The contribution of diet, physical activity and

Human Reproduction, Vol.28, No.8 pp. 2276 –2283, 2013 Advanced Access publication on June 15, 2013 doi:10.1093/humrep/det256

ORIGINAL ARTICLE Reproductive epidemiology

The contribution of diet, physical activity and sedentary behaviour to body mass index in women with and without polycystic ovary syndrome 1

Women’s Public Health Research, School of Public Health and Preventive Medicine, Monash University, Clayton, VIC 3168, Australia 2The Robinson Institute, Discipline of Obstetrics and Gynaecology, University of Adelaide, 55 King William Road, North Adelaide, SA 5006, Australia 3 Centre for Physical Activity and Nutrition Research, School of Exercise and Nutrition Sciences, Deakin University, Melbourne, VIC 3125, Australia 4School of Human Movement Studies, University of Queensland, Brisbane, QLD 4072, Australia 5Diabetes Unit, Southern Health, Clayton, VIC 3168, Australia *Correspondence address. Tel: +61-8-8313-1352; Fax: +61-8-8161-7652; E-mail: [email protected]

Submitted on April 6, 2013; resubmitted on May 17, 2013; accepted on May 22, 2013

study question: What is the contribution of diet, physical activity and sedentary behaviour to body mass index (BMI) in women with and without polycystic ovary syndrome (PCOS)? summary answer: PCOS status, higher energy intake and glycaemic index and lower physical activity were independently associated with BMI.

what is known already: Obesity worsens the clinical features of PCOS and women with PCOS have an elevated prevalence of overweight and obesity. It is not known whether there is a contribution of lifestyle factors such as dietary intake, physical activity or sedentary behaviour to the elevated prevalence of obesity in PCOS. study design, size, duration: This study is a population-based observational study with data currently collected at 13 year followup. The study commenced in 1996. For this analysis, data are analysed at one time point corresponding to the Survey 5 of the cohort in 2009. At this time 8200 participants remained (58% retention of baseline participants) of which 7466 replied to the questionnaire; 409 self-reported a diagnosis of PCOS and 7057 no diagnosis of PCOS. participants/materials, setting, methods: Australian women born in 1973–1978 from the Australian Longitudinal Study on Women’s Health.

main results and the role of chance: Mean BMI was higher in women with PCOS compared with non-PCOS (29.3 + 7.5 versus 25.6 + 5.8 kg/m2, P , 0.001). Women with PCOS reported a better dietary intake (elevated diet quality and micronutrient intake and lower saturated fat and glycaemic index intake) but increased energy intake, increased sitting time and no differences in total physical activity compared with non-PCOS. PCOS status, higher energy intake and glycaemic index and lower physical activity, as well as age, smoking, alcohol intake, occupation, education and country of birth, were independently associated with BMI.

limitations, reasons for caution: The weaknesses of this study include the self-reported diagnosis of PCOS, and the women not reporting PCOS not having their control status clinically verified which is likely to underrepresent the PCOS population. We are also unable to determine if lifestyle behaviours contributed to the PCOS diagnosis or were altered in response to diagnosis. wider implications of the findings: The strengths of this study include the community-based nature of the sample which minimizes selection bias to include women with a variety of clinical presentations. These results are therefore generalizable to a broader population than the majority of research in PCOS examining this research question which are performed in clinic-based populations. This study is in agreement with the literature that PCOS is independently associated with elevated BMI. We provide new insights that diet quality is subtly improved but that sedentary behaviour is elevated in PCOS and that PCOS status, higher energy intake and glycaemic index and lower physical activity are independently associated with BMI. & The Author 2013. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please email: [email protected]

Downloaded from http://humrep.oxfordjournals.org/ at Monash University on May 28, 2014

L.J. Moran 1,2,*, S. Ranasinha 1, S. Zoungas 1, S.A. McNaughton 3, W.J. Brown4, and H.J. Teede1,5

2277

Association of diet, exercise and BMI in PCOS

study funding/competing interest(s): L.J.M. was supported by a South Australian Cardiovascular Research Development Program (SACVRDP) Fellowship (AC11S374); a program collaboratively funded by the National Heart Foundation of Australia, the South Australian Department of Health and the South Australian Health and Medical Research Institute, S.A.M. was funded by an Australian Research Council Future Fellowship (FT100100581), S.Z. was funded by a Heart Foundation Career Development Fellowship (ID CR10S5330) and H.J.T. was funded by an NHMRC fellowship (ID 545888). None of the authors has any conflict of interest to declare.

trial registration number: Not applicable. Key words: polycystic ovary syndrome / dietary intake / physical activity / sedentary behaviour / body mass index

Introduction

Materials and Methods Study population This study is based on data from the Australian Longitudinal Study on Women’s Health (ALSWH)—a longitudinal population-based study of three age cohorts of Australia women. Women were randomly selected from the national health insurance scheme (Medicare) database, which includes almost all people who are permanent residents of Australia, with national recruitment and intentional over-sampling from rural and remote areas (Lee et al., 2005). Further details of the methods and characteristics of the sample have been reported elsewhere (Brown et al., 1998; Lee, 2001; Powers and Loxton, 2010). A comparison of women who participated in the baseline survey with data from women in the same age range from the Australian census of 1996 showed that the ALSWH participants were reasonably representative of the general population, although they were slightly more likely to be Australian born and to have a post school qualification when first recruited in 1996 (Brown et al., 1998; Lee et al., 2005). The Human Research Ethics Committees of the University of Newcastle and the University of Queensland approved the study methods and informed written consent was obtained from each participant. The current study uses data from the cohort of younger women (born 1973 – 1978; n ¼ 14,779 at Survey 1) who first completed a mailed survey in 1996 (Lee et al., 2005). For this analysis, data are from Survey 5 (2009, n ¼ 8200, 58% retention of baseline participants). We analysed data from 7466 women who completed Survey 5 and responded to the question on PCOS diagnosis (‘In the last 3 years have you been diagnosed with or treated for PCOS’) of which n ¼ 409 were classified as PCOS and n ¼ 7057 as non-PCOS. The analyses in this study are based on cross-sectional analysis of diet, physical activity and BMI at Survey 5 in women with and without PCOS. Variables for which there are missing data are reported in Supplemental data, Table SI. Participants were excluded who had incomplete food frequency questionnaire (FFQ) data (.10% of items missing responses; n ¼ 3) or those who reported daily energy intake of .14 700 kJ/day or ,2100 kJ/day (n ¼ 160). No specific inclusion or exclusion criteria were applied to this cohort and all women were included irrespective of pregnancy, medication, country of birth and language spoken.

