.

18 Bentivoglio G, Melica F, Cristoforoni P. Folinic acid in the treatment of human male infertility.

Fertil Steril 1993;60:698-701.

Dietary Patterns and Semen Quality | 29

19 Verkleij-Hagoort AC, de Vries JH, Ursem NT, de Jonge R, Hop WC, Steegers-Theunissen RP.

Dietary intake of B-vitamins in mothers born a child with a congenital heart defect. Eur J Nutr

2006;45:478-86.

20 Feunekes GI, Van Staveren WA, De Vries JH, Burema J, Hautvast JG. Relative and biomarker-

based validity of a food-frequency questionnaire estimating intake of fats and cholesterol. Am

J Clin Nutr 1993;58:489-96.

21 Donders-Engelen M, van der Heijden L, Hulshof KF. Maten Gewichten en Codenummers. In:

University HNoTaW, ed. Informatorium voor voeding en dietetiek. Wageningen: Human

Nutrition of TNO and Wageningen University, 2003.

22 Netherlands Nutrition Centre. NEVO: Dutch food composition database. The Hague, the

Netherlands, 2001.

23 Schofield WN. Predicting basal metabolic rate, new standards and review of previous work.

Hum Nutr Clin Nutr 1985;39:5-41.

24 Goldberg GR, Black AE, Jebb SA, et al. Critical evaluation of energy intake data using funda-

mental principles of energy physiology: 1. Derivation of cut-off limits to identify under-

recording. Eur J Clin Nutr 1991;45:569-81.

25 World Health Organization. WHO laboratory manual for the examination of human semen and

sperm–cervical mucus interaction. 4 ed. Cambridge: Cambridge University Press, 1999.

26 Smit M, Dohle GR, Hop WC, Wildhagen MF, Weber RF, Romijn JC. Clinical correlates of the

biological variation of sperm DNA fragmentation in infertile men attending an andrology

outpatient clinic. Int J Androl 2007;30:48-55.

27 Schrijver J, Speek AJ, Schreurs WH. Semi-automated fluorometric determination of pyridoxal-

5'-phosphate (vitamin B6) in whole blood by high-performance liquid chromatography

(HPLC). Int J Vitam Nutr Res 1981;51:216-22.

28 Pfeiffer CM, Huff DL, Gunter EW. Rapid and accurate HPLC assay for plasma total homocysteine

and cysteine in a clinical laboratory setting. Clin Chem 1999;45:290-2.

29 Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol

2002;13:3-9.

30 Slimani N, Fahey M, Welch AA, et al. Diversity of dietary patterns observed in the European

Prospective Investigation into Cancer and Nutrition (EPIC) project. Public Health Nutr

2002;5:1311-28.

31 Willett WC, Howe GR, Kushi LH. Adjustment for total energy intake in epidemiologic studies.

Am J Clin Nutr 1997;65:1220S-1228S; 1229S-1231S.

32 Kaiser HF. The Varimax Criterion for Analytic Rotation in Factor Analysis. Psychometrika

1958;23:187-200.

33 Vitvitsky V, Prudova A, Stabler S, Dayal S, Lentz SR, Banerjee R. Testosterone regulation of renal

cystathionine beta-synthase: implications for sex-dependent differences in plasma homocys-

teine levels. Am J Physiol Renal Physiol 2007;293:F594-600.

34 Mendiola J, Torres-Cantero AM, Moreno-Grau JM, et al. Food intake and its relationship with

semen quality: a case-control study. Fertil Steril 2008;91:812-8.

35 Boxmeer JC, Smit M, Weber RF, et al. Seminal plasma cobalamin significantly correlates with

sperm concentration in men undergoing IVF or ICSI procedures. J Androl 2007;28:521-527.

36 Henricks DM, Gray SL, Owenby JJ, Lackey BR. Residues from anabolic preparations after good

veterinary practice. Apmis 2001;109:273-83.

37 Black AE, Goldberg GR, Jebb SA, Livingstone MB, Cole TJ, Prentice AM. Critical evaluation of

energy intake data using fundamental principles of energy physiology: 2. Evaluating the

results of published surveys. Eur J Clin Nutr 1991;45:583-99.

38 van Dam RM, Grievink L, Ocke MC, Feskens EJ. Patterns of food consumption and risk factors

for cardiovascular disease in the general Dutch population. Am J Clin Nutr 2003;77:1156-63.

The preconception Mediterranean dietary pattern

in couples undergoing IVF/ICSI treatment

increases the chance of pregnancy

Marijana Vujković, Jeanne H. de Vries, Jan Lindemans, Nick S. Macklon, Peter J. van der

Spek, Eric A. Steegers, and Régine P. Steegers-Theunissen

based on Fertil Steril 2010 (Feb 18

th

, epub ahead of print)

CHAPTER

3

32 | Chapter 3

ABSTRACT

Objective

To investigate associations between preconception dietary patterns and IVF/ICSI outcomes validated

by biomarkers of the homocysteine pathway.

Methods

An observational prospective study was set up in a tertiary referral fertility clinic at the Erasmus Univer-

sity Medical Centre, Rotterdam, the Netherlands. One hundred sixty-one couples undergoing IVF/ICSI

treatment were included. Main outcome measures consisted of dietary patterns, blood and follicular

fluid concentrations of folate, vitamin B12, vitamin B6, and homocysteine, fertilization rate, embryo

quality, and pregnancy.

Results

Two dietary patterns were identified in women. The Health Conscious – Low Processed dietary pattern

(variance explained 12.1%) was characterized by high intakes of fruit, vegetables, fish, and whole grains

and low intakes of snacks, meats, and mayonnaise, and positively correlated with red blood cell folate

(β 0.07). The Mediterranean dietary pattern (variance explained 9.1%), that is, high intakes of vegetable

oils, vegetables, fish, and legumes and low intakes of snacks, was positively correlated with red blood

cell folate (β 0.13), and vitamin B6 in blood (β 0.09) and follicular fluid (β 0.18). High adherence by the

couple to the Mediterranean diet increased the probability of pregnancy with approximately 40%,

odds ratio 1.4 (95%CI1.0; 1.9).

Conclusion

The adherence to a preconception Mediterranean diet in couples undergoing IVF/ICSI treatment

contributes to the success of achieving pregnancy.

Dietary Patterns and Chances of Pregnancy | 33

INTRODUCTION

Subfertility is an increasing problem in the reproductive population of industrialized countries, primar-

ily because of postponed childbearing.

1

This problem is accentuated by unhealthy lifestyles, such as

smoking, alcohol use, and malnutrition.

2

A nutritionally unbalanced diet characterized by low intakes

of minerals and vitamins has previously been associated with adverse fertility outcomes.

3

Especially the

B-vitamins folate, vitamin B6, and vitamin B12, are important because of their role in the homocysteine

pathway.

4-6

A deficiency of these vitamins may cause an accumulation of homocysteine concentra-

tions, which can ultimately lead to hyperhomocysteinemia.

3,5-7

This biochemical derangement seems

to be detrimental for reproductive outcome as elevated total homocysteine (tHcy) concentrations in

follicular fluid have been inversely associated with in vitro fertilization (IVF) / intracytoplasmic sperm

injection (ICSI) outcomes, that is, number of preantral follicles, the number of retrieved oocytes, and

embryo quality.

8-10

Thus, it seems biologically plausible that the homocysteine pathway is at least one of the intermediate

mechanisms between nutritional intake and reproductive outcome. Given that most micronutrients

are present in common food sources, such as fruit, vegetables, and cereals, we are interested in the

impact of the overall diet on reproductive outcome after fertility treatment.

