Chapter 12 — Exercises

Part D3 is the one worth doing even if you skip the rest — it asks you to check whether your reaction to this chapter came from the evidence or from somewhere else.

Difficulty legend: ⭐ basic recall · ⭐⭐ applied · ⭐⭐⭐ synthesis and judgment · ⭐⭐⭐⭐ extension

Selected answers in Appendix J.

Part A — Conceptual Understanding ⭐

A1. How many kcal per gram is ethanol? What storage form does the body have for it?

A2. Write out the metabolic pathway from ethanol to acetate. Which intermediate matters most, and why?

A3. What is the ALDH2 variant, roughly how common is it and in which populations, and what does it cause?

A4. Name three metabolic consequences of alcohol clearance being prioritized.

A5. Describe the J-curve. What does it appear to show?

A6. Define sick-quitter bias in your own words. Why does it push the abstainer group's risk up?

A7. Name the three problems with the J-curve, and say which is most damaging.

A8. What is Mendelian randomization, and why is the ALDH2 variant a good instrument for alcohol?

A9. List the cancers with established alcohol associations. Which one has a dose-response beginning at low intakes?

A10. What does "no safe level" precisely mean, and what does it not mean?

A11. Give the ethanol content of a US standard drink, a UK unit, and an Australian standard drink.

A12. Name five harms of alcohol that aren't cancer or cardiovascular disease.

Part B — Applied Analysis ⭐⭐

B1. Convert to grams of ethanol: (a) 175 ml wine at 13%; (b) a 568 ml pint at 5%; (c) 330 ml beer at 4.5%; (d) 50 ml spirits at 40%. Show the formula each time.

B2. Theo drinks 27 g of ethanol daily. Calculate his weekly grams, US standard drinks, UK units, weekly calories, and annual calories. Then express the annual figure in pounds of fat equivalent, with the caveat from Chapter 4 §4.6 about why that's an overestimate.

B3. A study compares moderate drinkers to non-drinkers and finds lower mortality in the drinkers, adjusted for age, sex, smoking, BMI, and income. Name the specific problems remaining, in order of severity.

B4. Explain why Mendelian randomization is immune to sick-quitter bias. Then name one assumption MR makes that could be violated.

B5. Someone says: "Alcohol is Group 1 like asbestos, so a glass of wine is like handling asbestos." Correct this precisely, without minimizing the real risk.

B6. Calculate the absolute breast cancer risk increase for a woman drinking two drinks a day, starting from a 12.5% baseline and using ~7–10% relative increase per 10 g/day. Show your working and state your assumptions.

B7. Two people drink 14 units a week — one daily, one all on Saturday. Compare their risk profiles and say what their sharing a study category does to the literature.

B8. A 72-year-old on lisinopril, metformin, and occasional paracetamol drinks three glasses of wine most evenings. List every concern from §12.7 and §12.11 that applies.

Part C — Skills & Practice ⭐⭐–⭐⭐⭐

C1. Do the alcohol audit. The Project Checkpoint: honest weekly count, converted to grams of ethanol, calories, percentage of intake, and annual cost. Then your own risk/benefit call, written down.

C2. Convert your drinks properly. Use the formula — ml × ABV% × 0.789 ÷ 100 — on the actual drinks you actually pour, including the size of glass you actually use. Most people find their "one glass" is 1.5–2 standard drinks.

C3. Price it. Weekly spend × 52. Compare to the annual cost of every other intervention in this book: Walt's supplements ($2,244), the metabolism aisle ($2,880), closing a fiber gap ($80).

C4. Track the cue. For one week, record not what you drank but what preceded it — arriving home, cooking, sitting down, a specific time, a specific person, a specific feeling. Identify your top two cues. That's the intervention point (§12.12).

C5. The two-week experiment. If you drink regularly, halve it for two weeks. Record: sleep quality, morning energy, weight trend, spend, and how difficult it was. This is a personal N-of-1 and it's rung 3 evidence — but it's rung 3 evidence about the only person whose decision this is.

C6. Audit the messaging. Find three pieces of media from the last five years making a health claim about moderate drinking or red wine. For each: what evidence is cited, is it observational or genetic, and is sick-quitter bias mentioned anywhere?

