Who eats chocolate, and how much

Published analyses of dietary surveys, most of them American. They answer a narrower question than the headline figures do: of the people asked, how many ate chocolate on the days recorded, and how much.

7 studies2011 to 2021Each checked against its published abstract

What these studies add up to

On any given day most people eat no chocolate. About one American adult in eight ate chocolate candy on a recall day, and 11% reported any chocolate across two days.
Those who do eat about 40 g; the average over everyone is under 6 g. Which of the two a headline quotes changes it sevenfold.
Dark chocolate is a small minority: 1.4% of American adults reported it.
Cocoa is a minor source of the flavanols studied for health: about 3% of European adults' intake, against 62% from tea.
Chocolate eaters differ from non-eaters in many ways, which is why surveys linking chocolate to lower weight or better mood cannot show cause.

On a given day one US adult in eight ate chocolate candy: 5.7 g a head, about 40 g among eaters

O'Neil and colleagues, 2011. Nutrition Research. doi:10.1016/j.nutres.2011.01.007

Question: How much total, chocolate and sugar candy do US adults eat, and how does candy consumption relate to diet and health markers?
How: Dietary survey analysis: adults 19 years and older (n = 15,023) in the 1999-2004 National Health and Nutrition Examination Survey, intake determined from twenty-four-hour dietary recalls.
Finding: A total of 21.8%, 12.9% and 10.9% of adults consumed total, chocolate and sugar candy respectively on the recall day. Mean daily per capita intake was 9.0, 5.7 and 3.3 g; intake among consumers was 38.3, 39.9 and 28.9 g. Candy consumers had higher energy intake (9973 vs 9027 kJ) but lower body mass index (27.7 vs 28.2 kg/m2) than non-consumers.

Does not establish: The per capita figure is an average over mostly zeros, not what a typical person eats. The lower BMI among consumers does not show that candy is harmless or slimming; people who are overweight may avoid or under-report it.

Limits: A single 24-hour recall classifies people by one day's eating, so 'consumer' means 'ate it that day'. Recall data under-report snack foods. Cross-sectional. The abstract does not state the funding source.

On this site: How much chocolate people eat · Chocolate nutrition

US children averaged 4.8 g of chocolate candy a day and adolescents 7.0 g

O'Neil and colleagues, 2011. Food & Nutrition Research. doi:10.3402/fnr.v55i0.5794

Question: How much candy do US children and adolescents eat, and is it associated with diet quality, weight and cardiovascular risk markers?
How: Dietary survey analysis: children 2-13 years (n=7,049) and adolescents 14-18 years (n=4,132) in the 1999-2004 National Health and Nutrition Examination Survey; intake from twenty-four hour dietary recalls; diet quality by Healthy Eating Index-2005.
Finding: In younger children total, chocolate and sugar candy consumption was 11.4 g, 4.8 g and 6.6 g; in adolescents 13.0 g, 7.0 g and 5.9 g. Total candy consumers had higher energy intake (2248.9 vs 1993.1 kcals).

Does not establish: The study's conclusion that candy did not adversely affect health risk markers rests on a one-day snapshot and cannot show that habitual candy eating is without effect on children's weight or teeth.

Limits: Single-day recalls, reported by a parent for young children; cross-sectional; means include non-consumers.

On this site: How much chocolate people eat · Chocolate nutrition

How often US adults ate candy showed no link with obesity, with caveats from the authors

Murphy and colleagues, 2013. Nutrition Journal. doi:10.1186/1475-2891-12-53

Question: Is the typical frequency of candy consumption associated with body weight status and cardiovascular risk factors in US adults?
How: Cross-sectional survey analysis of adults in the 2003-2006 National Health and Nutrition Examination Surveys, classed as infrequent (3 eating occasions per month or fewer), moderate (up to 3.5 per week) or frequent (more than 3.5 per week) candy consumers from a questionnaire on chocolate and other candy over the previous 12 months. The abstract gives no sample size.
Finding: Frequency of candy consumption was not associated with the risk of obesity, elevated waist circumference, blood pressure, LDL or HDL cholesterol, triglycerides or insulin resistance. Increased frequency was associated with higher energy intakes and higher energy-adjusted intakes of total and added sugars and saturated fatty acids.

Does not establish: The authors state that it cannot be concluded that candy consumption does not cause obesity. The paper is an example of why a cross-sectional null finding on confectionery and weight is uninformative.

Limits: Frequency, not amount, was analysed; cross-sectional; the authors note that people who are overweight may eat less candy because of dieting or professional advice.