Therefore, our aims were to 1) identify patterns in food consumption that explain the largest propor-

tion of variation, i.e., dietary patterns, in women of subfertile couples undergoing IVF/ICSI treatment; 2)

validate these dietary patterns with biomarkers of the homocysteine pathway in blood and follicular

fluid; and 3) determine associations between the dietary patterns in subfertile couples and IVF/ICSI

outcomes.

METHODS

Study Population

The FOod, Lifestyle and Fertility Outcome study (FOLFO) was designed to investigate the relationship

between nutrition and lifestyle and IVF/ICSI outcome.

11

In short, between September 2004 and January

2007 subfertile couples undergoing IVF/ICSI treatment at the Erasmus University Medical Centre,

Rotterdam, the Netherlands, were invited to participate. Sixty-four percent (n 161) of the couples were

included in the analysis. Of the eligible 251 couples 15 couples dropped out because of oocyte

donation, endometrioma, hydrosalpinx, medication error, or pregnancy before start of the treatment.

Couples of whom no oocytes could be retrieved were also excluded (n 20). Because lifestyle factors,

including dietary patterns, are culturally determined, we excluded couples of non-European origin (n

55) as well.

12

Ethnicity was categorized into Dutch native, European other, and non-European accord-

ing to the definitions of Statistics Netherlands (http://www.cbs.nl/). The study protocol was approved

by the Dutch Central Committee for Human Research and the Medical Ethical and Institutional Review

board of the Erasmus University Medical Centre in Rotterdam. Participants provided written informed

consent and the obtained materials and questionnaires were processed anonymously.

34 | Chapter 3

Questionnaires

All participants filled out a general questionnaire that generated the following information: age,

educational level, medical history, body mass index as kg/m

2

(BMI), ethnicity, medication use, smoking,

folic acid, and vitamin supplement use. All participants filled out a food frequency questionnaire (FFQ)

to estimate food intake of the previous 4 weeks. This FFQ was developed by the division of Human

Nutrition, Wageningen University, Wageningen, the Netherlands and validated for intake of energy, B-

vitamins, and fatty acids.

13,14

The FFQ was provided on the day of oocyte retrieval or semen sample

collection and returned on the day of embryo transfer.

In Vitro Fertilization Procedure

We used three stimulation treatments. Women were assigned either to one of two types of conven-

tional ovarian stimulation, or a mild ovarian stimulation treatment as previously described.

11

In all three

regimens a single dose of 5,000 or 10,000 IU human chorionic gonadotropin subcutaneously (hCG,

Pregnyl, NV Organon, the Netherlands) was administered to induce oocyte maturation as soon as the

leading follicle reached a diameter of at least 18 mm and at least one additional follicle reached a

diameter of 15 mm. Oocyte retrieval was performed 35 hours after hCG injection by transvaginal

ultrasound-guided puncture of the follicles. Intravaginal luteal phase supplementation of 600 mg/day

progesterone was started on the evening following oocyte pickup, and was continued for 12 days

thereafter. On day 3 after oocyte pickup a maximum of two embryos was transferred.

Follicular Fluid and Blood Sample Collection

During oocyte retrieval, follicular diameters were measured and monofollicular fluid from the largest

follicle was aspirated from each ovary and collected separately.

11

The oocytes were washed and

transferred to a separate droplet of medium to monitor embryo quality. The monofollicular fluids were

centrifuged for 10 minutes at 1,700 × g to separate red blood cells (RBC), leucocytes, and granulosa

cells. The samples were frozen without preservatives and stored at –20°C until assayed. Venous blood

samples were drawn on cycle day 2 before the first injection of recombinant FSH and on the day of

hCG administration. At both time points folate, vitamin B6, vitamin B12, tHcy, and estradiol were

determined. Baseline FSH levels were determined on cycle day 2 only.

11

We previously demonstrated

that folate concentrations increase on average to 22.5 nmol/L when women of reproductive age use

250 μg folic acid daily for 4 weeks.

16

A blood concentration above this value was therefore classified as

regular folic acid use and below as no folic acid use.

The primary endpoints of the study were the biomarker concentrations in blood and follicular fluid

and the IVF/ICSI treatment outcomes fertilization rate, embryo quality, and positive pregnancy test.

Fertilization was determined on day 1 after the IVF/ICSI procedure, and was calculated as the number

of fertilized oocytes divided by the total number of oocytes retrieved. On day 3 post-oocyte retrieval

embryo quality scores were assigned ranging from 1 (best quality) to 5 (to poorest quality without

transfer).

17

Pregnancy confirmation was assessed by a biochemical pregnancy test in urine 15 days

after oocyte retrieval.

Dietary Patterns and Chances of Pregnancy | 35

Statistics

All 195 food items from the FFQ data of all participants were classified into 22 food groups and

adjusted for total energy intake.

15,18

This was followed by principal component factor analysis (PCA)

applied on the energy-adjusted food groups of the women to construct overall dietary patterns by

explaining the largest proportion of variation in the food group intake.

19,20

The two most prevalent

dietary patterns were selected by rotating the solution with varimax method.

21

Table 1 | Characteristics of two dietary patterns in 161 women undergoing IVF/ICSI treat-

ment

food group

Health Conscious –

Low Processed diet

Mediterranean diet

alcoholic drinks −

0.15

0.13

cereals

0.13 −

0.05

butter

0.04 −

0.04

dairy

−

0.02

0.07

eggs −

0.02

0.16

fish

0.39

2

0.53

2

fruit

0.62

2

−

0.07

legumes

0.23

1

0.53

2

margarine

0.15 −

0.17

mayonnaise −

0.61

2

−

0.05

meat −

0.23

2

−

0.15

non-alcoholic drinks

0.19 −

0.11

nuts −

0.05

0.15

refined grains

0.03 −

0.10

potatoes

0.03

0.04

sauces

0.01

0.06

snacks −

0.49

2

−

0.30

2

soup

0.13

0.07

sugar

0.02

0.19

vegetable oil −

0.14

0.75

2

vegetables

0.45

2

0.57

2

whole grains

0.43

2

−

0.02

explained variance

12.1%

9.1%

Data are presented by correlation coefficients.

1

p ≤0.05,

2

p ≤0.01

Each woman was assigned two personal scores for the two respective factors, calculated as the

product of the food group value and its factor loadings summed across foods. According to their

personal score the group of 161 women was divided into tertiles and classified as low, intermediate, or

high adherence to the respective dietary pattern. The strength of adherence indicates the resem-

blance of the woman's diet compared with the respective dietary pattern identified by PCA. To

36 | Chapter 3

elucidate the relative contribution of each food group to the two dietary patterns Pearson's r correla-

tion coefficients were calculated (Table 1). The associations between lifestyle characteristics and the

identified dietary patterns are described in Table 2.

