. 2016 Oct 21;20(4):578–597. doi: 10.1017/S1368980016002639

Table 1.

Summary of all studies identified to assess sugar-sweetened beverages (SSB): design, population studied, dietary assessment instruments used and details of testing for validity and/or reproducibility. Studies were selected to be potentially suitable to assess SSB intake based on (i) the instrument was tested for validity and/or reproducibility and (ii) the instrument was used in more than two countries simultaneously which represent a range of European regions; and are indicated by ticks in the last column. Where validation or reliability data was not available for SSB specifically, this is highlighted in bold font

Study	Design	Population	Countries	Instrument(s)	Tested for validity	>2 countries/range	Validity	Reproducibility	Instrument selected
Adults
CNSHS⁽ ⁵⁶ ^, ⁵⁷ ⁾	Cross-sectional	Adults/students (n 2651) Age range NR	4 (Germany, Denmark, Poland, Bulgaria)	FFQ		X	No test of validity was performed, but the questionnaire was similar to other FFQ that had been tested for validity	No details^†
ENERGY⁽ ⁵¹ ^, ⁷⁷ ^, ¹⁰⁸ ⁾	Cross-sectional	Adults/parents or guardians (n 6002) Age range NR	7 (Belgium, Greece, Hungary, Netherlands, Norway, Slovenia, Spain)	Questionnaire with FFQ and 24-HDR^‡	X⁽ ⁵¹ ⁾	X	No details^†	The reliability and content validity of the parent questionnaires were tested separately in all participating countries, in five schools per country, using approximately 50 parents per country for the reliability study and 20 parents for the construct validity study Unpublished data	✓
EPIC⁽ ²⁹ ^, ⁴⁷ ⁴⁹ ^, ¹⁰⁹ ⁾	Longitudinal	Adults (n 519 978) 30–70 years	10 (Italy, Spain, Netherlands, Germany, Sweden(Malmo)/Sweden (Umea), Denmark, France, Greece, Norway, England)	FFQ, 24-HDR (EPIC-SOFT)	X⁽ ⁹¹ ⁾	X	Country-specific FFQ were tested for validity⁽ ⁹² ⁾. Data obtained from highly standardised 24-HDR, EPIC-SOFT, carried out in a random sample of each EPIC cohort, were used to account for differences in the FFQ, reducing the measurement error of the FFQ by calibration⁽ ¹⁰⁹ ⁾ EPIC-SOFT was tested for validity against biomarkers as part of the EFCOVAL project⁽ ⁹¹ ⁾ No data on SSB		✓
ESCAREL⁽ ¹¹⁰ ⁾	Cross-sectional	Adults (n 3187) 18–35 years	7 (France, Spain, Italy, UK, Finland, Latvia, Estonia)	FFQ^‡			All questionnaires were tested for validity in pilot studies	No details^†
Finbalt Health Monitor⁽ ⁶⁰ ⁾	Cross-sectional	Adults (n 25 044) 20–64 years	4 (Estonia, Finland, Latvia, Lithuania)	FFQ^‡			No details^†	No details^†
Finnish and Russian Karelia study⁽ ⁶⁵ ⁾ (as reported as for the 2002 study)	Cross-sectional	Adults (n 1201) 25–64 years	2 (Russia, Finland)	FFQ^‡			No details^†	No details^†
Food4Me⁽ ¹⁰³ ^– ¹⁰⁵ ⁾	Randomised controlled trial	Adults (n 5562) 17–79 years	7 (Ireland, Netherlands; Spain, Greece, UK, Poland, Germany)	FFQ (web-based)	X⁽ ¹⁰⁴ ^, ¹⁰⁵ ⁾	X	FFQ was tested for validity against 4d non-consecutive weighed records⁽ ¹⁰⁴ ⁾ using crude correlations and exact level of agreement Good agreement with 4d weighed food record Tested for validity against the EPIC-Norfolk FFQ⁽ ¹⁰⁵ ⁾ using energy-adjusted correlations, mean/median differences and exact level of agreement Good agreement with EPIC-Norfolk FFQ, which has been tested for validity Note: validation data available only for ‘other beverages’ ^§	Reproducibility of the FFQ has been tested⁽ ¹⁰⁴ ⁾	✓
HAPIEE⁽ ⁵⁸ ⁾*	Cross-sectional	Adults (n 28 947) 45–69 years	3 (Russia, Poland, Czech Republic)	FFQ^‡	X⁽ ⁹³ ^, ⁹⁴ ⁾		The FFQ was based on the Whitehall II questionnaire tested for validity by Brunner et al.⁽ ⁹³ ⁾ and Willett et al.⁽ ⁹⁴ ⁾ No data on SSB	No details^†