Anthropometric and demographic variables Self-reported height, weight and BMI were reported with overweight and obesity defined by the World Health Organization criteria (BMI ≥ 25 kg/ m2 for overweight and obesity, BMI ≥ 30 kg/m2 for obesity; WHO, 2000). Demographic variables including parity, education, occupation and


Polycystic ovary syndrome (PCOS) affects up to 18% of reproductiveaged women (March et al., 2010) and is associated with reproductive (hyperandrogenism, menstrual irregularity, anovulation, infertility, pregnancy complications), metabolic (elevated risk factors for type 2 diabetes mellitus and cardiovascular disease and elevated prevalence of type 2 diabetes mellitus and cardiovascular disease) and psychological characteristics (worsened quality of life and elevated prevalence of anxiety and depression; Teede et al., 2011). Insulin resistance is a pathophysiological feature of PCOS and is proposed to be mechanistically distinct (or intrinsic) from obesity-associated insulin resistance. Insulin resistance present in the majority of women with PCOS, including lean women (Dunaif et al., 1989), and worsens the clinical presentation. Weight gain further compounds insulin resistance (Erdmann et al., 2008). Given the association between insulin resistance, obesity and the presentation of PCOS, weight management is a key initial treatment strategy for PCOS (Teede et al., 2011) and improves the reproductive, metabolic and psychological features (Moran et al., 2011). There is emerging evidence that women with PCOS have an elevated risk of being overweight and obese and have increased longitudinal weight gain compared with community controls (Teede et al., 2013). A recent meta-analysis has reported increased prevalence of overweight or obesity [1.95 risk ratio; 95% confidence interval (CI) 1.52 –2.50] for women with PCOS compared with controls (Lim et al., 2012). We have proposed that there is a bidirectional interaction between PCOS and weight with PCOS driving weight gain and weight gain contributing to an increased prevalence and severity of PCOS (Teede et al., 2013). Indeed, women with PCOS may have specific physiological or intrinsic barriers to maintaining a healthy weight such as insulin resistance, hyperinsulinaemia and hyperandrogenism, which can contribute to weight or abdominal fat gain (1988, Pasquali, 2006), reduce energy expenditure and increase food intake (Felig, 1984; Welle et al., 1988; FranssilaKallunki and Groop, 1992; Robinson et al., 1992; Carlson and Campbell, 1993; Kersten, 2001; Moran et al., 2004; Hirschberg et al., 2004; Georgopoulos et al., 2008; Ryan et al., 2008). Women with PCOS may additionally have altered energy balance caused by extrinsic factors, supported by reports of reduced physical activity (Eleftheriadou et al., 2012) and increased intake of high glycaemic index foods compared with controls (Douglas et al., 2006). Conversely, other research suggest no differences in energy or dietary intake such as macronutrient, micronutrient or food group intake, physical activity or muscle strength between lean or overweight women with or without PCOS (Wright et al., 2004; Douglas et al., 2006; Thomson et al., 2009). Given the prevalence and health burden of PCOS and the propensity to, and adverse impact of excess weight in PCOS, it is important to explore

potentially modifiable extrinsic or environmental factors that may contribute to obesity in women with PCOS to guide management. The aim of this study was to examine the association of demographic variables and extrinsic factors, including diet and physical activity, with body mass index (BMI) in a large community study of women with and without diagnosed PCOS.

2278 income were collected at Survey 5 and area of residence was measured at Survey 1.

Dietary intake and physical activity

Statistical analysis Data are reported as mean + SD or median + interquartile range (IQR) as documented and were analysed using Stata software version 11.2

(StataCorp, TX, USA). Prior to analysis, the normality assumptions of the data were examined. Univariate regression analyses were used to compare continuous variables and x2 tests were used to compare categorical variables between populations. Variables significant at a , 0.05 in the univariable analysis or based on hypothesis testing as being potentially associated with BMI as the primary outcome were included in the multivariable regression analysis. Multivariable linear regression analysis was used to examine the independent predictors of BMI at Survey 5. The interaction of PCOS status and energy intake was assessed for BMI. Analyses were conducted using survey commands for analysing data weighted by area of residence to adjust for the deliberate over-sampling in rural and remote areas.

Results Participant characteristics Participant characteristics are reported in Table I. The women with PCOS were younger, and were more likely to not have children than those without PCOS. The women with PCOS also reported a higher BMI than women without PCOS, with a different BMI distribution (a higher proportion in the obese and a lower proportion in the healthy BMI categories).