Table 2a | Characteristics of the 161 women undergoing IVF/ICSI treatment

Health Conscious – Low Processed diet

characteristic

low

(n 54 )

intermediate

(n 54 )

high

(n 53 ) p

age

1,2

34.2 (23.2 - 43.7) 35.6 (23.7 - 41.8) 36.4 (26.6 - 41.6) 0.06

BMI

1,3

24.1 (16.1 - 33.7) 23.0 (18.1 - 36.3) 22.0 (18.0 - 29.8) 0.03

high education

4

20 (37.7) 26 (48.1) 26 (49.1) 0.15

female subfertility

4

17 (31.5) 13 (24.1) 12 (22.7) 0.79

Dutch ethnicity

4

48 (88.9) 46 (85.2) 41 (77.4) 0.26

smoking

4

5 (9.3) 6 (11.1) 2 (3.8) 0.32

alcohol use

1,5

0.9 (0.0 - 17) 1.5 (0.0 - 32.2) 1.1 (0.0 - 18.3) 0.44

folic acid use

4

42 (77.8) 39 (72.2) 43 (81.1) 0.50

multivitamin use

4

24 (44.4) 29 (53.7) 25 (47.2) 0.61

energy intake

6,7

8083 (2198) 7649 (2146) 8614 (3537) 0.21

FSH

8,9

8.2 (4.2 - 18.5) 7.5 (4.1 - 30.3) 8.1 (0.4 - 14.1) 0.20

estradiol

8,10

142 (72 - 273) 132 (63 - 338) 146 (41 - 443) 0.17

duration of subfertility

1

33 (9 - 115) 40 (3 - 135) 36 (6 - 121) 0.94

fertilized oocytes

6,11

0.7 (0.3 - 1.0) 0.6 (0.1 - 1.0) 0.6 (0.2 - 1.0) 0.68

biochemical pregnancy

12

14 (34.1) 10 (26.3) 11 (26.2) 0.66

stimulation scheme

4

0.95

PO2-150

40 (75.5) 35 (68.6) 36 (70.6)

PO5-150

6 (11.3) 4 (7.8) 7 (13.7)

DLP225

7 (13.2) 12 (23.5) 8 (15.7)

fertilization by IVF

4

32 (66.7) 33 (68.8) 38 (76.0) 0.31

Data are presented by

1

median (range)

2

years,

3

kg/m

2

,

4

n (%),

5

units/week,

6

mean (sd),

7

kJoule/day,

8

cycle day 2,

9

U/L,

10

pmol/L,

11

percentage,

12

per started treatment.

Continuous variables with skewed distributions are displayed as median with range. Normally distrib-

uted variables are presented by mean and standard deviation. P values were estimated from a linear

regression model. Categorical variables are presented in frequencies with percentages and tested with

chi-square test.

Dietary Patterns and Chances of Pregnancy | 37

The associations between dietary patterns and biomarkers are shown in Table 3. Because the bio-

markers showed skewed distributions they are displayed as median with 95% confidence intervals

(95%CI). P values were estimated from a multivariable linear regression model in which logarithmically

transformed biomarkers were chosen as the dependent, whereas dietary pattern, age, BMI, and

vitamin use were included as covariates. Biomarkers in follicular fluid were adjusted for gram of protein

to eliminate potential confounding of the oocyte maturation status. Statistically significant β estimates

are given textually in the Results section. Finally we investigated whether the couple’s dietary pattern

was associated with IVF/ICSI outcome parameters (data not shown). The personal scores for the two

dietary patterns were now calculated for both the woman and man and analyzed as the couple’s

dietary pattern score by taking the average of these personal factor scores. Linear and logistic regres-

sion analyses were used in which fertilization rate, average embryo quality, and pregnancy, were

chosen as the dependent. Covariates included woman’s age, smoking, vitamin use, and type of fertility

treatment. In the final regression model BMI, alcohol use and stimulation scheme were added as

covariates. Statistical analysis was performed using SPSS 15.0 for Windows software (SPSS Inc., Chicago,

IL, US).

RESULTS

Two major dietary patterns are identified with PCA in 161 women. The first is labelled Health Con-

scious – Low Processed, containing high intakes of fruit, vegetables, whole grains, fish, and legumes,

but low intakes of mayonnaise, snacks, and meat products. The second dietary pattern, called Mediter-

ranean, comprises of high intakes of vegetable oil, fish, legumes, and vegetables but low intakes of

snacks (Table 1). The Health Conscious – Low Processed and Mediterranean diet explain proportions of

12.1% and 9.1%, respectively, of the total variation in the nutritional intake of the women.

Table 2 depicts the general characteristics for women with low, intermediate, or high adherence to

both diets. Women with high adherence to the Health Conscious – Low Processed diet show a lower

BMI compared with women with low adherence. Women with high adherence to the Mediterranean

diet are generally older, higher educated, consume more alcoholic drinks (e.g., wine), are more fre-

quently of non-Dutch origin, and undergo more often IVF treatment. An increase in RBC folate is

observed among women with a high adherence to the Health Conscious – Low Processed diet (β 0.07,

p 0.05) and Mediterranean diet (β 0.13, p ≤0.01) (Table 3). Furthermore, a high adherence to the

Mediterranean diet is positively associated with vitamin B6 in blood (β 0.09, p 0.04) and follicular fluid

(β 0.16, p 0.02). Neither the Health Conscious – Low Processed nor the Mediterranean diet are associ-

ated with fertilization rate (β 0.00, p 0.44; β 0.00, p 0.31, respectively) and embryo quality (β −0.03, p

0.95, β 0.01, p 0.35, respectively).

However, high adherence of the couple to the Mediterranean diet substantially increases the probabil-

ity of pregnancy, odds ratio (OR) 1.4 (95%CI 1.0; 1.9). This association is not present in couples with

high adherence to the Health Conscious – Low Processed diet, OR 0.8 (95%CI 0.6; 1.0). All Ors have

been adjusted for the confounders age, BMI, smoking, alcohol use, IVF/ICSI treatment, and stimulation

38 | Chapter 3

scheme. All associations are not significantly affected by the characteristics of the men, i.e., age, BMI,

smoking, and alcohol use (data not shown).

Table 2b | Characteristics of the 161 women undergoing IVF/ICSI treatment

Mediterranean diet

characteristic

low

(n 54 )

intermediate

(n 54 )

high

(n 53 ) p

age

1,2

35.2 (23.2 - 43.7 ) 33.9 (23.7 - 40.6) 37.2 (29.3 - 42.1) ≤0.01

BMI

1,3

23.5 (18.5 - 36.3) 22.8 (18.1 - 34.4) 22.3 (16.1 - 29.8) 0.34

high education

4

18 (33.3) 21 (39.6) 33 (62.3) 0.03

female subfertility

4

17 (31.5) 9 (16.7) 16 (30.2) 0.07

Dutch ethnicity

4

47 (87.0) 49 (90.7) 39 (73.6) 0.04

smoking

4

4 (7.4) 4 (7.4) 5 (9.6) 0.89

alcohol use

1,5

0.3 (0.0 - 13.4) 0.7 (0.0 - 32.2) 2.1 (0.0 - 18.3) 0.02

folic acid use

4

38 (70.4) 44 (81.5) 42 (79.2) 0.33

multivitamin use

4

23 (42.6) 26 (48.1) 29 (54.7) 0.45

energy intake

6,7

8040 (1922) 8087 (2234) 8212 (3707) 0.86

FSH

8,9

8.1 (4.5 - 30.3) 7.6 (2.4 - 18.4) 8.6 (0.4 - 18.5) 0.53

estradiol

8,10

139 (49 - 338) 142 (68 - 307) 140 (41 - 443) 0.78

duration of subfertility

1

42 (3 - 135) 36 (10 - 121) 34 (6 - 115) 0.52

fertilized oocytes

6,11

0.6 (0.1 - 1.0) 0.7 (0.2 - 1.0) 0.6 (0.2 - 1.0) 0.22

biochemical pregnancy

12

11 (25) 13 (32.5) 11 (29.7) 0.75

stimulation scheme

4

0.09

PO2-150

34 (63) 36 (73.5) 41 (78.8)

PO5-150

8 (14.8) 3 (6.1) 6 (11.5)

DLP225

12 (22.2) 10 (20.4) 5 (9.6)

fertilization by IVF

4

30 (61.2) 28 (60.9) 45 (88.2) ≤0.01

Data are presented by

1

median (range)

2

years,

3

kg/m2,

4

n (%),

5

units/week,

6

mean (sd),

7

kJoule/day,

8

cycle day 2,

9

U/L,

10

pmol/L,

11

percentage,

12

per started treatment.