On this site: How much chocolate people eat · Chocolate nutrition

Cocoa supplies about 3% of the flavan-3-ols in European diets; tea supplies 62%

Vogiatzoglou and colleagues, 2014. The British Journal of Nutrition. doi:10.1017/s0007114513003930

Question: What is the habitual intake of flavan-3-ols in the European Union and what are the main food sources?
How: Dietary survey analysis using the EFSA Comprehensive European Food Consumption Database: adults aged 18-64 years from fourteen European countries, linked to a flavanol food composition database built for the study. The abstract gives no participant total.
Finding: Mean habitual intake of flavan-3-ol monomers, theaflavins and proanthocyanidins ranged from 181 mg/d (Czech Republic) to 793 mg/d (Ireland). Main sources were tea (62%), pome fruits (11%), berries (3%) and cocoa products (3%). The average habitual intake is considerably below the amounts used in most dietary intervention studies.

Does not establish: The study does not measure chocolate consumption itself. It shows that the flavanol doses used in trials are not reached by ordinary European diets, and that what flavanol people do get comes mainly from tea and fruit, not chocolate.

Limits: National surveys used different methods; food composition values for flavanols vary widely by product and processing.

On this site: How much chocolate people eat · Chocolate nutrition

In US dietary recalls 1.4% of adults reported eating dark chocolate

Jackson and colleagues, 2019. Depression and Anxiety. doi:10.1002/da.22950

Question: Is chocolate consumption associated with depressive symptoms in US adults?
How: Cross-sectional dietary survey analysis: 13,626 adults (20 years and over) in the National Health and Nutrition Examination Survey between 2007-08 and 2013-14; daily chocolate consumption from two 24-hr dietary recalls; depressive symptoms by PHQ-9 with scores of 10 or more counted as clinically relevant.
Finding: Overall, 11.1% of the population reported any chocolate consumption, with 1.4% reporting dark chocolate consumption. Dark chocolate consumers had lower odds of clinically relevant depressive symptoms (OR = 0.30, 95% CI 0.21 to 0.72); non-dark chocolate was not significantly associated. The highest quartile of chocolate consumption was 104-454 g/day.

Does not establish: The study cannot show that dark chocolate protects against depression. People who buy dark chocolate differ from others in income, education and diet, and the direction of cause is unknown, as the authors state.

Limits: Cross-sectional; the dark chocolate group is 1.4% of the sample; two recall days; people with depression may change what they eat.

On this site: How much chocolate people eat · Chocolate nutrition

Only 11.1% of US adults reported any chocolate on either of two recall days

Smith and colleagues, 2020. The American Journal of Medicine. doi:10.1016/j.amjmed.2020.01.020

Question: Is chocolate consumption associated with body mass index and waist circumference in US adults?
How: Cross-sectional dietary survey analysis: 13,626 nondiabetic adults (20 years and over) in the National Health and Nutrition Examination Survey, 5 study cycles from 2005-2006 through 2013-2014; chocolate consumption from 2 24-hour dietary recalls; BMI and waist measured.
Finding: Overall, 11.1% of the population self-reported any chocolate consumption in either of their 2 24-hour dietary recalls. Those who reported any chocolate had 0.92 kg/m2 (95% CI 0.53 to 1.32) lower BMI and 2.07 cm (95% CI 1.22 to 2.92) lower waist circumference than those who reported none.

Does not establish: The association with lower BMI does not show chocolate reduces body fat. A prospective cohort found the opposite direction over time, and the authors themselves say longitudinal or experimental work is needed to establish the direction of causation.

Limits: Cross-sectional; people with diabetes excluded; two recall days do not capture habitual intake of an occasional food.

On this site: How much chocolate people eat · Chocolate nutrition

By questionnaire, 46.4% of Korean adults counted as chocolate consumers

Kim and Jeon, 2021. Nutrition Research and Practice. doi:10.4162/nrp.2021.15.1.80

Question: Is chocolate consumption related to metabolic syndrome in Korean adults?
How: Cross-sectional survey analysis: 6,982 adults aged 19-64 years free of major chronic diseases in the 2014-2016 Korea National Health and Nutrition Examination Survey; chocolate consumption from food frequency questionnaires; non-consumers (n = 3,741, 53.6%) compared with consumers (n = 3,241, 46.4%).
Finding: Chocolate consumers had healthier lifestyles and dietary patterns and a lower prevalence of metabolic syndrome than non-consumers. No dose-response between chocolate intake and metabolic syndrome was found after excluding non-consumers. The abstract gives no intake in grams.

Does not establish: The share of 'consumers' here cannot be compared with the 11% to 13% found in US 24-hour recalls: a frequency questionnaire asks about habitual intake over a long period and a recall asks about one day. The study does not show chocolate lowers metabolic risk; consumers were healthier in other ways.

Limits: Cross-sectional; adults with major chronic disease were excluded; a food frequency questionnaire counts anyone who eats chocolate at all over the reference period.

On this site: How much chocolate people eat · Chocolate nutrition

What is still open

What intake looks like outside the United States, Europe and Korea. No analysis from elsewhere was found.
How much chocolate dietary surveys miss. Snacks are known to be under-reported; none of these studies measures by how much.
Whether the share of people eating dark chocolate has changed. The American data end in 2014.

These are questions the studies read for this page leave unanswered. They are not a claim that nobody has answered them.

All research areas