Dietary intake and physical activity The dietary intake and physical activity levels of the participants are reported in Tables I and II. Women with PCOS had a better diet quality as indicated by a higher diet quality score and higher energy, fibre, folate, iron, calcium, magnesium, niacin, phosphorus, potassium, sodium, vitamin E and zinc intake and lower percentage energy from saturated fat intake, glycaemic index and retinol intake than women without PCOS. Analyses adjusted for mean intake and energy adjusted intake gave very similar results, except differences in the levels of sodium (P ¼ 0.764), folate (P ¼ 0.119), calcium (P ¼ 0.503), zinc (P ¼ 0.217), niacin (P ¼ 0.210) and potassium (P ¼ 0.056) between women with and without PCOS were no longer statistically significant (data not reported). Women with PCOS reported greater sitting time than women without PCOS. No differences in overall reported physical activity levels between women with and without PCOS were observed.

Associations of demographic factors, dietary intake and physical activity with BMI The independent associations of demographic variables, dietary intake and physical activity with BMI are reported in Table III. In univariable analysis, PCOS status, age, physical activity, energy intake, alcohol intake, country of birth, education, occupation, % fat, % protein, % carbohydrate, glycaemic index and smoking status were all associated with BMI. In multiple variable analyses, PCOS status remained independently associated with an increased BMI. A 1000 MET/min increase in physical activity was associated with a 0.42 kg/m2 decrease in BMI and a 1000 kJ increase in energy intake was associated with a 0.44 kg/m2 increase in BMI. Increased age, higher glycaemic index, lower alcohol intake and smoking status also remained independently associated with increased BMI. Women born in Asia were more likely to have lower BMI than women born in Australia, women with a degree or higher were more likely to have a lower BMI than women with no formal education and associate professionals were more likely to have a higher BMI than professionals. There was no evidence of any interaction between BMI and


At Survey 5, self-reported dietary intake data were collected from the Dietary Questionnaire for Epidemiological Studies (DQES) Version 2, an FFQ developed by The Cancer Council of Victoria previously validated in young Australian women (Hodge et al., 2000). The DQES collects information regarding frequency of consumption of 80 food and beverage items utilizing a 10-point frequency scale (‘never’ to ‘three or more times per day’) over the previous 12 months and 10 additional questions on fruit, vegetable, sugar, eggs, milk, cheese, bread and fat spread consumption. Portion size photographs were provided to adjust the portion size. Nutrient intakes were calculated using Australian food composition data (Lewis et al., 1995). A diet quality index was calculated using the previously described dietary guidelines index (DGI) (McNaughton et al., 2008, 2009). The DGI reflects compliance with Dietary Guidelines for Australian adults and comprises dietary indicators of vegetables and legumes, fruit, total cereals, meat and alternatives, total dairy, saturated fat, alcoholic beverages, added sugars and ‘extra foods’ which are defined as foods that are not essential to provide nutrient requirements and contain too much fat, sugar and salt such as soft drinks, cordials, fruit juice drinks, mayonnaise and dressing, chips, jam and marmalade, confectionery, chocolate, hamburgers, hot chips, meat pies, pizza, cakes and muffins, pies and pastries, puddings, ice cream, cream and biscuits. The indicators used were based on the dietary guidelines, cut-points and food groups guided by the Australian Guide to Healthy Eating that provides age- and sex-specific recommendations for the consumption of five core food groups (vegetables, fruits, cereals, meat and alternatives and dairy) and ‘extra foods’ and national recommendations for saturated fat and added sugars (Department of Health and Ageing, 1998). Each component was scored from 0 to 10 with 10 indicating an optimal intake. The total score was the sum of 13 indicators with the DGI having a possible range of 0 – 130 with a higher score indicating increased compliance with the dietary guidelines. Nutrients were analysed as mean intakes. Intakes of micronutrients, diet quality, glycaemic index, glycaemic load, fibre, cholesterol, sodium and alcohol were adjusted for energy intake using the nutrient density method as recommended in validation studies (Willett, 2001). Physical activity data were recorded by self-reported recall and collected on frequency and duration of a variety of leisure time and active transport activities, including walking briskly (for recreation or exercise or to get to or from places), moderate-intensity leisure-time physical activity (like social tennis, moderate exercise classes, recreational swimming, dancing) and vigorous-intensity leisure-time physical activity (that makes you breathe harder or puff and pant) in the last week, for activities lasting 10 min or more. Physical activity was calculated as the sum of the products of total weekly minutes in each of the three categories of physical activity and the metabolic equivalent value (MET) assigned to each category: [(walking minutes × 3.0 METs) + (moderate-intensity physical activity minutes × 4.0 METs) + (vigorous-intensity physical activity minutes × 7.5 METs). Outliers were truncated at 28 h/week for total physical activity. Sedentary activity was calculated based on sitting time which was assessed with the following question: ‘How many hours in total do you typically spend sitting down while doing things like visiting friends, driving, reading, watching television, or working at a desk or computer?’: (a) on a usual weekday; and (b) on a usual weekend-day. This is similar to that used in the International Physical Activity Questionnaire which has been shown to be valid and reliable (Rosenberg et al., 2008).

Moran et al.

2279


Table I Characteristics for women with and without PCOS. Overall (n 5 8200)

PCOS (n 5 409)

Non-PCOS (n 5 7057)

P-value

............................................................................................................................................................................................. Age (years)

33.7 (1.5)

Current smoker

1168 (14%)

BMI (kg/m2)

25.8 (5.9)

29.3 (7.5)

Normal (BMI , 25 kg/m2)

4309 (53%)

141 (35%)

3822 (54%)

Overweight(BMI ≥ 25 , 30 kg/m2)

1969 (24%)

89 (22%)

1731 (25%)

Obese (BMI ≥ 30 kg/m2)

1527 (19%)

163 (40%)

1270 (18%)

71.3 (17)

80.4 (21)

70.8 (17)

,0.001 0.9

Weight (kg)

33.5 (1.4) 57 (14%)

33.7 (1.5)