DISCUSSION

This study demonstrates that Dutch subfertile couples with high adherence to the Mediterranean

dietary pattern have a 40% increased probability of achieving pregnancy after IVF/ICSI treatment. The

adherence to this diet is reflected by relatively high concentrations of folate and vitamin B6 in blood

and follicular fluid. The Health Conscious – Low Processed dietary pattern was positively associated

with RBC folate, but did not affect IVF/ICSI outcomes.

Dietary Patterns and Chances of Pregnancy | 39

Labelling the extracted factors as the Health Conscious – Low Processed and Mediterranean dietary

pattern is part of a continuing effort to achieve consistency in the use of dietary patterns in medical

research. Northstone et al.

22

recently investigated dietary patterns during pregnancy in a large cohort

study and identified a Health Conscious, Traditional, and Processed dietary pattern. Our first factor

consists of high intakes of fruit, vegetables, fish, and whole grains and low intakes of snacks, meat, and

mayonnaise, and resembles a combination of Northstones Health Conscious and Processed diet, and

was therefore labelled as Health Conscious – Low Processed.

Table 3a | Biochemical determinants of the homocysteine pathway in blood in women

undergoing IVF/ICSI treatment

1

Health Conscious – Low Processed diet

biomarker

low

(n 32)

intermediate

(n 25)

high

(n 28)

p

9

folate

2,3

1553 (488 - 2461) 1149 (531 - 3116) 1672 (783 - 3611) 0.05

folate

2,4

32.1 (9.8 - 82.4) 38.0 (11.7 - 82.3) 37.0 (15.8 - 119.5) 0.09

vitamin B6

2,5

78.5 (43 - 231) 75 (51 - 173) 81 (46 - 250) 0.33

vitamin B12

6,7

316 (74 - 1856) 325 (140 - 724) 302 (143 - 863) 0.55

tHcy

8

9.7 (6.1 - 75.3) 9.2 (5.5 – 37.0) 8.4 (5.8 - 12.6) 0.72

Mediterranean diet

biomarker

low

(n 27)

intermediate

(n 27)

high

(n 31)

p

9

folate

2,3

1078 (531 - 3293) 1462 (488 - 2244) 1665 (909 - 3611) ≤0.01

folate

2,4

32.6 (11.7 - 119.5) 33.9 (9.8 - 108.5) 37.3 (17.1 - 92.0) 0.28

vitamin B6

2,5

72 (50 - 231) 79 (43 - 173) 89 (43 - 250) 0.04

vitamin B12

6,7

297 (74 - 724) 333 (183 - 688) 353 (140 - 1856) 0.14

tHcy

8

10.1 (6.1 - 75.3) 8.9 (5.5 - 18.3) 8.7 (5.8 - 12.9) 0.48

Data are presented by

1

median with range,

2

nmol/L,

3

red blood cell,

4

serum,

5

whole blood,

6

pmol/L,

7

plasma,

8

μmol/L.

9

Adjusted for age, BMI, and use of vitamin supplements.

The second factor shares many features with the classical Mediterranean diet,

23

defined by high intakes

of vegetable oils, vegetables, fruit, nuts, fish, and legumes, low dairy intake, and moderate intake of

alcohol. We examined the magnitude of comparability by scoring the factor to the 10-point Mediter-

ranean diet scale as introduced by Trichopoulou et al.

23

Analysis revealed that women with high

adherence to the second factor scored positively in 7 of 10 dietary criteria. This indicates good compa-

rability with the traditional definition of Mediterranean, justifying the labelling as the Mediterranean

diet.

The two dietary patterns in our study show a remarkable overlap in foods, for example, high intakes of

vegetables, fish, and legumes and low intakes of snacks. However, only the Mediterranean diet seems

40 | Chapter 3

to increase the chance of pregnancy after IVF/ICSI treatment. There are two major differences between

these two diets that may explain this finding.

First, when comparing differences in food intakes, the high intake of vegetable oils in the Mediterra-

nean diet is outstanding. Vegetable oils are generally rich in linoleic acid and belong to the family of

omega-6 fatty acid molecules, which can only be obtained by the diet. They are precursors of different

prostaglandins, important for the initiation of the menstrual cycle, growth, and development of

preantral follicles and ovulation. Prostaglandins are also involved in the maintenance of pregnancy by

optimizing endometrial receptivity.

24-27

This may imply that a higher intake of linoleic acid perhaps

positively affects the implantation of the fertilized ovum.

Table 3b | Biochemical determinants of the homocysteine pathway in follicular fluid

in women undergoing IVF/ICSI treatment

1

Health Conscious – Low Processed diet

biomarker

low

(n 36 )

intermediate

(n 32)

high

(n 39)

p

8

folate

2,3

29.9 (7.3 - 84) 32.7 (8.9 - 200) 35.5 (12.5 - 190) 0.23

vitamin B6

2,4

68.0 (17.5 - 310) 86.4 (14.5 - 310) 84 (22.5 - 310) 0.97

vitamin B12

5,6

193 (59 - 3695) 247 (123 - 31013) 209 (88 - 1673) 0.59

tHcy

6,7

6.3 (3.8 - 70.2) 6.6 (3.3 - 27.2) 6.1 (3.6 - 14) 0.72

Mediterranean diet

biomarker

low

(n 34)

intermediate

(n 37)

high

(n 36)

p

8

folate

2,3

30.5 (8.9 - 199.6) 32.3 (7.3 - 69) 35.3 (12.0 - 190) 0.14

vitamin B6

2,4

70.5 (14.5 - 310) 74.0 (17.5 - 310) 91.5 (22.5 - 310) 0.02

vitamin B12

5,6

206 (59 - 31013) 196 (104 - 4199) 230 (88 - 2959) 0.89

tHcy

6,7

6.3 (3.3 - 70.2) 6.4 (3.8 - 15.5) 6.2 (3.6 – 14.0) 0.20

Data are presented by

1

median with range,

2

nmol/L,

3

serum,

4

whole blood,

5

pmol/L,

6

plasma,

7

μmol/L.

8

Adjusted

for age, BMI, use of vitamin supplements, and total protein in follicular fluid.

The second difference between the two dietary patterns is found in their effect on the biomarkers of

the homocysteine pathway. Both diets increase folate concentrations, but the Mediterranean diet

shows an additional rise in vitamin B6 in blood and follicular fluid. Vitamin B6 is a versatile coenzyme

involved in many biochemical pathways. Research has shown that giving vitamin B6 to subfertile

women increases reproductive performance, that is, a 40% increased chance of conception and a 30%

lower risk of miscarriage early in pregnancy.

28

The positive association between vitamin B6 and the

success of IVF/ICSI treatment is herewith in line.

Dietary Patterns and Chances of Pregnancy | 41

Thus, this study has provided novel insights in the relationship between dietary patterns and IVF/ICSI

outcomes, in which vitamin B6 and fatty acids may play an important role. Further research, however,

is needed to validate the current findings and investigate optimal dosage effects before any dietary

preparation can be generally recommended.

Self-reported dietary assessment methods are susceptible to measurement errors, where habitual

energy intake tends to be underestimated.