I.Family Project⁽ ⁶⁶ ^, ⁶⁹ ^, ⁷⁰ ⁾	Prospective cohort study (successor of IDEFICS study)	Adults/parents (n>7000) Age range not determined	8 (Belgium, Cyprus, Estonia, Germany Hungary, Italy, Spain, Sweden)	A diet questionnaire (FFQ) was included as part of the parent questionnaire Online 24-HDR (SACANA)	X	X	Instruments are similar to those used in the IDEFICS project and these were tested for validity	No details^†
Kolarzyk et al. ⁽ ¹⁰⁷ ⁾	Cross-sectional	Adults/students (n 1517) Age range NR	4 (Poland, Belarus, Russia, Lithuania)	FFQ^‡			The FFQ has been tested for validity and recommended by the National Food and Nutrition Institute in Warsaw, Poland	No details^†
MEDIS⁽ ⁶¹ ⁾	Cross-sectional	Adults (n 1190) 65–80+ years	2 (Cyprus, Greece)	FFQ	X⁽ ⁸⁹ ⁾		Tested for validity⁽ ⁹⁵ ⁾ No data on SSB	Tested for reproducibility⁽ ⁹⁵ ⁾ No data on SSB
MGSD⁽ ⁶² ⁾	Cross-sectional	Adults (n 4254) Non-diabetics (n 1833) 35–60 years	6 (Greece, Italy, Algeria, Bulgaria, Egypt, Yugoslavia (only diabetics in Yugoslavia))	Dietary history method using questionnaire	X⁽ ⁴⁰ ⁾		100 subjects from various participating centres were evaluated using the 3d diet diary (two weekdays and a Sunday) in order to test the questionnaire for validity No data on SSB	No details^†
SENECA⁽ ⁴² ^, ⁴³ ^, ⁵⁰ ^, ¹⁰² ⁾	Mixed design (longitudinal and cross-sectional)	Adults/elderly (n ~2600) 70–75 years	12 (Belgium, Denmark, France, Greece, Hungary, Italy, Netherlands, Norway, Poland, Portugal, Spain, Switzerland)	Modified dietary history method comprising a 3d estimated record and meal-based frequency checklist	X⁽ ⁵⁴ ^, ⁹⁶ ⁾	X	The cross-check dietary history method has been tested for validity⁽ ⁹⁶ ⁾ No data on SSB	No details^†	✓
ToyBox⁽ ⁸⁰ ^– ⁹⁰ ⁾	Intervention multifactorial study	Adults/parents or guardians (n 7056) (providing data at baseline)	6 (Belgium, Bulgaria, Germany, Greece, Poland, Spain)	Primary caregiver’s FFQ (PCQ)^‡	X⁽ ⁸³ ⁾	X	No details^†	Interval: 2 weeks Reliability of the PCQ was tested⁽ ⁸³ ⁾ Note: validity data available only for ‘water’ and ‘beverages’
Adolescents
HBSC 2009/10 survey⁽ ⁶⁷ ⁾	Cross-sectional	Adolescents (n 209 320) 11-, 13- and 15-year-olds	37 (England, Norway, Macedonia, Iceland, Netherlands, Portugal Wales, Italy, Sweden, Latvia, Switzerland, Denmark, Estonia Scotland, Slovenia, Ukraine, Belgium, Finland, Greece, Croatia, Hungary Lithuania, Poland, Germany, Greenland, Russia, Armenia, Austria, Belgium, Spain, France, Romania Turkey, Czech Republic, Ireland, Luxembourg, Slovakia)	FFQ^‡	X⁽ ⁹⁷ ⁾	X	Tested for validity against 7d diet record using crude correlations, mean/median differences and exact level of agreement⁽ ⁹⁷ ⁾ Moderate agreement for ‘soft drinks’	Interval: 7–15 d⁽ ⁹⁷ ⁾ Provides reproducible estimates of food group intake	✓
HELENA⁽ ⁴⁴ ^, ⁴⁵ ^, ⁵² ^– ⁵⁵ ^, ⁶⁸ ⁾	Cross-sectional	Adolescents (n 3000) 13–17 years	9 (Greece, Germany, Belgium, France, Hungary, Italy, Sweden, Austria, Spain) 8 countries used the 24-HDR (as above, except Hungary) Only 5 (Austria, Belgium, Greece, Sweden, Germany) pilot-tested the online FFQ	24-HDR using the HELENA-DIAT (Dietary Assessment Tool)^‡ Online FFQ	X⁽ ⁵² ^– ⁵⁴ ⁾ X⁽ ⁵⁴ ⁾	X X	24-HDR tested for validity against 1d food records and 24-HDR⁽ ⁵² ⁾ using crude correlations, mean/median differences and exact level of agreement Moderate to good agreement for ‘soft drinks’ Self-report 24-HDR was compared with interview-administered 24-HDR using crude correlations and mean/median differences, with good agreement⁽ ⁵³ ⁾ Online FFQ tested for validity against four 24-HDR using crude correlations, mean/median differences⁽ ⁵⁴ ⁾ Good agreement for ‘soft drinks’	Interval: 1–2 weeks HELENA FFQ has adequate reliability⁽ ⁵⁴ ⁾	✓