0.01

351 (5%)

0.3

25.6 (5.8)

,0.001

Country of birth Australia

379 (93%)

6529 (93%)

357 (4%)

19 (5%)

331 (5%)

Asian

122 (2%)

6 (2%)

106 (2%)

Low (6– 36k)

2856 (35%)

146 (36%)

2494 (35%)

Medium (36–77k)

2022 (25%)

89 (22%)

1785 (25%)

High (.77k)

2924 (36%)

157 (38%)

2545 (36%)

Personal income ($AUS) 0.9

Education No formal education/school/year (Y10/Y12)

1666 (21%)

82 (20%)

1447 (21%)

Trade/diploma

2109 (26%)

100 (24%)

1854 (26%)

Degree or higher

4140 (51%)

213 (52%)

3616 (51%)

Professional

3438 (43%)

180 (44%)

2980 (42%)

Associate professional

1437(18%)

72 (18%)

1264 (18%)

Clerical/trade

1387 (17%)

74 (18%)

1205 (17%)

No paid job

1671 (21%)

74 (18%)

1477 (21%)

0.7

Main occupation 0.34

Parity ,0.001

No Children

2923 (36%)

187 (46%)

2552 (36%)

One or more

5146 (64%)

222 (54%)

4490 (64%)

Physical activity (METs/min)

820 (898)

814 (875)

820 (895)

0.75

Sedentary behaviour (h/day)

5.8 (2.8)

6.3 (2.8)

5.8 (2.9)

0.008

Data are mean (SD) or number (%). Data were analysed by survey-weighted univariable regression analysis to compare continuous variables and x2 test to compare categorical variables between populations. BMI, body mass index; MET, metabolic equivalent value.

energy intake by PCOS status (b ¼ 0.00013, 95% CI 20.00002 to 0.0005, P ¼ 0.5).

Discussion Here we confirm prior reports of greater weight and a greater prevalence of overweight and obesity in women with PCOS compared with controls and an independent association of PCOS status with BMI (Glueck et al., 2003; Teede et al., 2013). We advance our knowledge in this field by reporting in a large population-based cohort of women using detailed dietary and physical activity assessment tools that women with PCOS had a better dietary intake as reflected by improved diet quality, lower saturated fat and glycaemic index intake and higher fibre and micronutrient intake, but a higher energy intake and increased amount of sedentary time compared with controls. We report here for the first time that diet quality was marginally better in women with PCOS compared with controls. This occurred despite a

greater energy intake for the women with PCOS in contrast to prior reports (Douglas et al., 2006; Colombo et al., 2009; Altieri et al., 2012). As reported in general population data, this better diet quality was characterized by greater fibre and micronutrient intake and less saturated fat intake (McNaughton et al., 2008) although we note the modest magnitude of these differences between women with PCOS. We also note that overall both women with and without PCOS had high saturated fat levels (15% of total energy) compared with dietary recommendations (U.S. Department of Agriculture and U.S. Department of Health and Human Services, 2010). Diet quality is a global assessment of dietary intake and inversely associated with an unfavourable metabolic risk profile and higher risk of all cause and chronic disease mortality (McNaughton et al., 2009; Wirt and Collins, 2009). However, a 10 unit greater DGI has been associated with lower systolic blood pressure (0.53 mmHg), total cholesterol (0.042 mmol/l), fasting glucose (0.003 mmol/l) and insulin levels (0.014 U/l; McNaughton et al., 2009). This suggests that the observed 1.4 unit difference in DGI in


7490 (93%)

English speaking/European

2280

Moran et al.

Table II Dietary intake for women with and without PCOS. Overall (n 5 7466)

PCOS (n 5 409)

Non-PCOS (n 5 7057)

P-value

............................................................................................................................................................................................. DGI

86.7 (11.3)

88.1 (11.9)

86.7 (11.2)

0.004

Energy (kJ/day)

7040 (2275)

7245 (2577)

7030 (2265)

0.02

Protein (% energy)

20.9 (3.3)

21.1 (3.2)

20.9 (3.3)

0.34 0.75

Carbohydrate (% energy)

40.3 (5.7)

40.5 (6.0)

40.3 (5.6)

Fat (% energy)

39.9 (4.9)

36.6 (5.2)

37.0 (4.9)

0.3

SFA (% energy)

15.4 (3.1)

15.1 (3.1)

15.4 (3.1)

0.01

MUFA (% energy)

13.1 (2.1)

13.1 (2.3)

13.1 (2.1)

0.86

PUFA (% energy)

5.1 (1.6)

5.2 (1.6)

5.1 (1.6)

0.3

Glycaemic index

50.8 (3.9)

50.3 (4.1)

50.7 (3.9)

0.005

Glycaemic load

86.9 (33.5)

Fibre (g/day) Sodium (mg/day)

20.2 (8.4)

86.9 (33.5)

0.13

19.0 (7.0)

0.001

266 (109)

270 (119)

265 (109)

0.17

2236 (825)

2296 (923)

2233 (819)

0.048

Alcohol (g/day)

9.2 (13.3)

8.4 (13.4)

9.3 (13.4)

Folate (mg/day)

243 (88.4)

253 (107)

242 (87)

0.009

0.40

Iron (mg/day)

11.7 (4.6)

12.3 (5.4)

11.6 (4.5)

0.003

Zinc (mg/day)

10.7 (3.9)

11.2 (4.6)

10.7 (3.9)

0.013

Calcium (mg/day)

850 (290)

874 (324)

848 (288)

0.048

Magnesium (mg/day)

258 (86)

272 (104)

258 (85)

0.002

Phosphorus (mg/day)

1405 (450)

1471 (536)

1401 (444)

0.002

Potassium (mg/day)