29

However, we consider it unlikely that these measurement

errors would have produced unvalid associations. Because of the prospective study character, expo-

sure determinants were measured before the outcomes were known. Therefore, energy and nutri-

tional intake are unlikely to be differentially underreported or differently reported between couples

with and without successful IVF/ICSI treatment outcome, e.g. ongoing pregnancy. The findings from

this Dutch subfertile study group cannot be equally generalized to the overall reproductive population

because couples seeking fertility treatment generally have a higher age and education. These results

should also not be extrapolated to non-Europeans for which studies in different ethnic populations are

required.

General Conclusion

A high adherence to the Mediterranean dietary pattern by the couple may improve the chance of

pregnancy after IVF/ICSI treatment. These findings are important with regard to the development of

nutritional interventions to further improve fertility treatment and success rates.

REFERENCES

1 Baird DT, Collins J, Egozcue J, et al. Fertility and ageing. Hum Reprod Update 2005;11:261-76.

2 Homan GF, Davies M, Norman R. The impact of lifestyle factors on reproductive performance

in the general population and those undergoing infertility treatment: a review. Hum Reprod

Update 2007;13:209-23.

3 Ebisch IM, Thomas CM, Peters WH, Braat DD, Steegers-Theunissen RP. The importance of

folate, zinc and antioxidants in the pathogenesis and prevention of subfertility. Hum Reprod

Update 2007;13:163-74.

4 Scholl TO, Johnson WG. Folic acid: influence on the outcome of pregnancy. The Am J Clin

Nutr2000;71:S1295-303.

5 Thaver D, Saeed MA, Bhutta ZA. Pyridoxine (vitamin B6) supplementation in pregnancy.

Cochrane Database Syst Rev 2006:CD000179.

6 Refsum H. Folate, vitamin B12 and homocysteine in relation to birth defects and pregnancy

outcome. Br J Nutr 2001;85:S109-13.

7 Timmermans S, Jaddoe VW, Hofman A, Steegers-Theunissen RP, Steegers EA. Periconception

folic acid supplementation, fetal growth and the risks of low birth weight and preterm birth:

the Generation R Study. Br J Nutr 2009;102:777-85.

8 Forges T, Monnier-Barbarino P, Alberto JM, Gueant-Rodriguez RM, Daval JL, Gueant JL. Impact

of folate and homocysteine metabolism on human reproductive health. Hum Reprod Update

2007;13:225-38.

9 Ebisch IM, Peters WH, Thomas CM, Wetzels AM, Peer PG, Steegers-Theunissen RP. Homocys-

teine, glutathione and related thiols affect fertility parameters in the (sub)fertile couple. Hum

Reprod 2006;21:1725-33.

42 | Chapter 3

10 Szymanski W, Kazdepka-Zieminska A. [Effect of homocysteine concentration in follicular fluid

on a degree of oocyte maturity]. Ginekol polska 2003;74:1392-6.

11 Boxmeer JC, Macklon NS, Lindemans J, et al. IVF outcomes are associated with biomarkers of

the homocysteine pathway in monofollicular fluid. Hum Reprod 2009;24:1059-66.

12 Cassidy CM. Walk a mile in my shoes: culturally sensitive food-habit research. The Am J Clin

Nutr 1994;59:S190-7.

13 Feunekes GI, Van Staveren WA, De Vries JH, Burema J, Hautvast JG. Relative and biomarker-

based validity of a food-frequency questionnaire estimating intake of fats and cholesterol. Am

J Clin Nutr 1993;58:489-96.

14 Verkleij-Hagoort AC, de Vries JH, Ursem NT, de Jonge R, Hop WC, Steegers-Theunissen RP.

Dietary intake of B-vitamins in mothers born a child with a congenital heart defect. Eur J Nutr

2006;45:478-86.

15 Brouwer IA, van Rooij IA, van Dusseldorp M, et al. Homocysteine-lowering effect of 500 microg

folic acid every other day versus 250 microg/day. Ann Nutr Metabol 2000;44:194-7.

16 Huisman GJ, Fauser BC, Eijkemans MJ, Pieters MH. Implantation rates after in vitro fertilization

and transfer of a maximum of two embryos that have undergone three to five days of culture.

Fertil Steril 2000;73:117-22.

17 Slimani N, Fahey M, Welch AA, et al. Diversity of dietary patterns observed in the European

Prospective Investigation into Cancer and Nutrition (EPIC) project. Public Health Nutr

2002;5:1311-28.

18 Willett WC, Howe GR, Kushi LH. Adjustment for total energy intake in epidemiologic studies.

Am J Clin Nutr 1997;65:S1220-31.

19 Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Op Lipidol

2002;13:3-9.

20 Vujkovic M, Ocke MC, van der Spek PJ, Yazdanpanah N, Steegers EA, Steegers-Theunissen RP.

Maternal Western dietary patterns and the risk of developing a cleft lip with or without a cleft

palate. Obstet Gynecol 2007;110:378-84.

21 Kaiser HF. The Varimax Criterion for Analytic Rotation in Factor Analysis. Psychometrika

1958;23:187-200.

22 Northstone K, Emmett P, Rogers I. Dietary patterns in pregnancy and associations with socio-

demographic and lifestyle factors. Eur J Clin Nutr 2008;62:471-9.

23 Trichopoulou A, Costacou T, Bamia C, Trichopoulos D. Adherence to a Mediterranean diet and

survival in a Greek population. N Engl J Med 2003;348:2599-608.

24 Abayasekara DR, Wathes DC. Effects of altering dietary fatty acid composition on prostaglandin

synthesis and fertility. Prostaglandins Leukot Essent Fatty Acids 1999;61:275-87.

25 Achache H, Revel A. Endometrial receptivity markers, the journey to successful embryo

implantation. Hum Reprod 2006;12:731-46.

26 Sorbera LA, Asturiano JF, Carrillo M, Zanuy S. Effects of polyunsaturated fatty acids and pros-

taglandins on oocyte maturation in a marine teleost, the European sea bass (Dicentrarchus

labrax). Biol Reprod 2001;64:382-9.

27 Staples CR, Burke JM, Thatcher WW. Influence of supplemental fats on reproductive tissues

and performance of lactating cows. J Dairy Sc 1998;81:856-71.

28 Ronnenberg AG, Venners SA, Xu X, et al. Preconception B-vitamin and homocysteine status,

conception, and early pregnancy loss. Am J Epidemiol 2007;166:304-12.

29 Black AE, Goldberg GR, Jebb SA, Livingstone MB, Cole TJ, Prentice AM. Critical evaluation of

energy intake data using fundamental principles of energy physiology: 2. Evaluating the

results of published surveys. Eur J Clin Nutr 1991;45:583-99.

Increased preconception omega

-

3

polyunsaturated fatty acid intake improves

embryo morphology

Fatima Hammiche, Marijana Vujković, Willeke Wijburg, Jeanne H. de Vries, Joop S. Laven,

and Régine P. Steegers-Theunissen

submitted for publication

CHAPTER

4

46 | Chapter 4

ABSTRACT

Objective

This study investigates associations between preconception dietary intake of omega-6 and omega-3

poly-unsaturated fatty acids (LC-PUFAs) on estradiol levels and IVF/ICSI outcome.

Methods

An observational prospective study was set up in a tertiary referral fertility clinic at the Erasmus Univer-

sity Medical Centre, Rotterdam, the Netherlands. Two-hundred-thirty-five women undergoing IVF/ICSI

treatment were included. Main outcome measures consisted of estradiol in blood, number of follicles

and embryo morphology.