I.Family Project ⁽ ⁶⁶ ^, ⁶⁹ ^, ⁷⁰ ⁾†	Prospective cohort study (successor of the IDEFICS study)	Adolescents (n>9000 children of IDEFICs study and their siblings) 12–17 years	8 (Belgium, Cyprus, Estonia, Germany Hungary, Italy, Spain, Sweden)	A diet questionnaire (FFQ) was included as part of the teen questionnaire Online 24-HDR (SACANA)			The instruments are similar to those used and tested for validity as part of the IDEFICS project	No details^†
Larsson et al. ⁽ ⁷¹ ⁾	Cross-sectional	Adolescents (n 2041) Age range NR	2 (Sweden, Norway)	FFQ^‡			No details^†	No details^†
Szczepanska et al. ⁽ ⁷² ⁾	Cross-sectional	Adolescents (n 404) Age range NR	2 (Poland, Czech Republic)	FFQ			No details^†	No details^†
TEMPEST⁽ ⁷³ ^– ⁷⁵ ⁾	Cross-sectional	Adolescents (n 2764) 12–17 years	4 (Netherlands, Poland, UK, Portugal)	FFQ^‡		X	No details^†	No details^†
Children
Cinar and Murtomaa⁽ ⁷⁶ ⁾	Cross-sectional	Children (n 619) 10–12 years	2 (Turkey, Finland)	Youth FFQ^‡			No details^†	No details^†
ENERGY⁽ ⁵¹ ^, ⁷⁷ ⁾	Cross-sectional	Children (n 7234) 10–12 years	7 (Belgium, Greece, Hungary, Netherlands, Norway, Slovenia, Spain)	Questionnaire with FFQ and 24-HDR^‡	X⁽ ⁹⁸ ⁾	X	Construct validity assessed by comparing self-complete questionnaire against questionnaire completed by interview using exact level of agreement⁽ ⁹⁸ ⁾ Moderate to good construct validity for ‘soft drinks’	Interval: 1 week Moderate to good reliability	✓
EYHS⁽ ⁴⁶ ^, ⁴⁸ ^, ⁷⁸ ⁾	Cross-sectional	Children (n 4000) 9 and 15 years	4 (Denmark, Portugal, Estonia, Norway) (sourced studies involve only Denmark)	FFQ, 24-HDR, 1d qualitative food record	X	X	A 24-HDR preceded by a qualitative 1d food record has been found to be valid for generating estimates of children’s food intake for the purpose of group comparison⁽ ⁹⁹ ⁾ Test for validity was not conducted in European population	No details^†
IDEFICS⁽ ³⁹ ^, ⁴⁰ ^, ⁷⁹ ⁾	Prospective cohort study with an embedded intervention	Children (n 16 224) 2–9 years	8 (Belgium, Cyprus, Estonia, Germany Hungary, Italy, Spain, Sweden)	CEHQ-FFQ^‡ SACINA 24-HDR^‡	X⁽ ⁴⁰ ^, ⁵³ ^, ¹⁰¹ ^, ¹¹² ^– ¹¹⁴ ⁾	X	SACINA is based on the YANA-C instrument tested for validity as part of the HELENA study⁽ ⁵³ ⁾ SACINA was tested for validity by Börnhorst et al. ⁽ ¹¹⁴ ⁾ using the doubly labelled water technique No data on SSB The validity of the FFQ was assessed by Huybrechts et al. ⁽ ¹⁰¹ ⁾ No data on SSB CEHQ-FFQ was tested for validity against 24-HDR using crude correlations, mean/median differences and exact level of agreement⁽ ⁴⁰ ⁾ Low agreement for ‘soft drinks’	No fixed interval CEHQ-FFQ provides reproducible estimates of food group intake⁽ ¹¹² ⁾	✓
I.Family Project⁽ ⁶⁶ ^, ⁶⁹ ^, ⁷⁰ ⁾	Prospective cohort study (successor of the IDEFICS study)	Children (n>9000 children of IDEFICs study and their siblings) 2–11 years	8 (Belgium, Cyprus, Estonia, Germany Hungary, Italy, Spain, Sweden)	A diet questionnaire (FFQ) was included as part of the Children’s Questionnaire Online 24-HDR (SACANA)	X	X	Instruments are similar to those used and tested for validity as part of the IDEFICS project	No details^†
ISAAC⁽ ⁶³ ^, ⁶⁴ ⁾	Cross-sectional	Children (n ~63 000) 8–12 years	15 (Albania, France, Estonia, Germany, Georgia, Greece, Iceland, Italy, Latvia, Netherlands, Norway, Spain, Sweden, Turkey, UK)	FFQ as part of a general questionnaire		X	No details^†	No details^†
ToyBox⁽ ⁸⁰ ^– ⁹⁰ ⁾	Intervention multifactorial study	Children (n 7056) (providing data at baseline) 3·5–5·5 years	6 (Belgium, Bulgaria, Germany, Greece, Poland, Spain)	Children’s FFQ^‡	X⁽ ¹⁰⁰ ⁾	X	The children’s questionnaire is based on a FFQ previously tested for validity and developed by Huybrechts et al. ⁽ ¹⁰⁰ ⁾ Tested for validity by comparison with 3d diet record using crude and de-attenuated correlations, mean/median differences and exact level of agreement Moderate to good agreement for ‘sugared drinks’	Interval: at least 5 weeks FFQ provides reproducible estimates of food group intake	✓