2535 (777)

2630 (914)

2528 (770)

0.004

Beta-carotene (mg/day)

2509 (1260)

2545 (1325)

2508 (1264)

0.19

Niacin (mg/day)

19.2 (7.8)

20.1 (9.2)

19.2 (7.6)

0.013

Retinol (mg/day)

310 (146)

295 (135)

311 (146)

0.026

Riboflavin (mg/day)

2.2 (0.9)

2.3 (1.0)

2.2 (0.8)

0.12

Thiamine (mg/day)

1.4 (0.6)

1.4 (0.7)

1.4 (0.6)

0.24

Vitamin C (mg/day)

103.5 (56)

106 (63)

103 (57)

0.15

Vitamin E (mg/day)

5.6 (2.1)

5.9 (2.3)

5.6 (2.1)

0.001

Data are mean (SD). Data were analysed by survey-weighted univariable regression analysis to compare continuous variables and x2 test to compare categorical variables between populations. DGI, dietary guideline index; MUFA, monounsaturated fat; PUFA, polyunsaturated fat; SFA, saturated fat.

this study is likely to be of minimal clinical relevance. Women with PCOS reported more optimal fibre, micronutrient, glycaemic index and saturated fat intake. This is consistent with previous studies that have reported reduced % fat and higher fibre intake independent of energy intake for women with PCOS (Altieri et al., 2012). Conversely, prior conflicting reports have noted no differences in macronutrient, micronutrient and glycaemic index (Douglas et al., 2006; Colombo et al., 2009), greater intakes of high glycaemic index foods, fat, animal fat, saturated fat or cheese intake and less fibre intake (Wild et al., 1985; Douglas et al., 2006; Colombo et al., 2009; Altieri et al., 2012) in women with or without PCOS. These discrepancies may be explained by methodological differences across the studies, including smaller sample sizes (n ¼ 294 maximum), predominantly clinic-based populations and varying dietary intake assessment. Our finding of better dietary intake in PCOS may indicate increasing recognition and uptake of healthy lifestyle recommendations for management of PCOS from national and international guidelines and position statements (Moran et al., 2009; Teede et al., 2011). Of note,

the observed better dietary intake occurred despite elevated BMI and reported energy intake for PCOS compared with controls. Indeed, the improved dietary intake for PCOS, in this study, may be partially related to higher food quantity as indicated by the removal of significant differences between women with and without PCOS for some micronutrients on when analyses were adjusted for energy intake. This is in contrast to some, but not all, studies where reduced energy intake was associated with improved dietary quality in the general population and higher fat, saturated fat and sugar intake were associated with elevated energy intake in PCOS (McNaughton et al., 2008; Wolongevicz et al., 2010; Barr et al., 2011). This may indicate that women with PCOS are receptive to education to optimize dietary intake but do not adequately translate education with respect to appropriate portion sizes. We report here for the first time that sedentary behaviour, as assessed by sitting time, is increased in women with PCOS. Sedentary behaviour is being increasingly recognized, in addition to total physical activity, as a contributor both to elevated adiposity and to obesity-associated


Cholesterol (mg/day)

89 (36.3)

19 (7.1)

2281


Table III Multiple variable linear regression analysis with BMI as the outcome variable and demographic factors, dietary intake and physical activity as explanatory variables. Crude b (95% CI)

P-value

Adjusted b (95% CI)

P-value

............................................................................................................................................................................................. PA (METs/min) (×1000) Energy (kJ/day) (×1000) Alcohol (g/day) Fat (%) Protein (%) Carbohydrate (%) Glycaemic Index DGI PCOS Age (years)

20.43 (20.56 to 20.29) 0.2 ( 0.06– 0.3) 20.25 (20.36 to 20.15)

,0.001 ,0.001 ,0.001

20.42 (20.6 to 20.2) 0.44 (0.05– 0.8) 20.03 (20.04 to 20.02)

,0.001 0.03 ,0.001

0.076 (0.05–0.12)

,0.001

0.17 (20.44 to 0.77)

0.60

0.25 (0.2–0.28)

,0.001

0.51 (20.1 to 1.1)

0.10

20.14 (20.16 to 20.11) 0.08 (0.045– 0.12) 20.0095 (20.02 to 0.003) 3.5 (2.7–4.4) 0.09 (0.002– 0.18)

,0.001

0.18 (20.46 to 0.81)

,0.001

0.13 (0.07– 0.17)

0.143 ,0.001

ns 3.6 (2.6–4.3) 0.12 (0.02– 0.22)

20.043 (20.65 to 0.56)

ns ,0.001 0.018

Country of Birth Australia

1

1

Europe/English speaking

20.28 (20.88 to 0.30)

0.34

Asia

22.98 (23.6 to 22.32)

,0.001

22.9 (23.6 to 22.1)

0.98 ,0.001

Education No formal education/school (Y10/Y12) Trade/diploma Degree or higher

1 20.46 (20.9 to 20.005) 22.5 (22.9 to 22.1)

1 0.043 ,0.001

20.25 (20.8 to 0.26) 21.9 (22.3 to 21.4)

0.33 ,0.001

Occupation Professional Associate professional Clerical/trade No paid job Smoke (year/n)

1 1.1 (0.62–1.6)

1 ,0.001

0.002

20.13 (20.64 to 0.39)

0.63

,0.001

0.14 (20.29 to 0.56)

0.53

,0.001

0.55 (0.08– 1.03)

0.02

0.091 (20.36 to 0.54)

0.65

20.86 (21.2 to 20.49) 1.1 (0.7–1.5)

0.84 (0.3–1.4)

Parity No children One or more

1 0.075 (20.21 to 0.35)