Results

Estradiol on cycle day 2 was positively associated with a high intake of total omega-3 LC-PUFA (β 68.5,

se 34.8, p 0.05), in particular ALA (β 90.4, se 35.7, p ≤0.01). A lower estradiol response on the hCG day

was observed in the groups with the highest EPA (β -1062, se 492, p 0.03) and DHA (β -1006, se 485, p

0.04) intakes. The number of follicles was inversely associated with high intakes of EPA (β -1.75, se 0.87,

p 0.05) and DHA (β -1.78, se 0.85, p 0.04). Positive associations were established between embryo

morphology and total omega-3 (linear β 0.63, se 0.26, p 0.02), ALA (β 0.56, se 0.26, p 0.03) and DHA (β

0.17, se 0.09, p 0.05) LC-PUFAs intakes. Estradiol and fertility outcome parameters were not associated

with omega-6 LA intake.

Conclusion

Omega-3 LC-PUFA intake in women undergoing IVF/ICSI treatment is associated with improved

embryo morphology.

Omega-3 Fatty Acids and Embryo Morphology | 47

INTRODUCTION

Dietary intake of long chain poly-unsaturated fatty acids (LC-PUFAs) are beneficial in the prevention of

cardiovascular disorders.

1,2

The role of LC-PUFAs in human fertility has received little attention thus far.

3

Several animal studies, however, reported that dietary fats influence oocyte maturation, corpus luteum

function and embryo development.

4-6

LC-PUFAs are essential of cell membranes and after activation by hormones and growth factors they

become precursors of eicosanoids, such as prostaglandins, leukotrienes and tromboxanes, which are

important mediators in inflammatory, trombogenic and vascular mechanisms.

3,4,7

Based on their

chemical structure we distinguish omega-3 and omega-6 LC-PUFAs. Omega-3 LC-PUFAs comprise

alpha-linolenic acid (ALA), eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). ALA –

present in green vegetables – can be converted in EPA and DHA.

7,8

However, this conversion is

insufficient to meet daily EPA and DHA needs.

Therefore, the intake of fish as rich dietary source of these omega-3 LC-PUFAs is recommended. Its

consumption, however, is rather low in Western countries and results in an increased ratio of omega-6

to omega-3 LC-PUFAs (10:1).

4

The most important omega-6 PUFA is linoleic acid (LA) serving as

precursor of arachidonic acid (AA) and present in nearly all vegetable oils, while substantial amounts of

AA are present in meat and eggs.

8

The effects of individual and various amounts of omega-3 and

omega-6 LC-PUFAs on human reproduction, however, are limited.

9

Therefore, the aim of this study was to investigate associations between the periconception maternal

dietary intake of omega-3 and omega-6 LC-PUFAs on estradiol levels and reproductive outcome

parameters in a periconception prospective observational study of women undergoing in vitro

fertilization (IVF) or intracytoplasmatic injection treatment (ICSI).

METHODS

Study Population

The Food Lifestyle and Fertility Outcome Study (FOLFO Study) is a prospective preconception observa-

tional study which focuses on the influence of nutrition and lifestyle on fertility and pregnancy out-

come. The design of the study has been described previously.

10

In summary, between September 2004

and January 2007 subfertile couples undergoing IVF/ICSI treatment at the Erasmus University Medical

Centre, Rotterdam, the Netherlands were invited to participate. Of the eligible IVF/ICSI population, 66%

of the couples participated in the FOLFO study (n 251). We excluded couples who suffered from

known conditions that may influence IVF/ICSI treatment outcome, such as oocyte donation, endome-

triosis and hydrosalpinx resulting in 235 women for this study.

The study protocol was approved by the Central Committee for Human Research in The Hague, the

Netherlands and the Medical Ethical and Institutional Review Board of the Erasmus University Medical

48 | Chapter 4

Centre in Rotterdam, the Netherlands. All participants gave their written informed consent and all

obtained materials and questionnaires were processed anonymously.

General Questionnaire

All participants filled out a general questionnaire from which the following data were extracted:

height, weight, ethnicity, education level, vitamin use, and other lifestyle factors. Ethnicity and educa-

tion level were classified according to the definitions of Statistics Netherlands.

11

Education level was

divided into three categories: low (primary, lower vocational, or intermediate secondary), intermediate

(intermediate vocational or higher secondary) and high (higher vocational, or university). Ethnicity was

divided into Dutch Native, European other and Non-European.

11

Food Frequency Questionnaire

All participants filled out a food frequency questionnaire (FFQ) to estimate habitual food intake over

the previous four weeks. This FFQ was originally developed at the division of Human Nutrition, Wagen-

ingen University, the Netherlands and validated for intake of energy, B-vitamins and fats.

12,13

The FFQ

was provided to the subfertile woman on the day of oocyte retrieval and was returned on the day of

embryo transfer. The researcher verified the completeness of the FFQ. In case of missing or unclear

information about type and amount of foods consumed, additional questions were asked by tele-

phone. The intake of energy and fatty acids were compared with the Dietary Reference Intake (DRI) for

the Netherlands.

14

To evaluate the existence of underreporting the ratio of total energy intake to basal

metabolic rate (BMR) was calculated using the new Oxford equation for women aged 30-60 years: BMR

(mJoule/day) = 0.0407 x weight (in kilogram) + 2.90.

15

This value is an estimation of the physical activity

level (PAL) of a sedentary lifestyle. The physical activity level was then calculated by dividing the mean

reported energy intake by the mean BMR. According to Goldberg et al. (1991) a cut-off point for

underreporting for a sedentary lifestyle is a ratio of ≤1.35.

16

In Vitro Fertilization Procedure

In our study population, three IVF stimulation treatments were used. In one group women started

ovarian stimulation with daily injections of 150 IU recombinant Follicle Stimulating Hormone (rFSH)

subcutaneous on cycle day 2 (Puregon, Schering Plough, Houten, the Netherlands or Gonal-F, Merck

Serono Benelux BV, Schiphol-Rijk, the Netherlands).

Administration of daily subcutaneous Gonadotropin Releasing Hormone (GnRH) antagonist (Or-

galutran, NV Schering Plough, or Cetrotide, Merck Serono Benelux BV) was started when at least one

follicle was ≥14 mm. In another treatment group women were randomized for either conventional

ovarian stimulation or mild ovarian stimulation. Patients assigned to the conventional ovarian stimula-

tion started the GnRH agonist 0.1 mg/day subcutaneous on cycle day 21 of the menstrual cycle. After

two weeks of the GnRH regimen, co-treatment with rFSH 225 IU/day subcutaneous was started.

Patients assigned to the mild ovarian stimulation were treated with a fixed dose of 150 IU/day rFSH

subcutaneous started on cycle day 5. As soon as the leading follicle reached a diameter of 14 mm the

GnRH-antagonist of 0.25 mg/day subcutaneous was added to the regimen.

Omega-3 Fatty Acids and Embryo Morphology | 49

To induce final oocyte maturation a single dose of 10.000 IE human chorionic gonadotropin (hCG)

subcutaneous was administered in all three regimens as soon as the leading follicle reached a diame-

ter of at least 18 mm and at least one additional follicle reached a diameter of 15 mm.