C7. Check the guidelines yourself. Look up your own country's current official guidance. When was it last revised? Does it use a threshold model or a continuum model? Compare it to the four in §12.9.

Part D — Synthesis & Critical Thinking ⭐⭐⭐

D1. Mendelian randomization disagrees with thirty years of observational data. Chapter 2 §2.10 says that's a serious signal. But MR has its own assumptions. Construct the strongest case that the observational data is right and MR is misleading here. Does it survive?

D2. The chapter explicitly refuses to tell you what to do, having pushed hard on fiber in Chapter 11. Is that inconsistent? The chapter itself says it's "not entirely comfortable" with the difference. Adjudicate — is the different posture justified by the evidence, or by the topic being socially loaded?

D3. The halted NIH trial is presented as showing both a real problem and a working correction mechanism. Which reading is more important? What should happen now — should someone fund it independently, or is the question adequately answered by MR?

D4. "No safe level" is technically defensible and reads as absolutist. Write a public health message that is equally accurate and less likely to be dismissed. Then explain why yours probably wouldn't be adopted.

D5. §12.8 argues that pleasure and social connection are legitimate inputs to the decision. Push back: is a nutrition text entitled to weigh enjoyment against cancer risk on a reader's behalf? Where should the line be between informing and deciding?

Part M — Mixed & Interleaved Practice ⭐⭐–⭐⭐⭐

M1. (With Chapter 2.) Sick-quitter bias, healthy-user bias, and exposure misclassification all appear in §12.4. Match each to the Chapter 2 section that introduced it, and explain how they compound rather than merely add.

M2. (With Chapter 9.) The J-curve's mechanism ran through HDL. Using Chapter 9 §9.10, explain why that mechanism was weaker than it appeared — and note that this was knowable before the MR studies.

M3. (With Chapter 3.) Theo's juice cleanse (Chapter 3, Case Study 1) changed eight things at once. Using this chapter, argue for which one was doing most of the work, and explain the mechanism.

M4. (With Chapters 4 and 6.) Explain how alcohol contributes to fat gain without being converted to fat. Name the Chapter 6 concept that makes this precise.

M5. (With Chapter 11.) Compare the absolute risk arithmetic for fiber (~1 fewer CRC case per 100) and alcohol (~1 extra breast cancer case per 75). Same shape, opposite direction. Did you find them equally persuasive? If not, why not?

M6. (With Chapter 10.) §12.12 recommends reduction over elimination. Connect this to Chapter 10 §10.9's reset-on-failure rule and §10.12's reversibility test.

M7. (With Chapter 8.) Why does alcohol matter more for an older adult than for a younger one? Name at least three mechanisms drawn from both chapters.

M8. (Revisit at Chapter 26.) Predict now: what will Chapter 26 say about alcohol and type 2 diabetes risk? Seal it.

Part E — Research & Extension ⭐⭐⭐⭐

E1. Find Mendelian randomization studies on alcohol and cardiovascular outcomes — both those using ALDH2 in East Asian populations and those using alcohol-related variants in large biobanks. What did they find, and what assumptions do the authors state?

E2. Read the 2018 Lancet Global Burden of Disease alcohol analysis, which concluded that the level of consumption minimizing health loss is zero. Then read at least two critical responses. Where does the disagreement sit — the data, the modelling, or the framing?

E3. Investigate the halted NIH moderate alcohol trial. Read the reporting and the institutional review findings. What exactly went wrong, at which stage, and what has changed as a result?

E4. Compare the 2023 Canadian (CCSA) guidance with the previous Canadian guidance and with your own country's. What evidence drove the revision, and how was it received? This is a live case of guidance moving substantially in one country and not others.

E5. Research the French paradox. What was claimed, what explanations have been proposed (including differences in death certification), and what is the current view? A good case study in a compelling narrative outliving its evidence.

E6. Look into the alcohol industry's role in funding research and shaping messaging — including the history of "drink responsibly" campaigns and industry-funded social aspects organizations. Then apply Chapter 1 §1.5's rule symmetrically: are there comparable funding influences on the abstinence side?