CNSHS, Cross National Student Health Survey; ENERGY, EuropeaN Energy balance Research to prevent excessive weight Gain among Youth; EPIC, European Prospective Investigation into Cancer and Nutrition; ESCAREL, European Study in Non-Carious Cervical Lesions; HAPIEE, Health, Alcohol and Psychosocial factors in Eastern Europe; MEDIS, MEDiterranean Islands Study; MGSD, Mediterranean Group for the Study of Diabetes; SENECA, Survey in Europe on Nutrition and the Elderly; a Concerted Action; HBSC, Health Behaviour in School-aged Children; HELENA, Healthy Lifestyle in Europe by Nutrition in Adolescence; TEMPEST, ‘Temptations to Eat Moderated by Personal and Environmental Self-regulatory Tools’; EYHS, European Youth Heart Study; IDEFICS, Identification and prevention of Dietary-and lifestyle-induced health EFfects In Children and infantS; ISAAC, International Study of Asthma and Allergies in Childhood; NR, not reported; 24-HDR, 24 h recall; PCQ, Primary Caregiver’s Questionnaire; CEHQ, Children’s Eating Habits Questionnaire; EFCOVAL, European Food Consumption Validation; YANA-C, Young Adolescents’ Nutrition Assessment on Computer.

Funded by the Wellcome Trust programme grant entitled ‘Determinants of Cardiovascular Diseases in Eastern Europe: A multi-centre cohort study’ (reference number 064947/Z/01/Z) and developed by Martin Bobak, Anne Peasey, Hynek Pikhart (UCL), Ruzena Kubinova, Lubomíra Milla Novosibirsk, Sofia Malyutina, Oksana Bragina (Prague), Andrzej Pajak, Aleksandra Gilis-Januszewska (Krakow).

^†

Validity or reproducibility of the instrument was not reported in the article and no reference to validation or reproducibility studies was provided.

^‡

Original instrument was obtained for review.

^§

‘Other beverages’ includes everything except milk, alcoholic beverages, tea and coffee.