Table 1 | General characteristics of the study population

characteristic

participants

(n 235)

age

1,2

35.0 ± 4.2

BMI

1,3

23.7 ± 3.7

Dutch ethnicity

4

164 (70.1)

high education

4

103 (44.0)

smoking

4

21 (8.9)

medication

4

36 (15.3)

alcohol

4

215 (91.5)

folic acid use

4

207 (88.5)

cause of subfertility

4

female

51(21.7)

male

86 (36.6)

male and female

15 (6.4)

idiopathic

83 (35.3)

fertilization by IVF

4

146 (62.1)

stimulation scheme

4

P02-150

174 (76.7)

P05-150

32 (14.1)

DLP-225

21 (9.1)

estradiol

5,6,7

138.5 (41.0 - 2051.0)

estradiol

5,6,8

2484 (233 - 20018)

Data are presented by

1

mean ± sd,

2

years,

3

kg/m

2

,

4

n (%),

5

pmol/L,

6

median (min-max),

7

at baseline,

8

after stimula-

tion.

Oocytes were retrieved 35 hours after hCG administration by transvaginal ultrasound-guided puncture

of follicles. Prior to ovum pick up the total number of follicles were counted by transvaginal ultra-

sound. Seventy two hours after oocyte retrieval embryo morphology scores were assigned according

to previously described criteria.

17

These

scores ranged from one (best quality) to five (poor quality

and

50 | Chapter 4

inadequate for embryo transfer). Embryos with an assigned score of 1 or 2 are classically denoted as

perfect and respectively good embryos.

Estradiol Assessment

Venous blood samples were drawn from each woman on cycle day 2 before the first injection of rFSH

and at the day of hCG administration for the determination of estradiol. Venous blood samples were

drawn into dry vacutainer tubes and allowed to clot. After centrifugation at 2000 g, serum was col-

lected before being assayed. Estradiol was determined by using coated radioimmunoassay (RIA,

Diagnostic Products Corporation). Intra- and inter-assay coefficients of variation were 5% and 7% for

estradiol, respectively.

Statistics

Dietary intakes of the LC-PUFAs were log-transformed to obtain normal distributions. The log-

transformed intakes were adjusted for energy intake according to the residual method of Willet et al.

18

Multivariable linear regression models were used to assess the relationship between the logarithmi-

cally transformed LC-PUFA intakes and the estradiol levels on cycle day 2 and hCG day, number of

follicles and embryo morphology. In the linear regression model the potential confounders considered

comprised ethnicity, age, BMI, smoking, alcohol use, folic acid supplement use, IVF/ICSI treatment, and

ovarian stimulation regimen (n 3). Using a stepwise backward regression model covariates were

included when they reached significance defined as a p ≤0.1. The associations between LC-PUFAs with

estradiol level on cycle day 2 and number of follicles are adjusted for all covariates. Estradiol on hCG

day, i.e., estradiol response after ovarian stimulation treatment, was also adjusted for estradiol on cycle

day 2. The LC-PUFA associations with embryo morphology are adjusted for ethnicity, BMI, age, IVF/ICSI

treatment, and ovarian stimulation treatment. All data in the table are presented as standardized

adjusted regression coefficients (ß) with standard error (se) both linear and after dichotomization of

the LC-PUFA intake into 15

th

(p15) and 85

th

(p85) percentiles, representing respectively low and high

intake. The latter was conducted to evaluate possible threshold values for the effects on the fertility

outcome parameters. A p ≤0.05 was considered statistically significant. All statistical analyses were

performed using SPSS 15.0 for Windows (SPSS Inc, Chicago, II, US).

RESULTS

Data from 235 subfertile women were evaluated and the general characteristics are shown in Table 1.

Women had a mean age of 35.0 years (sd 4.2), a BMI of 23.7 kg/m

2

(sd 3.7). Furthermore, the majority

was of Dutch origin (70.4%), highly educated (44.0%), consumed alcoholic drinks on a frequent basis

(91.5%), and used folic acid and/or a folic acid containing multivitamin supplement (88.5%). Only 8.9%

of the women smoked. In Table 2 total energy, macronutrient and LC-PUFA intakes are depicted and

compared with the DRI for women between 19-40 years of age as reference.

19

The average total energy was slightly lower, fat, protein and carbohydrate intake were higher than the

recommendations. The omega-3 LC-PUFA intake – in particular of EPA and DHA – was lower whereas

the omega-6 LA LC-PUFA intake was higher than the recommendations for women in the age group

Omega-3 Fatty Acids and Embryo Morphology | 51

of 19-40 years and pregnant women. The omega-6/omega-3 ratio is 12.1/1.14; this is higher than

recommended. To evaluate whether the relatively low omega-3 intake were due to the general

underreporting of food intake, we calculated the PAL measure.

Table 2 | Nutrient Intakes

nutrient

unit participants DRI

3

energy intake

kJoule/day 7861 (1999 - 29814) 8100

total fat

g/day 70.0 (13.4 - 28.7) 50

adjusted 81.2 (30.5 - 124.7)

total protein

g/day 70.8 (24.2 - 175.3) 50-52

adjusted 74.7 (43.7 - 100.7)

total carbohydrates

g/day 223 (59 - 980) 270

adjusted 247 (136 - 386)

LA

g/day 12.1 (1.5 - 49.8) 12.0

adjusted 13.6 (5.6 - 33.5)

ALA

g/day 0.98 (0.18 - 6.61) 1.1

adjusted 1.06 (0.47 - 5.37)

EPA

g/day 0.04 (0.00 - 0.38)

adjusted 0.05 (0.00 - 0.39)

EPA + DHA > 0.4

DHA

g/day 0.07 (0.00 - 0.51)

adjusted 0.08 (0.00 - 0.52)

EPA + DHA > 0.4

total omega-3

1

g/day 1.14 (0.28 - 7.42 )

adjusted 1.26 (0.49 - 1.63)

omega-6: omega-3 ratio

2

10.1 (3.5 - 20.7)

Data are presented by median (min-max).

1

EPA + DHA + ALA.

2

LA / (ALA + EPA + DHA).

3

Dietary Reference Intakes:

energy, The Hague: Health Council of the Netherlands 2001/19R (corrected edition June 2002)

It revealed that the PAL was 1.44 above the cut-off level of 1.35; therefore underreporting is not very

likely. Table 3 shows associations, linear and after dichotomization in high (>p85) and low (<p15) LC-

PUFA intakes in relation to baseline and response estradiol levels and reproductive outcome parame-

ters. Baseline estradiol on cycle day 2 was positively associated with high intakes of omega-3 LC-PUFA

(β 68.5, se 34.8, p 0.05). An inverse estradiol response on the day of hCG administration was associated

with highest intakes of EPA (β -1062, se 492, p 0.03) and DHA (-1006, se 485, p 0.04) also substantiated

with a linear association of DHA (β -395, se 200, p 0.03).

52 | Chapter 4

The number of follicles was inversely associated with high intakes of EPA (β -1.75, se 0.87, p 0.05) and

DHA (-1.78, se 0.85, p 0.04). Positive associations were established between embryo morphology and

intakes of omega-3 LC-PUFA (linear β 0.63, se 0.26, p 0.02), in particular ALA (β 0.56, se 0.26, p 0.03) and

DHA (β 0.17, se 0.09, p 0.05). Estradiol and fertility outcome parameters were not associated with

omega-6 LA intake and the omega-6:omega-3 ratio.