0.61

ns

ns

Data were analysed by multivariable linear regression analysis. Variables significant at a , 0.05 in the univariable (crude) analysis were included in the multivariable regression (adjusted) analysis. DGI, dietary guideline index; NS, not significant; PA, physical activity; PCOS, polycystic ovary syndrome; MET, metabolic equivalent value.

metabolic diseases such as type 2 diabetes mellitus (Hu et al., 2003). The clinical implications of the 0.5 h/day greater sitting time for women with PCOS remain unclear. We note that prior research suggested that each 2 h/day increment in TV watching is associated with a 23% increase in obesity and 14% increase in type 2 diabetes risk (Hu et al., 2003). Given the increase in obesity and obesity-associated metabolic diseases in PCOS, sedentary behaviour provides an additional therapeutic target for lifestyle interventions in PCOS. We confirm some (Wright et al., 2004; Teede et al., 2013) but not all reports (Wild et al., 1985; Eleftheriadou et al., 2012) of no difference in total physical activity between women with and without PCOS. Research in PCOS has primarily focused on modification of dietary intake, although recent evidencebased guidelines for the management of PCOS have emphasized the need for specific structured physical activity advice (Teede et al., 2011). This suggests a greater awareness of dietary versus exercise recommendations for PCOS. Preliminary findings also indicate reduced cardiopulmonary capacity as a marker of fitness and potential physical activity tolerance (Orio et al., 2006) in women with PCOS

compared with weight or BMI-matched controls. This may indicate physiological or psychological barriers to uptake of optimal physical activity recommendations. Optimizing dietary intake, physical activity and sedentary behaviour will improve the success of weight management programmes in PCOS and thus focusing on all these factors in future interventions is clearly warranted. We confirm here that PCOS is independently associated with BMI (Teede et al., 2013). Lifestyle variables independently associated with BMI for all women included higher energy intake and lower physical activity levels, as would be expected through modulating energy balance. As previously reported, elevated glycaemic index (Murakami et al., 2007) and lower alcohol intake were also associated with elevated BMI (Murakami et al., 2007). However, we report here no associations between diet quality and BMI in contrast to prior reports of diet quality being associated with the development of overweight and obesity (Wolongevicz et al., 2010). As a primary contributor to adiposity, we observed no interaction between energy intake and PCOS status on BMI. This indicates energy intake is a similar independent predictor of


0.046

0.59 ,0.001

2282

Supplementary data Supplementary data are available at http://molehr.oxfordjournals.org/

Acknowledgements The research on which this paper is based was conducted as part of the Australian Longitudinal Study on Women’s Health, which was conceived and developed by groups of inter-disciplinary researchers at The University of Newcastle and The University of Queensland. We are grateful to the Australian Government Department of Health and Ageing for funding and to the women who provided the survey data.

Authors’ roles L.J.M., S.R., S.Z., S.A.M., W.J.B. and H.J.T. designed the research, S.R. and S.Z. performed statistical analysis, L.J.M. wrote the paper and had primary responsibility for final content and all authors read and approved the final manuscript.

Funding L.J.M. was supported by a South Australian Cardiovascular Research Development Program (SACVRDP) Fellowship (AC11S374); a program collaboratively funded by the National Heart Foundation of Australia, the South Australian Department of Health and the South Australian Health and Medical Research Institute. S.A.M. was funded by an Australian Research Council Future Fellowship (FT100100581), S.Z. was funded by a Heart Foundation Career Development Fellowship (ID CR10S5330) and H.J.T. was funded by an NHMRC fellowship (ID 545888).

Conflict of interest All authors report no conflict of interest.

References Department of Health and Ageing. The Australian Guide to Healthy Eating. Canberra, Australia: Australian Government Department of Health and Ageing, 1998. Altieri P, Cavazza C, Pasqui F, Morselli AM, Gambineri A, Pasquali R. Dietary habits and their relationship with hormones and metabolism in overweight and obese women with polycystic ovary syndrome. Clin Edocrinol 2012; 78:52 – 59. Barr S, Hart K, Reeves S, Sharp K, Jeanes YM. Habitual dietary intake, eating pattern and physical activity of women with polycystic ovary syndrome. Eur J Clin Nutr 2011;65:1126 –1132. Brown WJ, Bryson L, Byles JE, Dobson AJ, Lee C, Mishra G, Schofield M. Women’s Health Australia: recruitment for a national longitudinal cohort study. Women Health 1998;28:23– 40. Carlson MG, Campbell PJ. Intensive insulin therapy and weight gain in IDDM. Diabetes 1993;42:1700– 1707. Colombo O, Pinelli G, Comelli M, Marchetti P, Sieri S, Brighenti F, Nappi RE, Tagliabue A. Dietary intakes in infertile women a pilot study. Nutr J 2009;8:53. Douglas CC, Norris LE, Oster RA, Darnell BE, Azziz R, Gower BA. Difference in dietary intake between women with polycystic ovary syndrome and healthy controls. Fertil Steril 2006;86:411 – 417. Dunaif A, Segal KR, Futterweit W, Dobrjansky A. Profound peripheral insulin resistance, independent of obesity, in polycystic ovary syndrome. Diabetes 1989;38:1165 – 1174.