Table 3a | Associations between omega LC-PUFA and fertility outcome

nutrient

estradiol

cycle day 2

1

estradiol

hCG Day

2

no. of

follicles

1

embryo

morphology

3

LA

(n 188) (n 178) (n 194) (n 175)

linear

β (se) –49.2 (39.9) –647 (665) –0.71 (1.14) 0.48 (0.28)

p 0.22 0.33 0.54 0.09

>p85 (17.9 g)

β (se) 0.9 (38.1) –243 (620) –0.51 (1.08) 0.48 (0.3)

p 0.98 0.70 0.64 0.11

<p15 (10.6 g)

β (se) –11.2 (33.1) 500 (561) –0.18 (0.95) –0.17 (0.23)

p 0.74 0.37 0.85 0.47

ALA

linear

β (se) –2.7 (35.3) –647 (578) –0.45 (1.02) 0.56 (0.26)

p 0.94 0.27 0.66 0.03

>p85 (1.5 g)

β (se) 90.4 (35.7) –637 (609) –0.33 (1.05) 0.07 (0.27)

p 0.01 0.30 0.76 0.79

<p15 (0.8 g)

β (se) –11.3 (38.1) 5.79 (633) –0.01 (1.1) –0.39 (0.26)

p 0.77 0.99 1.00 0.13

EPA

linear

β (se) 9.1 (11.3) –294 (181) –0.28 (0.32) 0.1 (0.08)

p 0.42 0.11 0.38 0.21

>p85 (0.1 g)

β (se) –19.7 (30.4) –1062 (492) –1.75 (0.87) 0.13 (0.22)

p 0.52 0.03 0.05 0.54

<p15 (0.01 g)

β (se) –27.4 (39.4) 746 (635) –0.2 (1.09) –0.19 (0.27)

p 0.49 0.24 0.85 0.50

Adjusted for

1

ethnicity, age, BMI, smoking, alcohol, stimulation scheme, folic acid, IVF/ICSI treatment;

2

estradiol on

cycle day 2, ethnicity, age, BMI, smoking, alcohol, stimulation scheme, folic acid, IVF/ICSI treatment;

3

age, BMI,

ethnicity, IVF/ICSI, stimulation scheme.

Omega-3 Fatty Acids and Embryo Morphology | 53

DISCUSSION

To our knowledge this is the first study to evaluate omega-3 and omega-6 LC-PUFA intakes in associa-

tion with estradiol status and response, number of follicles, and embryo morphology in women

undergoing IVF/ICSI treatment. We demonstrate that in these women the dietary intake of omega-3

LC-PUFAs – in particular of EPA and DHA – is much lower than the dietary recommended intakes, in

contrast to the adequate intakes of omega-6 LC-PUFA. Women with the highest intake of the omega-3

LC-PUFA ALA showed a higher baseline estradiol level, and in particular the high intakes of EPA and

DHA reduced the estradiol response and number of follicles after ovarian stimulation treatment. This is

in line with the improved embryo morphology by high intakes of omega-3 LC-PUFA, in particular ALA

and DHA.

The high intake of the omega-3 LC-PUFA ALA was also associated with a higher baseline estradiol

level. The importance of the baseline estradiol level for reproductive outcomes is controversial.

20

Omega-3 LC-PUFAs are an essential source for the synthesis of cholesterol, which acts as precursor of

estradiol and other steroids. This is a possible pathway in which a high ALA intake increases follicular

steroid synthesis.

4

Other studies support this finding by showing that trans fatty acids, mono-

unsaturated fat and poly-unsaturated fat intake influence the levels of estradiol as well.

21,22

Two studies

investigated potential associations between maternal omega-3 and omega-6 LC-PUFAs on estradiol

levels during pregnancy, however, without significant results.

23,24

These negative results are suggested

to be due to the type of fatty acid intake.

In this study, we also showed that a high intake of EPA and/or DHA reduced the estradiol response

and number of follicles after ovarian stimulation treatment. In an animal study it has been shown that

consumption of high levels of omega-3 LC-PUFAs resulted in elevated ova release, whereas consump-

tion of moderate levels had no effect on ova release in rats.

25

However, in this study fish oil was used,

which included different omega-3 LC-PUFAs, therefore the enhancing effect couldn’t be attributable

to a specific omega-3 LC-PUFAs or their combination, as well as the dietary level. Our findings suggest

a beneficial effect of omega-3 LC-PUFA intake on fertility outcomes, since a more physiological

approach to ovarian stimulation, resulting in fewer dominant follicles, may allow only the healthiest

follicles and oocytes to develop in competent embryos.

26

In addition, the existence of an estradiol

window with an upper threshold at the time of hCG administration has been suggested.

20,27

An

elevation above this threshold could be deleterious for embryonic implantation. It has been shown

that uterine receptivity is affected in patients undergoing ovulation induction with high serum

estradiol concentrations on the day of hCG administration, regardless of the number of oocytes

retrieved and the progesterone concentration. It would be interesting to study the associations with

number of implantations and pregnancies as well. However, due to the limited power this was not the

aim of the current study.

Moreover, our finding is consistent with several studies in rats fed a diet high in EPA and DHA showing

a decrease in the frequency of ovulations.

28-31

EPA and DHA have been reported to elicit a reduction of

ovarian synthesis of prostaglandin F

2α,

which is involved in follicle growth and ovulation and therefore

54 | Chapter 4

may partly explain the inverse effect on number of follicles.

32-34

The mechanism, by which EPA and

DHA inhibit secretion of prostaglandin F

2α

is however not fully understood.

Table 3b | Associations between omega LC-PUFA and fertility outcome

nutrient

estradiol

cycle day 2

1

estradiol

hCG Day

2

no. of

follicles

1

embryo

morphology

3

DHA

(n 188) (n 178) (n 194) (n 175)

linear

β (se) 8.8 (12.1) –395 (200) –0.42 (0.35) 0.17 (0.09)

p 0.47 0.05 0.22 0.05

>p85 (0.2 g)

β (se) –21.1 (29.9) –1006 (485) –1.78 (0.85) 0.13 (0.22)

p 0.48 0.04 0.04 0.53

<p15 (0.02 g)

β (se) –27.4 (35.6) 1177 (582) 0.82 (0.99) –0.42 (0.26)

p 0.44 0.04 0.41 0.11

total omega-3

linear

β (se) 4.9 (35.4) –871 (576) –0.65 (1.02) 0.63 (0.26)

p 0.89 0.13 0.52 0.02

>p85 (1.7 g)

β (se) 68.5 (34.8) –847 (572) –1.31 (1.02) 0.36 (0.26)

p 0.05 0.14 0.20 0.17

<p15 (1.0 g)

β (se) –23.5 (31.8) 34.6 (525) –0.56 (0.92) –0.41 (0.23)

p 0.46 0.95 0.54 0.08

O6:O3 ratio

linear

β (se) –29.7 (42.4) 562.8 (698) 0.15 (1.23) –0.38 (0.31)

p 0.48 0.42 0.91 0.22

>p85 (16.2)

β (se) –39.7 (39.3) 596 (641) 1.49 (1.14) –0.43 (0.29)

p 0.31 0.35 0.19 0.14

<p15 (7.3)

β (se) –30.5 (32.9) –609 (553) 0.07 (0.96) 0.20 (0.24)

p 0.36 0.27 0.94 0.42

Adjusted for

1

ethnicity, age, BMI, smoking, alcohol, stimulation scheme, folic acid, IVF/ICSI;

2

estradiol on cycle day

2, ethnicity, age, BMI, smoking, alcohol, stimulation scheme, folic acid, IVF/ICSI treatment;

3

age, BMI, ethnicity,

IVF/ICSI, stimulation scheme.

Furthermore, the reduction of follicles results in less estradiol synthesizing granulosa cells leading to a

reduction of the estradiol level. In addition, omega-3 LC-PUFAs may affect the responses of the

granulosa cells to gonadotropins due to interactions with transcription factors involved in the steroi-

Dietary Patterns and Human Reproduction

Figures