adiposity for women both with and without PCOS. Given the subtle improvements in dietary intake, modest increase in sedentary behaviour and similar total physical activity for women with PCOS, elevated adiposity in PCOS may not be related to adverse lifestyle behaviours. It is possible that women with PCOS have a greater tendency to elevated adiposity due to intrinsic hormonal aberrations, including insulin resistance and hyperandrogenism (1988, Felig, 1984; Franssila-Kallunki and Groop, 1992; Robinson et al., 1992; Carlson and Campbell, 1993; Welle et al., 1988; Kersten, 2001; Moran et al., 2004; Hirschberg et al., 2004; Pasquali, 2006; Georgopoulos et al., 2008; Ryan et al., 2008). Alternatively, obesity in PCOS may be associated with alterations in homeostatic mechanisms of weight regulation, including dysregulation of appetite control, worsened subjective hunger and satiety and gut hormone regulation (Moran et al., 2004) or a reduced resting metabolic rate (Georgopoulos et al., 2008) and thermic effect of food (Robinson et al., 1992). Further research is needed to elucidate these mechanisms. The strengths of this study include the community-based nature of the sample which minimizes selection bias to include women with a variety of clinical presentations. The majority of research in PCOS examines clinicbased populations. Owing to the well-recognized association between adiposity and the features of PCOS, this may indicate a population with worsened reproductive, metabolic or psychological features, and a potential bias towards the women with PCOS being more overweight as reported by Ezeh et al. (2013), who note greater BMI and clinical and biochemical hyperandrogenism for women with clinic referred PCOS compared with population diagnosed PCOS. The weaknesses of this study include the self-reported diagnosis of PCOS and the women not reporting PCOS not having their control status clinically verified which is likely to underrepresent the PCOS population and contribute to the lower prevalence of PCOS observed here than in prior studies (March et al., 2010). In the young ALSWH cohort, PCOS status was ascertained at 10 and 13 years following recruitment (Surveys 4 and 5). It is therefore not possible to assess PCOS incidence due to the uncertain timing of PCOS diagnosis. Because of the study design, we are also unable to determine if lifestyle behaviours contributed to the PCOS diagnosis or were altered in response to diagnosis and report here only associations between dietary and physical activity variables and adiposity. In future studies, prospective longitudinal follow-up is warranted to assess the causality of relationships. We also note the 58% retention of participants in this study compared with baseline levels 13 years prior which may indicate bias and limit generalizability. However, prior research in this cohort has reported no differences between completors and noncompletors indicating a likely minimal effect of attrition on outcomes (Powers and Loxton, 2010). In conclusion, we report novel data showing that PCOS is independently associated with elevated BMI which may be partially attributable to higher energy intake and sedentary time in PCOS. In contrast, women with PCOS had a subtly better dietary intake and no differences in physical activity compared with controls. Independent predictors of adiposity in women with and without PCOS included elevated energy intake and glycaemic index and lower physical activity levels, highlighting future targets for lifestyle interventions in PCOS. These findings have implications for the pathophysiology of obesity development and management in PCOS. Adequately powered randomized controlled trials are needed to advise practice on optimal lifestyle management in PCOS.

Moran et al.


Androgen Excess and Polycystic Ovary Syndrome Society. Fertil Steril 2009;92:1966 – 1982. Moran LJ, Hutchison SK, Norman RJ, Teede HJ. Lifestyle changes in women with polycystic ovary syndrome. Cochrane Database Syst Rev 2011;2. CD007506. Murakami K, Sasaki S, Okubo H, Takahashi Y, Hosoi Y, Itabashi M. Dietary fiber intake, dietary glycemic index and load, and body mass index: a cross-sectional study of 3931 Japanese women aged 18 – 20 years. Eur J Clin Nutr 2007;61:986– 995. Orio F Jr, Giallauria F, Palomba S, Cascella T, Manguso F, Vuolo L, Russo T, Tolino A, Lombardi G, Colao A et al. Cardiopulmonary impairment in young women with polycystic ovary syndrome. J Clin Endocrinol Metab 2006;91:2967 – 2971. Pasquali R. Obesity and androgens: facts and perspectives. Fertil Steril, 2006; 85:1319– 1340. Powers J, Loxton D. The impact of attrition in an 11-year prospective longitudinal study of younger women. Ann Epidemiol 2010;20:318 – 321. Robinson S, Chan SP, Spacey S, Anyaoku V, Johnston DG, Franks S. Postprandial thermogenesis is reduced in polycystic ovary syndrome and is associated with increased insulin resistance. Clin Endocrinol 1992;36:537–543. Rosenberg DE, Bull FC, Marshall AL, Sallis JF, Bauman AE. Assessment of sedentary behavior with the International Physical Activity Questionnaire. J Phys Act Health 2008;5(Suppl 1):S30– 44. Ryan M, Livingstone MB, Ducluzeau PH, Salle A, Genaitay M, Ritz P. Is a failure to recognize an increase in food intake a key to understanding insulin-induced weight gain? Diabetes Care 2008;31:448– 450. Teede HJ, Misso ML, Deeks AA, Moran LJ, Stuckey BG, Wong JL, Norman RJ, Costello MF. Assessment and management of polycystic ovary syndrome: summary of an evidence-based guideline. Med J Austr 2011;195:S65–S112. Teede HJ, Joham AE, Paul E, Moran LJ, Loxton D, Jolley D, Lombard C. Longitudinal weight gain in women identified with Polycystic Ovary Syndrome: results of an observational study in young women. Obesity (2013) [Epub ahead of print]. Thomson RL, Buckley JD, Moran LJ, Noakes M, Clifton PM, Norman RJ, Brinkworth GD. Comparison of aerobic exercise capacity and muscle strength in overweight women with and without polycystic ovary syndrome. BJOG 2009;116:1242–1250. U.S. Department of Agriculture and U.S. Department of Health and Human Services. Dietary guidelines for Americans 2010, 7th edn. Washington, DC: US Government Printing Office, 2010. Welle S, Nair KS, Lockwood D. Effect of a sulfonylurea and insulin on energy expenditure in type II diabetes mellitus. J Clin Endocrinol Metab 1988; 66:593– 597. Wild RA, Painter PC, Coulson PB, Carruth KB, Ranney GB. Lipoprotein lipid concentrations and cardiovascular risk in women with polycystic ovary syndrome. J Clin Endocrinol Metab 1985;61:946 – 951. Willett W. Commentary: dietary diaries versus food frequency questionnaires-a case of undigestible data. Int J Epidemiol 2001;30:317–319. Weight Gain Associated with Intensive Therapy in the Diabetes Control and Complications Trial. The DCCT Research Group. Diabetes Care 1988; 11:567– 573. Wirt A, Collins CE. Diet quality—what is it and does it matter? Public Health Nutr 2009;12:2473 – 2492. Wolongevicz DM, Zhu L, Pencina MJ, Kimokoti RW, Newby PK, D’Agostino RB, Millen BE. Diet quality and obesity in women: the Framingham Nutrition Studies. Br J Nutr 2010;103:1223 – 1229. World Health Organization. Obesity: preventing and managing the global epidemic. Report of a WHO Consultation. WHO Technical Report Series 894. Geneva: World Health Organisation, 2000. Wright CE, Zborowski JV, Talbott EO, McHugh-Pemu K, Youk A. Dietary intake, physical activity, and obesity in women with polycystic ovary syndrome. Int J Obes 2004;28:1026– 1032.


Eleftheriadou M, Michala L, Stefanidis K, Iliadis I, Lykeridou A, Antsaklis A. Exercise and sedentary habits among adolescents with PCOS. J Pediatr Adolesc Gynecol 2012;25:172 – 174. Erdmann J, Kallabis B, Oppel U, Sypchenko O, Wagenpfeil S, Schusdziarra V. Development of hyperinsulinemia and insulin resistance during the early stage of weight gain. Am J Physiol Endocrinol Metab 2008;294:E568– E575. Ezeh U, Yildiz BO, Azziz R. Referral bias in defining the phenotype and prevalence of obesity in polycystic ovary syndrome. J Clin Endocrinol Metab 2013; 98:E1088–E1096. Felig P. Insulin is the mediator of feeding-related thermogenesis: insulin resistance and/or deficiency results in a thermogenic defect which contributes to the pathogenesis of obesity. Clin Physiol 1984;4:267 – 273. Franssila-Kallunki A, Groop L. Factors associated with basal metabolic rate in patients with type 2 (non-insulin-dependent) diabetes mellitus. Diabetologia 1992;35:962–966. Georgopoulos NA, Saltamavros AD, Vervita V, Karkoulias K, Adonakis G, Decavalas G, Kourounis G, Markou KB, Kyriazopoulou V. Basal metabolic rate is decreased in women with polycystic ovary syndrome and biochemical hyperandrogenemia and is associated with insulin resistance. Fertil Steril 2008; 92:250–255. Glueck CJ, Papanna R, Wang P, Goldenberg N, Sieve-Smith L. Incidence and treatment of metabolic syndrome in newly referred women with confirmed polycystic ovarian syndrome. Metabolism 2003;52:908 – 915. Hirschberg AL, Naessen S, Stridsberg M, Bystrom B, Holtet J. Impaired cholecystokinin secretion and disturbed appetite regulation in women with polycystic ovary syndrome. Gynecol Endocrinol 2004;19:79 – 87. Hodge A, Patterson AJ, Brown WJ, Ireland P, Giles G. The anticancer council of Victoria FFQ: relative validity of nutrient intakes compared with weighed food records in young to middle-aged women in a study of iron supplementation. Aust N Z J Public Health 2000;24:576 – 583. Hu FB, Li TY, Colditz GA, Willett WC, Manson JE. Television watching and other sedentary behaviors in relation to risk of obesity and type 2 diabetes mellitus in women. J Am Med Assoc 2003;289:1785– 1791. Kersten S. Mechanisms of nutritional and hormonal regulation of lipogenesis. EMBO Rep 2001;2:282– 286. Lee C. Women’s Health Australia: Progress on the Australian Longitudinal Study on Women’s Health 1995–2000. Brisbane: Australian Academic Press Pty Ltd, 2001. Lee C, Dobson AJ, Brown WJ, Bryson L, Byles J, Warner-Smith P, Young AF. Cohort profile: the Australian longitudinal study on women’s health. Int J Epidemiol 2005;34:987– 991. Lewis J, Milligan G, Hunt A. NUTTAB95: Nutrient Data Table for Use in Australia. Canberra, Australia: Australian Government Publishing Service, 1995. Lim SS, Davies MJ, Norman RJ, Moran LJ. Overweight, obesity and central obesity in women with polycystic ovary syndrome: a systematic review and meta-analysis. Hum Reprod Update 2012;18:618– 637. March WA, Moore VM, Willson KJ, Phillips DI, Norman RJ, Davies MJ. The prevalence of polycystic ovary syndrome in a community sample assessed under contrasting diagnostic criteria. Hum Reprod 2010;25:544–551. McNaughton SA, Ball K, Crawford D, Mishra GD. An index of diet and eating patterns is a valid measure of diet quality in an Australian population. J Nutr 2008;138:86 – 93. McNaughton SA, Dunstan DW, Ball K, Shaw J, Crawford D. Dietary quality is associated with diabetes and cardio-metabolic risk factors. J Nutr 2009; 139:734– 742. Moran LJ, Noakes M, Clifton PM, Wittert GA, Tomlinson L, Galletly C, Luscombe ND, Norman RJ. Ghrelin and measures of satiety are altered in polycystic ovary syndrome but not differentially affected by diet composition. J Clin Endocrinol Metab 2004;89:3337–3344. Moran LJ, Pasquali R, Teede HJ, Hoeger KM, Norman RJ. Treatment of obesity in polycystic ovary syndrome: a position statement of the

2283